Electronic Science and Technology ›› 2024, Vol. 37 ›› Issue (2): 23-29.doi: 10.16180/j.cnki.issn1007-7820.2024.02.004

Previous Articles     Next Articles

Estimation and Performance Analysis of Unscented Kalman Filter with Randomly Missing Measurements

BAI Rui1,REN Zhu2   

  1. 1. School of Computer Science and Technology,Zhejiang Sci-Tech University,Hangzhou 310018,China
    2. School of Information Science and Engineering,Zhejiang Sci-Tech University,Hangzhou 310018,China
  • Received:2022-10-25 Online:2024-02-15 Published:2024-01-18
  • Supported by:
    National Natural Science Foundation of China(61403347)


In engineering applications, wireless network control systems are mostly nonlinear systems. Due to long-distance transmission and unreliable communication networks, the measured values of system sensors may be lost in the transmission process, which influences accuracy estimation and system performance. In this study, the problem of unscented Kalman filtering for a class of nonlinear discrete stochastic systems affected by correlated noise and sensor measurement loss is studied. By introducing a random variable that obeys Bernoulli distribution and has known conditional probability to describe the random sensor measurement loss, an algorithm is proposed to compensate the data. The results are verified by standard numerical software. The results show that the filter compensated by the algorithm can estimate the system well, greatly reduce the impact of sensor measurement loss on the filter performance, and increase the accuracy of estimation.

Key words: wireless network control system, nonlinear discrete stochastic systems, state estimation, unscented Kalman filter, packet losses, system covariance, estimation error, system noise

CLC Number: 

  • TP393