Top Read Articles

Published in last 1 year |

In last 2 years |

In last 3 years |

All

Please wait a minute...


For Selected: Download Citations EndNote Reference Manager ProCite BibTeX RefWorks Toggle Thumbnails

Select

Swin-Transformer-Based Carotid Ultrasound Image Plaque Segmentation HE Zhiqiang, SUN Zhanquan Electronic Science and Technology    2024, 37 (9): 48-56.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.008 Abstract （298）   HTML （10）    PDF（pc） （2281KB）（89）       Knowledge map    Save The evaluation of carotid ultrasound image plaque requires a large number of experienced clinicians, and the ultrasound image has the characteristics of blurred boundary and strong noise interference, making the evaluation of plaques time-consuming and laborious. Therefore, a fully automated carotid plaque segmentation method is urgently needed to solve the problem of manpower scarcity. This study proposes a deep neural network model based on Swin-Transformer (Shifted-Windows Transformer) block for the automatic segmentation of carotid plaques. Based on the U-Net(U-Convolutional Network) architecture, the encoding part uses three convolutional blocks for image down-sampling to obtain feature images of different resolution sizes, and then adds six pairs of two consecutive Swin-Transformer blocks for more refined feature extraction. The decoding part up-samples the refined features output by the Swin-Transformer module step by step, and jump-joints them with the feature maps of each resolution level in the encoding part, respectively. The comparison experiments based on the data set of Tong Ren Hospital show that the Dice index of the proposed deep neural network model reaches 0.814 2, which is higher than that of other comparison networks. The results demonstrate that the proposed model can effectively extract the features of carotid ultrasound image plaques and achieve automated and high-precision plaque segmentation. Table and Figures | Reference | Related Articles | Metrics

Select

Identification Strategy of Power Grid Weak Links Based on Random Matrix Theory YANG Jie, SUN Weiqing, MA Meiling Electronic Science and Technology    2024, 37 (7): 1-8.   DOI: 10.16180/j.cnki.issn1007-7820.2024.07.001 Abstract （293）   HTML （16）    PDF（pc） （2265KB）（145）       Knowledge map    Save With the continuous expansion of the scale of bulk power grid, the operating characteristics of new power system has become more and more complex. Accurately identifying the weak links in the power grid is of great practical significance for improving the monitoring ability of the system and ensuring its reliable operation. Therefore, an identification strategy of weak links for power grid is proposed based on random matrix theory. The strategy uses the measured data in power grid operation to construct a random matrix and uses the weak links judgment index based on the random matrix theory to design the judgment method of weak nodes and weak branches. This method is analyzed from the perspective of data correlation, without considering the complex network structure and operation mechanism of the actual power grid, so there is no complex modeling process. To verify the feasibility of the method, an example simulation is established through the IEEE 39 node system and compared with the traditional methods. The results show that the proposed identification strategy has a good effect on identifying weak nodes and branches, and the accuracy is improved compared with the traditional methods. Table and Figures | Reference | Related Articles | Metrics

Select

Particle Size Analysis in A Coupled Multi-Physics Models for Lithium-Ion Batteries YU Runzhou, LI Peichao Electronic Science and Technology    2024, 37 (9): 1-7.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.001 Abstract （279）   HTML （20）    PDF（pc） （1438KB）（101）       Knowledge map    Save In order to deeply understand the multi-physical field coupling behavior inside the LIB(Lithium-Ion Battery) and better provide reference for the manufacturing and optimization design of the LIB, a more physically realistic coupled ETM(Electrochemical-Thermal-Mechanical) model of the LIB is established and solved in the finite element simulation software COMSOL Multiphysics by means of numerical simulation in the present study. The model takes into account the stress generation in both electrode and particle scales during battery operation, which solves the problem of difficult calculation of stress at the electrode level in previous models, and better correlates the relationship between stress and electrochemistry by considering the correction of stress on lithium diffusion and overpotential. Based on this model, the effect of different positive electrode particle sizes on the battery performance is discussed in the study. The numerical results show that the performance index of each physical field during the discharge of LIB is better and the energy density of the battery is improved when the positive electrode particle size is small, which proves that the use of smaller positive electrode particle size can improve the performance of LIB. Table and Figures | Reference | Related Articles | Metrics

Select

Low Power Consumption Wearable Electrocardiogram Monitoring System ZHANG Peng, JIANG Mingfeng, LI Yang Electronic Science and Technology    2024, 37 (10): 88-94.   DOI: 10.16180/j.cnki.issn1007-7820.2024.10.012 Abstract （278）   HTML （6）    PDF（pc） （2442KB）（83）       Knowledge map    Save Daily electrocardiogram monitoring is of great significance for the prevention of cardiovascular diseases. However, medical-grade electrocardiogram monitoring systems are expensive, complicated to operate, and not suitable for home-based electrocardiogram monitoring. Wearable devices often ignore the interference of noise, and the signals collected are difficult to analyze and diagnose. This study presents a low-power wearable electrocardiogram monitoring system, which consists of four parts:Power management module, electrocardiogram acquisition module, data processing module, and delay switch module. The system uses the BMD101 chip to acquire human electrocardiogram signals and sends the electrocardiogram data to a mobile device via low-power Bluetooth chip nrf52832. To address the muscle electrical interference noise that is easily introduced into the electrocardiogram signals, a denoising algorithm based on VMD (Variational Mode Decomposition) is proposed. The NLM(Non-Local Mean) filter is used to remove the noise in the low-frequency mode of the electrocardiogram signal, and the DWT(Discrete Wavelet Transform) threshold denoising algorithm is used to eliminate the high-frequency mode noise of the electrocardiogram signal. The reconstructed signal quality is significantly improved. The experimental results show that the proposed algorithm has the characteristics of low cost, easy portability and convenient use, and can obtain high quality electrocardiogramsignals, meet the needs of users for long-range monitoring, and solve the problems of inconvenient daily electrocardiogrammonitoring and low quality electrocardiogram signal acquisition. Table and Figures | Reference | Related Articles | Metrics

Select

YOLOv3 Lung Nodule Detection Based on Coordinate Attention WANG Xinyu, ZHAO Jingwen, LIU Xiang, SHI Yunyu, SHE Yunlang Electronic Science and Technology    2024, 37 (6): 1-7.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.001 Abstract （272）   HTML （21）    PDF（pc） （1807KB）（107）       Knowledge map    Save Lung nodules occupy fewer pixels in CT(Computed Tomography), which brings great difficulty to detection. In view of small target detection of lung nodule, this study proposes YOLOv3(You Only Look Once version 3) lung nodule detection algorithm based on coordinate attention. The backbone network adopts the improved YOLOv3, reduces the number of residual blocks and introduces the dilated convolution module to sense context information around the target. In the feature utilization, coordinate attention is introduced to capture the position, direction and cross-channel information, so as to locate lung nodules accurately. The loss function of YOLOv3 is improved, the boundary box is modeled as Gaussian distribution. Wasserstein distance is used to calculate the similarity between the two distributions instead of IoU(Intersection over Union), so as to improve the sensitivity of the target scale. The results on LUNA16 show that the average precision is 89.96% and the sensitivity is 95.37%. Compared with mainstream target detection algorithms, the precision and sensitivity of the proposed method are improved by 11.33% and 9.03%, respectively. Table and Figures | Reference | Related Articles | Metrics

Select

Motion Planning of Manipulator Based on Improved SoftActor-Critic Algorithm TANG Chao, ZHANG Fan Electronic Science and Technology    2024, 37 (11): 47-54.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.007 Abstract （270）   HTML （11）    PDF（pc） （2620KB）（51）       Knowledge map    Save In view of the problems such as low exploration efficiency, slow convergence speed or even non-convergence of deep reinforcement learning algorithm in the motion planning task of robot arm under the requirement of high dimensional state space and high precision, this study introduces asynchronous advantage mechanism based on SAC(Soft Actor-Critic) algorithm, and proposes an AA-SAC(Asynchronous Advantage Soft Actor-Critic) algorithm integrating asynchronous advantage. This algorithm replaces the original V network with a Qtarget network,which effectively reduces the variance of the Q network. The n independent processes can be trained in parallel, which improves the training efficiency. The study also divides the experience playback pool of the AA-SAC algorithm into two parts, store and sample high-quality empirical data separately to improve the utilization of effective empirical data. The simulation results show that AA-SAC algorithm has the best performance in convergence speed, success rate and stability. Compared with the SAC algorithm, the convergence time of AA-SAC algorithm is 3 000 rounds earlier. After convergence, the success rate of AA-SAC algorithm reaches 96%, which is 6% higher than SAC algorithm and 26% higher than DDPG(Deep Deterministic Policy Gradient) algorithm. Table and Figures | Reference | Related Articles | Metrics

Select

Design of High Precision Over-Temperature Protection Circuit for Power Management Chip DU Wenhe, XU Zheng, KANG Jiahao, YANG Ke, PAN Jingxue Electronic Science and Technology    2024, 37 (9): 79-86.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.012 Abstract （268）   HTML （17）    PDF（pc） （1026KB）（106）       Knowledge map    Save Power management chips is damaged to varying degrees when they work at the ambient temperature beyond their acceptable range, and the over-temperature protection circuit plays an important role in improving the reliability and robustness of this kind of chip. This study designs a high-precision over-temperature protection circuit with the dual functions of turning off when the temperature is too high and reminding when the temperature is too low. The positive and negative temperature coefficient voltages are used to detect the chip temperature in real time, and then four logic turnover points are obtained by comparing them with different reference voltages at the output end of the band-gap reference circuit. After processing by the high-precision comparator circuit and hysteresis logic circuit, the hysteresis logic signals are output to control the working state of the chip or to remind the low temperature. The design and relevant simulation have been carried out based on 0.18 μm BCD(Bipolar-Complementary Metal Oxied Semiconductor-Double diffused Metal Oxide Semiconductor) process, and the simulation results show that when the power supply voltage ranges from 3.0~ 5.5 V, the maximum offset of the temperature threshold corresponding to the hysteresis logic turnover signal at the output end of the circuit is within 0.3 ℃. The circuit has high precision and can be widely integrated in various power management chips requiring over-temperature protection. Table and Figures | Reference | Related Articles | Metrics

Select

Irreversible Expansion Model for Lithium-Ion Batteries and its Application WANG Yahui, LI Peichao, WANG Keyong Electronic Science and Technology    2024, 37 (9): 8-13.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.002 Abstract （268）   HTML （8）    PDF（pc） （968KB）（56）       Knowledge map    Save In view of the problem of rapid estimation of capacity attenuation after multiple charge and discharge cycles of LIB(lithium-ion battery), a new method based on the capacity attenuation model of lithium-ion battery is proposed to rapidly estimate its internal capacity attenuation using the external expansion displacement of the battery.Based on the capacity attenuation model of LIB, the radial irreversible expansion model is derived, and the cylindrical LIB Sanyo UR18650E is modeled and solved by COMSOL Multiphysics, and the numerical results are compared with the experimental data, so as to verify the proposed model. Based on the above model, the causes of capacity attenuation and irreversible expansion of battery due to side reactions during the charge-discharge cycle are analyzed.The results show that the change of side reaction rate causes the gradient distribution of the concentration of side reaction products in the negative electrode, and the accumulation of side reaction products causes the expansion displacement of the battery to increase linearly with the cycle.The function formula of the cell capacity attenuation and its radial displacement is obtained by using the side reaction product as a bridge, which provides a new method for the rapid estimation of the capacity attenuation. Table and Figures | Reference | Related Articles | Metrics

Select

Research on Object Detection Based on Radar and Video Fusion ZHU Yong, HUANG Yongming, HE Xing Electronic Science and Technology    2024, 37 (8): 1-7.   DOI: 10.16180/j.cnki.issn1007-7820.2024.08.001 Abstract （256）   HTML （11）    PDF（pc） （3310KB）（91）       Knowledge map    Save The object detection based on video has the problem of poor recognition effect in bad weather, so it is necessary to make up for video defects and improve the robustness of detection framework. In view of this problem, this study designs an object detection framework based on radar and video fusion. YOLOv5 (You Only Look Once version 5) network is used to obtain image feature map and image detection frame, density-based clustering is used to obtain radar detection frame, and radar data is encoded to get object detection results based on radar information. Finally, the detection boxes of the two are superimposed to obtain a new ROI (Region of Interest), and the classification vector after fusion radar information is obtained, which improves the detection accuracy in extreme weather. The experimental results show that the mAP(mean Average Precision) of the framework reaches 60.07%, and the parameter number is only 7.64×106, which indicates that the framework has the characteristics of lightweight, fast computing speed and high robustness, and can be widely used in embedded and mobile platforms. Table and Figures | Reference | Related Articles | Metrics

Select

Research Progress in Network Security Situation Awareness Models FANG Xiang Electronic Science and Technology    2024, 37 (6): 98-102.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.013 Abstract （254）   HTML （6）    PDF（pc） （596KB）（75）       Knowledge map    Save In response to the increasing number and forms of network attacks, different types and names of network security situational awareness models have received widespread attention and research from the academic community. In the context of the rapid popularization of information technology, hardware system, software vulnerabilities, and security vulnerabilities in daily application operations have led to an increasing number of ways and means of network attacks. However, a single type of network security monitoring and analysis tool is no longer suitable for the current development of network technology. By reviewing the research history and current status of network security situational awareness technology, this article summarizes and analyzes the theoretical development and engineering applications of situational awareness models, and discusses the shortcomings and deficiencies in relevant technical solutions. It also looks forward to the future research directions of network security situational awareness models. Reference | Related Articles | Metrics

Select

Pavement Pothole Detection Method Based on Improved YOLOv5 HE Xing, HUANG Yongming, ZHU Yong Electronic Science and Technology    2024, 37 (7): 53-59.   DOI: 10.16180/j.cnki.issn1007-7820.2024.07.007 Abstract （253）   HTML （20）    PDF（pc） （2169KB）（117）       Knowledge map    Save Pothole is a common road disease, it reduce driving safety, accurate and rapid detection of potholes is more important.In viewof the problem that the detection accuracy of existing pothole detection methods is not high in the scenario of small targets and dense targets, an improved YOLOv5(You Only Look Once version 5) model is proposed in this study.TheCBAM(Convolutional Block Attention Module) is introduced into YOLOv5's backbone network to improve the model's ability to pay attention to key features. The loss function of YOLOv5 is changed to EIoU(Efficient Intersection over Union) to improve the detection accuracy of the model.The experimental results show that the proposed model can detect Potholes quickly and accurately in the scenarios of small targets and dense targets, and the mAP(mean Average Precision) in the open source Annotated Potholes Image Dataset reaches 82%. Compared with YOLOv5 and other mainstream methods, it is also improved. Table and Figures | Reference | Related Articles | Metrics

Select

Emotion Recognition Based on Multimodal Fusion of the EEG and Peripheral Physiological Signals MA Zhuang, GAN Kaiyu, YIN Zhong Electronic Science and Technology    2025, 38 (2): 62-69.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.008 Abstract （250）   HTML （8）    PDF（pc） （2024KB）（141）       Knowledge map    Save Decoding human internal emotional states based on EEG(Electroencephalogram) and surrounding physiological signals is key in the field of emotional computing, but the performance of machine learning models using EEG or surrounding physiological signal modes may be limited. In this study, a multi-mode fusion strategy is proposed based on the single mode method. The differential entropy, statistical and complexity features are extracted from each EEG fragment, and these features are properly integrated with the surrounding physiological signals. Multiple modal features recorded in the DEAP(Database for Emotion Analysis using Physiological Signals) data set are incorporated in the proposed method. In terms of titer, the experimental accuracy of single EEG feature is 49.21%, the classification accuracy of two types of feature fusion is 56.39%, 55.24% and 56.98%, and the experimental accuracy of three types of mode fusion is 56.98%. In terms of arousal, the experimental accuracy of single EEG feature is 49.34%, the classification accuracy of two types of feature fusion is 54.53%, 54.53% and 59.39%, and the experimental accuracy of three types of feature fusion is 55.48%. The experimental results show that the classification accuracy of multi-modal features after the fusion of EEG features and peripheral physiological features is the highest, and the classification accuracy is improved by 7.77% and 10.05%, respectively, compared with the single EEG features. Table and Figures | Reference | Related Articles | Metrics

Select

Deep Learning with Noisy Labels Based on Co-Teaching XIA Qiangqiang, LI Feifei Electronic Science and Technology    2024, 37 (11): 1-6.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.001 Abstract （232）   HTML （18）    PDF（pc） （792KB）（78）       Knowledge map    Save When large-scale data is labeled artificially, labeling errors are easy to occur, which leads to the existence of noise labels in data sets, and further affects the generalization of deep neural network models. The sample selection mechanism in the existing research methods such as Co-teaching makes the noise samples easy to flow into the selected clean label sample subset, and it is difficult to control the deep neural network model's fitting to the selected clean sample subset in training. Therefore, this study presents a novel algorithm that improves upon Co-teaching. In this method, two regularization losses are added to prevent the model from placing too much trust in a single class and falling into a local optimal solution respectively. Additionally, the introduction of high learning rate attenuation training method makes the model more inclined to learn clean label sample features in the initial training to get better model parameters. Compared with the results of Co-teaching, the performance of the proposed model is improved on MNIST, CIFAR-10 synthetic noise data set and Animal10N realistic data set under 20% and 50% symmetric noise and 45% asymmetric noise environment. Table and Figures | Reference | Related Articles | Metrics

Select

Research on AES and ECC Algorithm Image Encryption Based on FPGA FANG Yingli, FANG Yuming Electronic Science and Technology    2024, 37 (6): 92-97.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.012 Abstract （227）   HTML （18）    PDF（pc） （2019KB）（62）       Knowledge map    Save With the increasing use of digital images, it is essential to protect confidential image data from unauthorized access. In view of the security problems of digital image in the fields of communication, storage and transmission, this study proposes a digital envelope technology encryption scheme with high security and high speed based on the advantages of symmetric algorithm model and asymmetric algorithm model. This method is based on AES(Advanced Encryption Standard) and ECC(Elliptic Curve Cryptography), and optimized ECC hardware architecture is used for symmetric key sharing to enhance the security of the key. The traditional AES is optimized by adding pseudo-random numbers, using column shift instead of column obfuscation, and three-dimensional S-box box to maintain the Shannon diffusion and obfuscation principle while reducing the time complexity. This study the digital image encryption simulation and performance test of AES algorithm are carried out on FPGA(Field Programmable Gate Array). The test results show that the proposed encryption scheme has the advantages of rapidity, high security and effectiveness, and can better achieve image encryption. Table and Figures | Reference | Related Articles | Metrics

Select

Dynamic Face Recognition System Design Based on RetinaFace and FaceNet LI Yunpeng, XI Zhihong Electronic Science and Technology    2024, 37 (12): 79-86.   DOI: 10.16180/j.cnki.issn1007-7820.2024.12.012 Abstract （226）   HTML （11）    PDF（pc） （2206KB）（56）       Knowledge map    Save This study proposes a dynamic face recognition system to address the problem of requiring the recognized individual's cooperation in existing static face recognition processes. The system uses the RetinaFace and FaceNet algorithms for dynamic face detection and recognition, respectively, and is optimized for high recognition accuracy and real-time performance. In particular, GhostNet is used as the backbone network for RetinaFace detection, and Adaptive-NMS(Non Max Suppression) non-maximum suppression is used for face bounding box regression. For FaceNet recognition, MobileNetV1 is used as the backbone network, and a joint loss function combining Triplet loss and cross-entropy loss is used for face classification. The optimized algorithm has excellent performance in detection and recognition The improved RetinaFace algorithm achieves detection accuracies of 93.35%, 90.84%, and 80.43% on the WiderFace dataset, with a frame rate of 53 frame·s-1. For dynamic face detection, the average detection accuracy is 96%, with a frame rate of 21 frame·s-1. When the FaceNet threshold is set to 1.15, the highest recognition rate is 98.23%. The average recognition accuracy of the dynamic recognition system is 98%, with a frame rate of 20 frame·s-1. The experimental results demonstrate that the proposed system fully addresses the problem of requiring cooperation from the recognized individual in static face recognition and achieves high recognition accuracy and real-time performance. Table and Figures | Reference | Related Articles | Metrics

Select

Quality Guidance Based Branch-Cut Phase Unwrapping Algorithm TAI Manli, LI Wenguo, LIU Tao, ZHONG Yongpeng Electronic Science and Technology    2024, 37 (7): 72-80.   DOI: 10.16180/j.cnki.issn1007-7820.2024.07.010 Abstract （218）   HTML （5）    PDF（pc） （4316KB）（83）       Knowledge map    Save Phase unwrapping is a key data processing step in phase measurement profilometry. In this study, a quality-guided branch-cut phase unwrapping algorithm is proposed based on the traditional Goldstein branch-cut method, aiming to obtain more accurate phase unwrapping results. The proposed algorithm uses the second-order differential of the wrapped phase as a supplement parameter for residual detection principle, and regards the mutation points in the second-order differential of the wrapped phase as non-polar residual points, and uses modulation as the criterion for residual point effectiveness judgment, regarding the residual points with low modulation as having higher effectiveness for local optimization to reduce the density of residual points. The optimized equivalent residual points are connected into branch-cut to block error propagation. The proposed algorithm calculates the quality of pixels using modulation to guide the order of phase unwrapping, with the unwrapping path circumventing the branch-cut and prioritizing high-quality pixels. Experimental results show that the proposed algorithm has higher accuracy and competitiveness in phase unwrapping results. Table and Figures | Reference | Related Articles | Metrics

Select

Pulse High-Frequency Voltage Injection Method Based on Rotor Polarity Judgment Optimization CHEN Yibin, JIN Hai, XU Shen Electronic Science and Technology    2024, 37 (6): 29-35.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.004 Abstract （217）   HTML （5）    PDF（pc） （1348KB）（77）       Knowledge map    Save When the permanent magnet synchronous motor operates at zero speed and low speed, the rotor position information can be accurately estimated by the pulse high-frequency injection method. The pulse high frequency injection method can not realize the rotor polarity judgment, while the traditional polarity judgment method has problems such as long response time, complex sampling and implementation process, and even polarity misjudgment. In this study, a method of judging the rotor polarity based on positive and negative voltage pulse injection is proposed. This method injects positive and negative voltage pulses into the motor, and judges the rotor polarity according to the response current peak value of the corresponding pulse. The pulse high frequency injection method combines the rotor position estimation with the polarity judgment result to obtain the accurate initial position of the rotor. This study verifies the effectiveness of the algorithm through Simulink simulation. The simulation results show that the proposed rotor polarity determination method can simplify the sampling method and algorithm complexity, reduce the response time for determining rotor polarity, and enable the pulse high-frequency injection method to accurately and quickly estimate the initial position of the rotor. Table and Figures | Reference | Related Articles | Metrics

Select

Bi-Level Optimal Scheduling Strategy of Integrated Energy System Considering EV SHAO Wenfeng, HE Yu, WEN Yongjun, NIE Xianglun, ZHANG Tangqian, KAN Chao Electronic Science and Technology    2024, 37 (11): 85-94.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.012 Abstract （216）   HTML （7）    PDF（pc） （4177KB）（54）       Knowledge map    Save In order to solve the optimal scheduling problem of large-scale electric vehicles into the network, a two-layer optimal scheduling strategy of integrated energy system with electric vehicles is proposed in this study. The upper layer is the optimal dispatching layer, in which electric vehicle agents group electric vehicles into clusters according to the dispatchable time and upload the cluster information to the system dispatching center, which cooperates with electric vehicle clusters and energy systems and builds an economic dispatching model with the goal of minimizing the dispatching cost by considering integrated demand response and ladder-type carbon trading mechanism. The lower layer is the power allocation layer, where electric vehicle agents build a power allocation model with the goal of satisfying users' travel demand, and guide electric vehicles to participate in system scheduling in an orderly manner. The simulation algorithm is constructed and solved by using CPLEX solver. The simulation results show that the proposed strategy can not only effectively reduce the scheduling cost of integrated energy system, smooth the system load curves and reduce carbon emissions, but also significantly reduce the cost of electricity consumption of customers on the basis of securing their travel demand, thus achieving a win-win situation for both supply and demand. Table and Figures | Reference | Related Articles | Metrics

Select

Omnidirectional Gait Generation Method for Biped Robot with Fusion of Imitation Learning FENG Zhen, MOU Haiming, XUE Jie, LI Qingdu Electronic Science and Technology    2025, 38 (1): 29-36.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.005 Abstract （214）   HTML （6）    PDF（pc） （2696KB）（49）       Knowledge map    Save Due to the complex high-dimensional dynamics and highly dynamic characteristics of bipedal robots, achieving omnidirectional gait is a difficult problem. In order to achieve omnidirectional walking of bipedal robots, this study proposes a gait training method of biped robot based on deep reinforcement learning. Based on expert experience and the periodicity of bipedal walking, periodic symmetric functions that can achieve different gait styles are designed for imitation learning. In order to make the bipedal robot capable of omnidirectional walking, the footstep planner in ROS (Robot Operating System) is used to generate target foothold points for imitation learning. The proposed method is validated on a self-designed bipedal robot. The experimental results show that the proposed method can realize four gait modes of biped robot including forward, side, diagonal and turn, and realize omnidirectional gait of biped robot, and can realize different styles of cycles. Table and Figures | Reference | Related Articles | Metrics

Select

Real-Time Hybrid Task Scheduling Algorithm in Embedded Multicore System LUO Guang, MAO Hang, ZHU Yangshuo, ZHANG Fengdeng Electronic Science and Technology    2024, 37 (8): 84-91.   DOI: 10.16180/j.cnki.issn1007-7820.2024.08.012 Abstract （214）   HTML （6）    PDF（pc） （996KB）（81）       Knowledge map    Save In this study, an algorithm based on BFZL(Boundary Fair until Zero Laxit) is proposed to solve the problem of reasonable scheduling of real-time mixed task set formed by periodic task and sporadic task. Based on the I-BF(Improved Boundary Fair) real-time mixed task algorithm, the relaxation parameter of LLF(Least Laxity First) algorithm is introduced to improve the priority of decision task. A heuristic algorithm based on relaxation and heuristic strategy is proposed to improve task assignment strategy. The experimental results show that the BFZL algorithm can satisfy the real-time performance of the system and achieve the purpose of algorithm optimization. Through data comparative analysis, compared with the original algorithm, the proposed algorithm reduces the average response time of sporadic tasks by about 26%, reduces the context switch and migration by about 28% and 50%, respectively. Additionally, the algorithm also has advantages in scheduling overhead. Table and Figures | Reference | Related Articles | Metrics

Select

A Transformerless Half-Bridge Lithium Battery Equalizer Based on Phase-Shift Strategy ZHOU Bin, KAN Jiarong, CHEN Heming, CHEN Weiwei, LI Yan, XU Sudong Electronic Science and Technology    2025, 38 (2): 1-9.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.001 Abstract （211）   HTML （39）    PDF（pc） （4454KB）（172）       Knowledge map    Save In view of the inconsistency between the cells of lithium battery and the difficulty and high cost of making the traditional half-bridge equalizer multi-winding transformer, a transformerless half-bridge lithium battery equalizer is proposed in this study, and the key factors affecting the energy flow between cells at high frequency are analyzed. Due to the use of phase-shifting control, shortest equalization path, and parallel equalization mode, the equalizer can quickly achieve simultaneous equalization between multiple battery cells. The equalization speed is fast, the equalization efficiency is high, and it is not affected by the number of battery cells in series,which makes the equalizer have the advantages of simple control, high equalization efficiency, high flexibility, and fewer switching devices. Based on the dynamic model of battery charging and discharging, an expression for energy balance between individual cells of the battery is derived, and verified by building a simulation model and experimental prototype. The simulation and experimental results show that the transformer free half bridge lithium battery equalizer based on phase shifting strategy can quickly achieve voltage equalization between battery cells. Table and Figures | Reference | Related Articles | Metrics

Select

Intelligent Path Planning Based on Ant Colony Algorithm TONG Yunhao, XI Zhihong Electronic Science and Technology    2025, 38 (1): 23-28.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.004 Abstract （210）   HTML （16）    PDF（pc） （3460KB）（81）       Knowledge map    Save In view of the problem that it is difficult to reasonably plan the path after the mobile robot completes its self-positioning and map construction, which leads to the disordered movement of the mobile robot and the waste of resources, ant colony algorithm is adopted to realize the path planning of mobile robot in this study. Ant colony algorithm is a probabilistic algorithm to solve the optimal path in a problem. However, in the general ant colony algorithm, all parameters of the ant colony algorithm are unchanged, resulting in the result of the ant colony algorithm dependent on the pheromone parameters set in the algorithm. In order to solve the above problems, the parameters of ant colony algorithm and pheromone allocation are improved, and the pheromone update standard is improved by changing the pheromone volatility coefficient and pheromone update standard in each iteration and combining with heuristic factors. Setting the adjustable pheromone volatile factor increases the adaptability of the algorithm. According to the meaningful parameter space, the path planning results of the traditional ant colony algorithm and the improved ant colony algorithm are compared under different environments. The path length of the improved ant colony algorithm is reduced by 4.48% and 8.54%, respectively, and no path crossover nodes are generated, which achieves the expected effect of reasonable path planning for mobile robots. Table and Figures | Reference | Related Articles | Metrics

Select

Quadcopter Sliding Mode Control Based on New Approximation Law ZHANG Niu, ZHOU Ying Electronic Science and Technology    2024, 37 (12): 1-8.   DOI: 10.16180/j.cnki.issn1007-7820.2024.12.001 Abstract （203）   HTML （17）    PDF（pc） （1210KB）（125）       Knowledge map    Save To solve the trajectory tracking and attitude control problems of quadcopter aircraft during flight, a sliding mode variable structure nonlinear controller is designed based on an improved new reaching law and sliding mode function. According to the underactuation, strong coupling and multi-variable characteristics of quadcopter aircraft, the whole force analysis of its system is carried out, and the system mathematical model is simplified based on Newton's second law and Euler's momentum equation. Based on the traditional exponential reaching law and combined with the double-power term, proportional term and improved variable exponential term, a new reaching law sliding mode controller is designed for the inner and outer loops of the system. The output of the outer loop position controller is taken as the expected input of the inner loop controller to form a closed loop system in which the outer loop controls the inner loop, and the stability of the system is proved by the Lyapunov stability theorem. Simulation results show that the proposed method has shorter response time, better tracking performance, and smaller steady-state error than traditional reaching law sliding mode controllers in fixed-point levitation and complex three-dimensional trajectory tracking experiments. Table and Figures | Reference | Related Articles | Metrics

Select

Improvement of YOLOv5s Algorithm for Steel Surface Defect Detection CUI Jingnan, HUANG Chunyan, LI Yanling Electronic Science and Technology    2024, 37 (12): 48-55.   DOI: 10.16180/j.cnki.issn1007-7820.2024.12.008 Abstract （203）   HTML （16）    PDF（pc） （4485KB）（73）       Knowledge map    Save In view of the problems such as low accuracy and slow recognition speed of existing steel surface defect detection methods, a defect detection method based on improved YOLOv5s(You Only Look Once version 5s) is proposed in this study. CBAM(Convolutional Block Attention Module) attention mechanism is introduced into the backbone feature extraction network to pay attention to the information of important regions and improve the model's learning ability of target defects. In order to improve the regression speed and localization accuracy of the target frame, the EIoU(Efficient Intersection over Union) boundary frame loss function combining distance loss and width-height loss are used to calculate the loss value. Transfer learning is used to accelerate the convergence speed of the model to improve the accuracy of defects detection. The experimental results on the data set NEU-DET show that compared with the original YOLOv5s network, the improved YOLOv5s network of the accuracy rate of the data set increased by 6.3 percentage point, the recall rate increased by 9.2 percentage point, and the mAP(mean Average Precision) reached 81.7%, which indicates the improved method has good performance for the detection of steel surface defects. The size of the steel surface defect detection model with the improved YOLOv5s algorithm is only 13.8 MB, which improves the detection accuracy on the basis of real-time performance and facilitates the deployment of the model in practical applications. Table and Figures | Reference | Related Articles | Metrics

Select

Intelligent Equalization Method of Battery Pack Based on Two-Level Equalization Circuit REN Zihao, TIAN Engang Electronic Science and Technology    2024, 37 (7): 9-15.   DOI: 10.16180/j.cnki.issn1007-7820.2024.07.002 Abstract （203）   HTML （9）    PDF（pc） （850KB）（69）       Knowledge map    Save In order to solve the problem of low efficiency of traditional balanced electric topology, a two-level balanced topology is proposed. This balancing topology divides the battery pack into two forms: Intra-group and inter-group. Buck-Boost equalization circuit is used in the group, and reconfigurable equalization circuit is used between the groups. The intra-group and inter-group can be equalized at the same time, which improves the equalization efficiency. Taking SOC(State of Charge) as the equalization variable, the intra-group equalization algorithm adopts the fuzzy logic control strategy based on SOC to reduce the equalization time and improve the equalization efficiency. MATLAB/Simulink software is used to model and simulate the circuit topology and the results are compared with the traditional Buck-Boost circuit. The simulation results show that compared with the traditional Buck-Boost circuit, the proposed algorithm and equalization topology reduce the equalization time by about 28% under the charge-discharge state, indicating that the equalization circuit and algorithm have good performance. Table and Figures | Reference | Related Articles | Metrics

Select

Heterogeneous Converged Network Access Algorithm Based on Deep Reinforcement Learning XIAO Xiong, LIU Hongyan, YI Mengjie Electronic Science and Technology    2024, 37 (11): 7-12.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.002 Abstract （199）   HTML （10）    PDF（pc） （2468KB）（60）       Knowledge map    Save With the increasingly mature communication networks in the fields of air, space and ground, cross-domain heterogeneous converged technology has become an important direction for the integrated development of future communication networks. Driven by the demand for cross-domain heterogeneous in the converged network of air, space and ground, this study aims to solve the problem of low spectrum resource utilization in heterogeneous networks. It uses deep reinforcement learning method to establish a heterogeneous converged network system model and designs intelligent agent access algorithm with fair scale. The system throughput is selected as the maximization objective. The communication network standards that meet the characteristics of air, space and ground integration are selected and corresponding access protocols are extracted. Non-dimensional channel parameters are set according to the principle of fairness and simulation scenarios are established. Multiple comparison strategies are introduced in the simulation to statistically analyze system throughput, collision rate, utilization rate and channel selection ratio. The simulation results show that the system throughput of cross-domain heterogeneous fusion network is increased by more than 60%, system channel utilization efficiency is increased by 20%, and the collision rate of service packets is maintained at 10%, which verifies the adaptability of the algorithm to different business scenarios. Table and Figures | Reference | Related Articles | Metrics

Select

Fault Identification of Complex Analog Circuit Based on Deep Learning HUANG Zehua, BI Guihong, ZHANG Zirui Electronic Science and Technology    2025, 38 (2): 23-34.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.004 Abstract （199）   HTML （16）    PDF（pc） （8538KB）（115）       Knowledge map    Save For complex analog circuits with complex fault transfer relationships and complex nonlinear relationships between fault types and fault features, which cause difficulties in feature extraction and fault identification. This study presents a fault diagnosis method for analog circuits based on two measuring points-CEEMDAN(Complete Ensemble Empirical Mode Decomposition with Adaptive Noise)-multi-scale false-color image-ALEXNet network. A new combined fault sample is constructed by connecting the output signals of two measuring points in a complex analog circuit. The combined fault sample data of two measuring points improves the ability to characterize the whole fault state of the complex analog circuit. The combined fault sample signals are decomposed in multi-scale, and the multi-scale data matrix is mapped to a two-dimensional false-color graph to form a multi-scale false-color image with abundant information and obvious features. Based on AlexNet's excellent image feature mining and learning ability, multi-scale false-color images of different fault types were input into AlexNet for model transfer training and fault identification. By comparing and analyzing the single and double faults and mixed faults of simple circuit and complex circuit, it is proved that the proposed method can achieve higher recognition accuracy for different fault types of complex analog circuit. Table and Figures | Reference | Related Articles | Metrics

Select

A Vehicle Detection Algorithm Based on Improved YOLOv4 LAI Ying, JU Zhiyong, YE Yuxin Electronic Science and Technology    2025, 38 (1): 81-87.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.011 Abstract （197）   HTML （21）    PDF（pc） （1202KB）（75）       Knowledge map    Save In the process of vehicle detection in traffic monitoring, there are some problems such as vehicles shielding each other and insufficient distance target size, which leads to missing detection and false detection. To solve this problem, this study proposes a traffic vehicle detection algorithm based on YOLOv4(You Only Look Once version 4) multi-scale fusion and attention mechanism. A new feature layer is added to YOLOv4's path aggregation network for multi-scale feature fusion to improve the model's ability to extract underlying texture features. The ECA (Efficient Channel Attention) channel attention module is embedded in front of YOLO Head detection head to reasonably suppress and enhance the aggregated features. The CIoU (Complete Intersection over Union) loss function is replaced by the Soft-CIoU loss function to improve the contribution of small target vehicles to the loss function. The experimental results on the publicly available vehicle data sets UA-DETRAC and KITTI show that compared to the original YOLOv4 algorithm, the average accuracy of the proposed algorithm improves by 2.45 percentage points and 1.14 percentage points, respectively, and the detection speed reaches 41.67 frame·s-1.The proposed algorithm performs well in detection accuracy when compared with other advanced algorithms. Table and Figures | Reference | Related Articles | Metrics

Select

Image Style Transfer Algorithm Based on Improved Generative Adversarial Network WANG Shengxiong, LIU Ruian, YAN Da Electronic Science and Technology    2024, 37 (6): 36-43.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.005 Abstract （193）   HTML （19）    PDF（pc） （3584KB）（71）       Knowledge map    Save Image style transfer has been a research hotspot in the field of image processing, but the current style transfer models have problems such as fuzzy details of generated image, the poor color effect of style texture and excessive model parameters. An image style transfer method based on improved cycle-consistent generative adversarial network is proposed in this study. The generator network architecture is improved by adding the Ghost convolution module and the inverted residual improved module to reduce the number of model parameters and computation cost, as well as enhance the feature extraction capability of the network. And the content style loss item, the color reconstruction loss item and the map identity loss item are added to the loss function for enhancing generative capability of the model and improving the quality of generated images. The experimental results show that the proposed method has a stronger ability for style transfer, which enhances the content details and color effect of style texture from generated images effectively, improves the image quality significantly, and the model performance has also been improved well. Table and Figures | Reference | Related Articles | Metrics

Select

Hybrid Image Super-Resolution Reconstruction with Multiple and Multi-Scale Attention KUAI Xinchen, LI Ye Electronic Science and Technology    2024, 37 (9): 34-42.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.006 Abstract （193）   HTML （17）    PDF（pc） （2627KB）（74）       Knowledge map    Save Image itself information is naturally robust to image reconstruction, yet most current super-resolution methods do not fully utilize global feature information. This study proposes a new image super-resolution model mixing multiple and multi-scale attentions, including two new modules: Multi-scale hybrid non-local attention upsampling module and residual dense attention block. Different from previous nonlocal methods, multi-scale hybrid non-local attention upsampling module mixes pixel-based and patch-based nonlocal attention and establishes patch-level upsampling mapping relationships at multiple scales, which enables a wider global search space. The residual dense attention block establishes attention associations in channel and spatial dimensions, which enhances the transfer and fusion of front-to-back attention information through dense connections. In this study, quantitative and qualitative evaluations are conducted on several benchmark datasets, and the experimental results show that the model outperforms similar super-resolution models in terms of performance and reconstruction quality. Table and Figures | Reference | Related Articles | Metrics

Select

Grading and Diagnostic Method for Colorectal Cancer Immunohistochemical Images MO Zhuorui, HUANG Qianghao, ZHANG Lin, CAO Yuqi, GE Weiting, YU Minghui Electronic Science and Technology    2024, 37 (12): 24-31.   DOI: 10.16180/j.cnki.issn1007-7820.2024.12.004 Abstract （190）   HTML （4）    PDF（pc） （4921KB）（44）       Knowledge map    Save Human tissue pathology examination is mainly used for the diagnosis and treatment of various tumors. Immunohistochemical technique has important clinical significance in the early screening of colorectal cancer. In order to accurately determine the expression level of the tumor suppressor gene p53 in colorectal cancer, this study proposes a grading diagnostic method based on transfer learning with block-wise fine-tuning strategy. The parameters of the cell nucleus segmentation model are transferred to the diagnostic framework through image preprocessing, supervised model pre-training, and block-wise fine-tuning. The generated cell nucleus segmentation mask is subjected to PCA(Principal Component Analysis) dimensionality reduction and SVM(Support Vector Machine) multivariate classification to obtain the final image diagnosis result. The proposed method is verified on colorectal cancer p53 protein IHC(Immunohistochemistry) image dataset. Dice value of the model reaches 0.887 6 and classification accuracy reaches 90.28%. The results show that the proposed method can effectively grade the immunohistochemical images of colorectal cancer, and provide reliable auxiliary information for doctors to read the film. Table and Figures | Reference | Related Articles | Metrics

Select

Multi-Beam Lidar Rotary Mirror Scanning System LIN Jiandong, SONG Yue Electronic Science and Technology    2025, 38 (1): 52-58.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.008 Abstract （188）   HTML （3）    PDF（pc） （1848KB）（47）       Knowledge map    Save To solve the problem of low spatial resolution of remote target detection caused by fixed angle resolution scanning of traditional multi-line lidar, an eight-line lidar scanning system with symmetrical arrangement of two light sources is proposed. The parameters of the rotating mirror and light source and the scanning trajectory of the beam are studied using the ray vector method, and the parameters of the eight-line lidar are optimized by combining the lidar equation. By strictly calculating the beam trajectories scanned by the mirror with different inclination angles, two light sources are symmetrically arranged on both sides of the eight-sided mirror for time-sharing luminescence detection to realize the eight-sided mirror rotation scanning system with high spatial resolution. The vector calculation results show that there are two scanning detection beams in front of the human body at a distance of 100 m, and there are still three scanning detection beams in the side and edge of the human body, indicating that the proposed eight-sided mirror rotating scanning system can detect long-distance targets more effectively using scanning beams, which is better than the Velodyne 64-line lidar. Table and Figures | Reference | Related Articles | Metrics

Select

Electromagnetic Signal Modulation Recognition Based on Complex-Valued Deep Neural Network YUAN Depin, ZHAO Liang, GE Xiansheng Electronic Science and Technology    2025, 38 (3): 1-6.   DOI: 10.16180/j.cnki.issn1007-7820.2025.03.001 Abstract （188）   HTML （50）    PDF（pc） （2836KB）（182）       Knowledge map    Save In the region of complex electromagnetic environment, it is difficult to obtain the signal modulation type. The traditional recognition and classification methods of modulated signals are not successful because of their own defects. The current deep learning methods usually used for signal modulation are based on real values for characterization and processing, which results in recognition bias due to the loss of the original intrinsic connection of complex values. To solve this situation, the complex deep neural network is applied to the modulation recognition of electromagnetic signals, complex convolutional deep neural networks such as complex convolutional deep neural networks, batch normalization and fully connected networks are designed, and the final classification task is completed by softmax function. The standard data set RML2016.10a is used to complete the training as well as testing of the network. The experimental results show that the trained complex deep neural network is significantly better than traditional recognition algorithms, and can effectively improve the recognition rate of electromagnetic signals. Table and Figures | Reference | Related Articles | Metrics

Select

Emotion Recognition Algorithm Based on Multimodal Cross-Interaction ZHANG Hui, LI Feifei Electronic Science and Technology    2024, 37 (10): 81-87.   DOI: 10.16180/j.cnki.issn1007-7820.2024.10.011 Abstract （187）   HTML （7）    PDF（pc） （1735KB）（59）       Knowledge map    Save Due to the limitations of single modality emotion recognition, many researchers have shifted their focus to the field of multimodal emotion recognition. Multi-modal emotion recognition focuses on two problems: The optimal extraction of the features of each mode and the effective fusion of the extracted features. This study proposes an emotion recognition method based on multimodal cross-interaction to capture the diversity of modality expressions. The editors of various modalities separately extract features with emotional information, and the stacked interaction modules based on the attention mechanism between modalities model the potential relationship among vision, text and audio. Experiments are conducted on CMU-MOSI and CMU-MOSEI datasets for emotion recognition based on text, audio and visual. The results show that the method achieved the scores of 86.5%, 47.7%, 86.4%, 0.718, 0.776, and 83.4%, 51.5%, 83.4%, 0.566, 0.737 on five indicators, Acc2(Accuracy2)、Acc7(Accuracy7)、F1、MAE(Mean Absolute Error) and Corr(Correlation). This demonstrates that the proposed algorithm significantly improves performance, and also validates that the cross-mapping mutual representation mechanism perform better than single-modal representation methods. Table and Figures | Reference | Related Articles | Metrics

Select

Overview of Research on Network Security Situation Prediction Technology LU Zhenyang Electronic Science and Technology    2024, 37 (8): 92-96.   DOI: 10.16180/j.cnki.issn1007-7820.2024.08.013 Abstract （186）   HTML （8）    PDF（pc） （607KB）（70）       Knowledge map    Save In order to further reduce the probability of multiple networks being attacked, different types of network security situation prediction models have received widespread attention and in-depth research from scholars both domestically and internationally. With the rapid development of situational awareness modeling technology, various novel technical solutions such as neural networks, time series, and support vector machines have been introduced into the prediction model of network security situations, deeply optimizing and improving the means and methods of situational prediction models, thereby further improving the accuracy of situational prediction models. This study reviews and sorts out the research history and development process of network security situation prediction technology, elaborates on the main principles and current development status of situation prediction models, analyzes the shortcomings and deficiencies of current technical solutions, and points out the future research directions of network security situation prediction model technology. Reference | Related Articles | Metrics

Select

Antenna Optimal Design Based on Flamingo Search Algorithm HUANG Zeming, SHAN Zhiyong Electronic Science and Technology    2024, 37 (9): 14-19.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.003 Abstract （183）   HTML （2）    PDF（pc） （2700KB）（58）       Knowledge map    Save In the field of antenna optimization design, traditional electromagnetic softwares adopt the method of sweeping with parameter, which causes the problems of large amount of calculation and low efficiency. In view of this problem, a joint optimized design based on flamingo search algorithm and HFSS (High Frequency Structural Simulator) is proposed in this study. This method is realized through the co-simulation of MATLAB and HFSS. MATLAB generates a VBS (Visual Basic Script) script for HFSS to call by writing the code of the design and simulation about antenna in MATLAB. HFSS returns the result of electromagnetic simulation to the function about fitness calculation. The flamingo search algorithm optimizes the particle position according to the fitness value until the optimal result is found and the optimal antenna size parameters are determined. In the co-simulation design, the optimal size parameters are found after 23 iterations, so that the fitness value of the optimized antenna in the two working frequency bands is optimized from -483.37 dB to -771.15 dB. The results of simulation show that the proposed algorithm has strong optimization ability and fast speed of convergence, which significantly improves the efficiency of the design of antenna optimization. Table and Figures | Reference | Related Articles | Metrics

Select

Research on Parameter Optimization of Cold Crucible Power Supply Based on Electromagnetic Simulation WANG Zexue, LI Yusong, LONG Haoqi, MING Yuzhou Electronic Science and Technology    2024, 37 (11): 70-77.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.010 Abstract （178）   HTML （3）    PDF（pc） （4634KB）（23）       Knowledge map    Save In view of the problems of high cost and complicated parameters, the magnetic field simulation model of φ100 mm cold crucible experimental device is established by COMSOL simulation software, and the simulation experiments of high-frequency power supply and coil parameters are carried out. The optimal combination of high-frequency power supply parameters is predicted by neural network and simulated annealing algorithm. The magnetic field distribution, eddy current loss and energy utilization of cold crucible are simulated under different current intensity, frequency, coil height, turns, diameter and spacing. Based on the above simulation results, the neural network can quickly predict the magnetic field and eddy current loss under various power supply parameters, and finally use the simulated annealing algorithm to predict the optimal combination of power supply parameters. The simulation results show that after parameter optimization, the magnetic field uniformity can be greatly improved, the melting effect can be improved, the energy utilization rate can be increased by 68%, and energy waste can be avoided. These results preliminarily verify the improvement effect of high-frequency power supply and coil parameter optimization on magnetic field distribution and energy use efficiency, which can provide an important basis for the optimization design of power supply parameters and coil parameters of the following cold crucible engineering prototype. Table and Figures | Reference | Related Articles | Metrics

Select

Emotion Recognition Method Based on EEG and Instantaneous Emotion Intensity Label GAN Kaiyu, YIN Zhong Electronic Science and Technology    2024, 37 (11): 78-84.   DOI: 10.16180/j.cnki.issn1007-7820.2024.11.011 Abstract （177）   HTML （6）    PDF（pc） （1779KB）（68）       Knowledge map    Save Revealing human brain activity through machine learning EEG(Electroencephalogram) has become an important scheme to explore the inner emotional state of humans. Because the change of emotion state is dynamic rather than constant, it is difficult to predict the change of emotion state in the field of emotion recognition. This study proposes a label generation framework for instantaneous emotion intensity. A set of supervised labels is generated by having subjects watch videos that stimulate and capture their instantaneous emotional intensity, and combine the supervised labels with EEG features to generate three sets of semi-supervised labels to correspond to the instantaneous emotional state changes of subjects. In this study, EEG features and various machine learning methods are used to analyze the applicability of four groups of labels to emotional state changes. The support vector machine model achieves 80.02%, 54.76% and 56.14% classification accuracy for two-class, three-class and four-class sentiment intensities on supervised label sets. The experimental results show that the supervised instantaneous emotion intensity labels are more universal for EEG data and emotional state changes across different subjects. Table and Figures | Reference | Related Articles | Metrics

Select

Privacy-Preserving Data Style Transfer Method for Artificial Intelligence of Things CHENG Jinke, LI Gaolei Electronic Science and Technology    2024, 37 (10): 1-5.   DOI: 10.16180/j.cnki.issn1007-7820.2024.10.001 Abstract （172）   HTML （21）    PDF（pc） （1628KB）（57）       Knowledge map    Save In the artificial intelligence of things, traditional privacy protection technologies mainly focus on the transmission, storage, and analysis stages of the data lifecycle, while ignoring the importance of protecting data privacy at the source. This study proposes a privacy-protecting data style transfer method for artificial intelligence of things. Based on cycle-consistent adversarial networks, a new loss function is added to obfuscate identity information, allowing real-style images and animation-style images to visually transform into each other. Animation-style data can be used to construct various virtual entities in the digital world (such as metaverses), and malicious users cannot reverse the original data based on the virtual entities or correctly identify the original data using the original deep learning model, thereby enhancing privacy protection for real entities in the physical world. Experimental results on a face dataset show that the transformed data reduces the accuracy of the ArcFace face recognition model by 30% without significantly reducing visual distortion. Table and Figures | Reference | Related Articles | Metrics

Select

Research on AGV Path Fusion of Object Detection and DWA Algorithm LI Jun, LIU Hu, YANG Haima, WANG Yuan, XU Wencheng, HUANG Hongxin Electronic Science and Technology    2024, 37 (7): 33-42.   DOI: 10.16180/j.cnki.issn1007-7820.2024.07.005 Abstract （171）   HTML （11）    PDF（pc） （4156KB）（97）       Knowledge map    Save In view of the path planning and navigation problem when the AGV forklift is in the situation of unknown environment information or dynamic change of environment, a method is proposed to obtain the target position by YOLOv5(You Only Look Once version 5) target detection algorithm. The global basic path is planned according to the target location, and the method of AGV path planning and navigation is integrated with DWA(Dynamic Window Approach) local dynamic path planning algorithm, so that the AGV forklift can quickly identify the target location and complete the path planning to reach the target location in the unknown environment or the environment with unknown local environment information. The experimental results show that compared with the previous method, the proposed method has good performance in terms of path length, time consumption and heading error of AGV forklift truck. The average path length is reduced by 12%, the average time consumption is reduced by about 5%, and the average error between the AGV heading and the target heading is within 5°. The proposed method can improve the working efficiency and flexibility of AGV forklift in unknown environment. Table and Figures | Reference | Related Articles | Metrics

Select

Persistent Clean-Label Backdoor Attack for Semi-Supervised Graph Node Classification YANG Xiao, LI Gaolei Electronic Science and Technology    2024, 37 (9): 57-63.   DOI: 10.16180/j.cnki.issn1007-7820.2024.09.009 Abstract （167）   HTML （3）    PDF（pc） （1479KB）（37）       Knowledge map    Save Semi-supervised graph learning aims to infer the class of unlabeled nodes or graphs by using various prior knowledge in a given graph. By improving the automation of data labeling, semi-supervised graph learning has high efficiency in node classification, but as a deep learning architecture, it also faces the threat of backdoor attacks, but no effective backdoor attack method has been developed for semi-supervised graph node classification tasks. This study propose a persistent clean-label backdoor attack method for semi-supervised graph node classification models, which generates poisoned samples by adaptively adding triggers and perturbations on unlabeled training data, and then trains to obtain poisoned semi-supervised graph node classification models without modifying the labels. The attacker can poison the model more stealthily with a poisoning rate no higher than 4%. To ensure the persistence of the backdoor in the model, a hyperparameter tuning strategy is also proposed to select the optimal value of the perturbation. Extensive experiments on several semi-supervised graph node classification models and open-source datasets show that the proposed approach achieves an attack success rate of up to 96.25% with little loss of classification accuracy of the model on normal samples. Table and Figures | Reference | Related Articles | Metrics

Select

Research Progress of Silicon-Based Optical Waveguide Beam Splitter CHEN Menglin, FENG Song, WANG Di, LIU Yong, HU Xiangjian, FENG Lulu Electronic Science and Technology    2025, 38 (1): 59-72.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.009 Abstract （167）   HTML （6）    PDF（pc） （2598KB）（66）       Knowledge map    Save As an important device in photonic integrated circuit, optical waveguide devices can be divided into active device with modulation function and passive device with static characteristic. Optical waveguide beam splitter is one of the important passive devices in optical communication networks. It is widely used in optical communication systems to realize efficient and fast information transmission. Silicon-based materials have become the main substrate materials for photonic integrated circuits due to their superior photoelectric properties and compatibility with semiconductor processes in integrated circuits. In this study, the current research status of silicon optical waveguide beam splitters is analyzed, the directional coupler, multimode interference coupler and subwavelength grating are summarized, and the dimensions and performance of different types of silicon optical waveguide beam splitters are discussed and compared. It provides a new idea for developing low loss, small size and wide band optical waveguide beam splitte. Table and Figures | Reference | Related Articles | Metrics

Select

Path Planning of Improved Artificial Potential Field Method Based on Deflection Angle Suppression JIN Tao, YU Lianzhi Electronic Science and Technology    2024, 37 (10): 15-22.   DOI: 10.16180/j.cnki.issn1007-7820.2024.10.003 Abstract （164）   HTML （7）    PDF（pc） （2462KB）（57）       Knowledge map    Save In view of the local minimum value, unreachable target and path oscillation in the practical application of traditional artificial potential field method, this study proposes an improved artificial potential field method based on deflection angle suppression. Based on the traditional artificial potential field method, this method adopts the improved repulsive potential field function to ensure that the target point is the lowest point of the whole situation. The deflection angle inhibitor is introduced in the path solution to suppress the excessive deflection angle during the driving process of the unmanned vehicle. After the unmanned vehicle falls into the local minimum point, a new virtual target point is added to the path solution at each step until the unmanned vehicle accumulates a certain deflection angle to get rid of the obstacle group. The simulation results show that the algorithm can reduce the volatility of the planning path without affecting the obstacle avoidance of the unmanned vehicle, and enable the unmanned vehicle to smoothly escape the complex obstacle group, smoothly reach the target point, and plan an effective, short and less oscillating path. Table and Figures | Reference | Related Articles | Metrics

Select

Numerical Simulation of Electrically Assisted Pressure Joining Based on Abaqus LI Yongfang, YANG Yali Electronic Science and Technology    2025, 38 (2): 17-22.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.003 Abstract （162）   HTML （9）    PDF（pc） （2896KB）（70）       Knowledge map    Save Based on the principle of conservation of energy and taking into account the influence of heat transfer, a universal theoretical model for heat transfer in EAPJ(Electrically Assisted Pressure Joining) of sheet metal lap joining is established. Taking Grade 1 titanium alloy rolling plate as the research object, according to the actual experimental setup process of EAPJ, simulation and analysis of the electric field, temperature field, and residual stress field during the EAPJ process of titanium alloy plate are completed using Abaqus finite element software. The experimental results show that when the current is applied, the temperature of the workpiece joint rises rapidly to the peak value, and then remains near the peak value until the power is stopped. After 20 s of natural cooling, the joint temperature tends to room temperature, and its residual stress is mainly concentrated near the joint, about 110 MPa. By comparing the experimental and simulation data of the thermal cycle curve at the center point of the EAPJ process joint, it is found that the numerical simulation results are in good agreement with the experimental results, which verifies the accuracy of the numerical simulation method, and provides a theoretical basis for the research of EAPJ mechanism and the actual production process. Table and Figures | Reference | Related Articles | Metrics

Select

Research on Large Twin-Rotor UAV Based on Fuzzy PID JIN Rui, JIN Hai, MO Songnan Electronic Science and Technology    2024, 37 (8): 54-59.   DOI: 10.16180/j.cnki.issn1007-7820.2024.08.008 Abstract （161）   HTML （8）    PDF（pc） （1606KB）（68）       Knowledge map    Save Compared with the quad-rotor UAV(Unmanned Aerial Vehicle), the large twin-rotor UAV has the advantages of lower price, longer endurance and higher efficiency, but it has serious coupling and poor anti-interference ability. To solve these problems, this study proposes a dual-rotor control algorithm using variable universe fuzzy PID (Proportional Integration Differentiation). In this study, the UAV system is divided into position control and attitude control. In attitude control, fuzzy PID is used to adjust the parameters of PID in real time to improve the stability of UAV. According to the motion mode of UAV, a new type of UAV structure is proposed, and the dynamic model of double-rotor is established, and the simulation model is built on Simulink for verification. The results show that the proposed algorithm reduces the overshoot by 50% when compared with the ordinary PID control algorithm, which indicates that the proposed algorithm has better stability on large twin-rotor UAV and can recover faster in case of interference. Table and Figures | Reference | Related Articles | Metrics

Select

Research on Action Recognition Method Based on Deep Learning XIN Tenghao, LI Feifei Electronic Science and Technology    2024, 37 (10): 64-70.   DOI: 10.16180/j.cnki.issn1007-7820.2024.10.009 Abstract （159）   HTML （14）    PDF（pc） （1324KB）（49）       Knowledge map    Save The key of current research on behavior recognition algorithms based on deep learning lies in enhancing the accuracy and stability of key point extraction, in order to achieve more accurate action recognition of targets. However, many current algorithms tend to just add attention mechanisms that appear to perform better in the feature extraction stage of the target, without considering the impact of different attention mechanisms on different models and tasks. Therefore, this study proposes an algorithmic model for pose estimation based on various attention mechanisms, which further highlights the importance of selecting an appropriate attention mechanism by comparing the impact of different attention mechanisms on the model. In addition, considering the stability of key point extraction, the initialization of the model is fine-tuned to select a more suitable initialization method that improves the performance by increasing the category of weights on network layer judgments. Compared with the performance of the benchmark network model, the model enhances all evaluation metrics on both multiscale and no-multiscale CrowdPose datasets, where the average accuracy improvement in both cases is more than 1%. Table and Figures | Reference | Related Articles | Metrics

Select

Face Image Super-Resolution Reconstruction Based on Conditional Priori Swin Transformer ZHENG Fangliang, WANG Yannian, LIAN Jihong, RUAN Pei Electronic Science and Technology    2025, 38 (2): 35-41.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.005 Abstract （157）   HTML （6）    PDF（pc） （2824KB）（85）       Knowledge map    Save In view of the problem that the existing image super resolution models based on Swin Transformer do not preprocess the face image, resulting in poor final super resolution results, this study proposes a face image super resolution reconstruction method based on conditional prior Swin Transformer. The method uses face Parsing Map and Swin Transformer model to preprocess the face image, uses conditional prior to optimize the face hyper-segmentation problem, and uses face parsing map to restrict the process so as to obtain more valuable prior information. In the stage of deep feature extraction, the channel space attention mechanism is integrated with Swin Transformer module to balance the speed and precision of feature group adjustment. Experimental results show that the proposed method achieves a PSNR(Peak Signal-to-Noise Ratio)of 32.21 dB on the test set. Compared with the existing model, this method has a certain improvement. It is proved that the improved model is more suitable for human face, and the generated result is clearer and more real, and more details of face image texture can be restored. Table and Figures | Reference | Related Articles | Metrics

Select

Output Characteristics of Magnetostrictive Displacement Sensor Considering Skin Effect XIAO Mo, WU Qinmu Electronic Science and Technology    2025, 38 (1): 1-5.   DOI: 10.16180/j.cnki.issn1007-7820.2025.01.001 Abstract （155）   HTML （16）    PDF（pc） （2078KB）（72）       Knowledge map    Save In view of the low detection accuracy of magnetostrictive displacement sensor, a calculation model of output voltage of magnetostrictive displacement sensor considering skin effect is proposed in this study. On the baisis of explaining the working principle of the magnetostrictive displacement sensor, the influence of skin depth is analyzed, and the Wiedemann effect and piezomagnetic effect are considered for theoretical analysis. The relationship between the excitation pulse current and the output voltage of the magnetostrictive displacement sensor is simulated by the skin depth generated by the excitation pulse signal acting on the waveguide wire at different frequencies. The optimal parameter value of excitation pulse current frequency is obtained using the experimental platform. Simulation experiments verify that the correctness of the output voltage model of the magnetostrictive displacement sensor considering the skin effect, indicating that the proposed model further improves the detection accuracy of the magnetostrictive displacement sensor and enriches the mathematical model. Table and Figures | Reference | Related Articles | Metrics

Select

Traffic Sign Detection Algorithm Incorporating Receptive Field Enhancement Module and Attention Mechanism YE Yuxin, JU Zhiyong, LAI Ying Electronic Science and Technology    2024, 37 (6): 8-16.   DOI: 10.16180/j.cnki.issn1007-7820.2024.06.002 Abstract （151）   HTML （6）    PDF（pc） （2956KB）（74）       Knowledge map    Save In view of the number of shortcomings of target detection algorithm in traffic sign detection, this study proposes a traffic sign detection algorithm that incorporating receptive field enhancement module and attention mechanism. The algorithm is improved on basis of YOLOv5(You Only Look Once version 5) algorithm, the RFB (Receptive Field Block) is used to replace the SPP(Spatial Pyramid Pooling) in the original backbone, the attention mechanism modules ECAM(Efficient Channel Attention Module) and CBA (Convolutional Block Attention Module) are embedded in the feature fusion network, and the Matrix NMS (Matrix Non-Maximum Suppression) is used to sift the candidate bounding-boxes. The experimental results show that there is no change in the number of model parameters when compared with the original network, meanwhile, mean average precision of the algorithm reaches 82.31%, which is 8.59% higher than the original network, and the detection speed reaches 51.89 frame·s-1. In addition, there is no false detection or missing detection in each test scenario, which proves that the generalization ability of the algorithm is also better than original algorithm, and the algorithm can perform real-time detection of traffic signs. Table and Figures | Reference | Related Articles | Metrics

Select

An Automatic Test Method for FPGA Interconnect Resource Based on An Improved EK Algorithm FU Mannan, CHEN Suting, XIE Weikun, LIN Xiaohui Electronic Science and Technology    2025, 38 (2): 10-16.   DOI: 10.16180/j.cnki.issn1007-7820.2025.02.002 Abstract （150）   HTML （30）    PDF（pc） （2526KB）（112）       Knowledge map    Save In the FPGA(Field Programmable Gate Array) IR(Interconnect Resource) testing, existing testing methods have problems such as multiple test vector configurations, high testing complexity,and low testing efficiency. In order to reduce the number of configurations and improve the efficiency of testing,an automatic test method for FPGA IR based on an improved EK(Edmonds-Karp) algorithm is proposed. This method achieves the goal of reducing the number of configurations by changing the search for the shortest path from the source point s to the endpoint t in the EK algorithm to the search for the longest path from s to t. A model based on the internal underlying IR structure of the FPGA is established,the improved EK algorithm is applied to the Kinex-7 series FPGA for automated routing path search,and the routing path is configured into the FPGA for simulation experiments. The experimental results show that the proposed method can detect the open circuit fault, short circuit fault and fixed fault in FPGA with less configuration times without reducing the fault coverage. Table and Figures | Reference | Related Articles | Metrics