一种挤压机能耗监测及预测系统研究

›› 2015, Vol. 28 ›› Issue (1): 85-.

一种挤压机能耗监测及预测系统研究

陀树青,梁鹏

(1.广东兴发铝业有限公司,广东佛山 528061;2.广东技术师范学院计算机科学学院,广东广州 510665)

出版日期:2015-01-15 发布日期:2015-01-22
作者简介:陀树青(1967—),男,工程师。研究方向:有色金属制造,仪表计量。E-mail:1937747978@qq.com。梁鹏(1981—),男,博士,讲师。研究方向:低碳制造,数据挖掘。
基金资助:
国家科技支撑计划课题基金资助项目(2012BAF12B10);[JP]广东省教育部产学研结合基金资助项目(2012B010500027)

Research on Energy Consumption Monitoring and Prediction System

TUO Shuqing,LIANG Peng

(1.Xingfa Aluminum Holdings Limited,Foshan 528061,China;
2.School of Computer Science,Guangdong Polytechnic Normal University,Guangzhou 510665,China)

Online:2015-01-15 Published:2015-01-22

摘要/Abstract

摘要：

针对铝型材生产过程中能耗较大,传统人工采集能耗数据频率低、速度慢等问题,文中提出一种铝型材挤压机生产实时能耗监测及异常检测系统。该系统分为实时能耗监测和能耗异常检测两部分,实时能耗监测使用串口通讯方式将生产现场电表、燃气表与交换机相连接,实现了对生产能耗数据的实时监测;能耗异常检测采用回归型支持向量机对历史生产数据进行训练学习,得到预测能耗模型,并根据历史数据计算出单位生产铝型材耗电量的置信区间,并采用误检率和漏检率用于对当前生产能耗数据作预测实验。实验结果表明,该系统可及时发现生产中的能源损失及生产参数不当等异常现象。

关键词: 能耗预测, 实时能耗监测, 回归型支持向量机

Abstract:

This paper proposes a real-time energy consumption monitoring and prediction system on aluminum casting furnace to deal with the great energy consumption in the production of aluminum and low efficiency of traditional manual data collection.A serial connection is used to connect field meter,gas meter and switches to achieve real-time monitoring of the production energy consumption data;on the other hand,historical production data is employed to train regression support vector machine,and confidence interval is computed by using historical production data.Finally,the trained SVM model of energy consumption is used for energy consumption prediction.The experimental result shows that our system can find abnormal incidents in the production such as energy loss and improper production parameters.

Key words: energy consumption prediction;real time energy consumption monitoring;regression support vector machine

中图分类号:

TP277

陀树青,梁鹏. 一种挤压机能耗监测及预测系统研究[J]. , 2015, 28(1): 85-.

TUO Shuqing,LIANG Peng. Research on Energy Consumption Monitoring and Prediction System[J]. , 2015, 28(1): 85-.