基于WOA-SVM模型的电磁继电器贮存寿命预测研究  

Research on Storage Life Prediction of Electromagnetic Relay Based on WOA-SVM Model

作　　者：乔青云 王召斌[1] 刘百鑫 陈康宁 QIAO Qingyun;WANG Zhaobin;LIU Baixin;CHEN Kangning(School of Automation,Jiangsu University of Science and Technology)

机构地区：[1]江苏科技大学自动化学院,江苏镇江212003

出　　处：《电器与能效管理技术》2023年第2期1-5,10,共6页Electrical & Energy Management Technology

基　　金：国家自然科学基金项目(51507074);江苏省研究生科研与实践创新计划资助项目(KYCX22_3809)。

摘　　要：可靠预测弹用电磁继电器的贮存寿命,对国防系统的运行有着重要现实意义。为了提高弹用电磁继电器贮存寿命的预测精度,提出一种基于鲸鱼优化算法支持向量机(WOA-SVM)的预测模型。在传统的支持向量机(SVM)模型的基础上,引入鲸鱼优化算法(WOA),对惩罚因子c和核函数参数g两个参数进行寻优,构建WOA-SVM预测模型,同时与基于飞蛾火焰优化算法支持向量机(MFO-SVM)模型和单一SVM模型进行对比研究。仿真结果表明,WOA-SVM预测模型对于弹用电磁继电器的贮存寿命预测结果更为准确,精度优于对比模型。Accurate prediction of the storage life of electromagnetic relays for missiles has important practical significance for the operation of national defense systems.In order to improve the prediction accuracy of the storage life of electromagnetic relay used in missiles,a prediction model based on the whale optimization algorithm support vector machine(WOA-SVM)is proposed.On the basis of the traditional support vector machine(SVM)model,the whale optimization algorithm(WOA)is introduced to optimize the penalty factor c and kernel function parameter g,and the required WOA-SVM prediction model is built.At the same time,the moth-flame optimization algorithm support vector machine(MFO-SVM)model and the single SVM model are used for comparative study.The simulation results show that WOA-SVM prediction model,for the storage life prediction of electromagnetic relay for missiles,is superior to other selected comparison models.

关 键 词：贮存寿命预测 鲸鱼优化算法 支持向量机 接触电阻 

分 类 号：TM581.3[电气工程—电器]