基于LSSVM的电磁直线执行器位移估算研究

Research on displacement estimation of electromagnetic linear actuator based on LSSVM

针对电磁直线执行器应用中由于位移传感器而产生的结构紧凑性下降、成本高以及可靠性降低等问题,提出了一种基于最小二乘支持向量机(LSSVM)的位移估算方法。应用多目标粒子群优化算法,以准确度和均方根误差(RMSE)为优化目标对正则化参数和核宽度参数进行多目标优化;在特定工况下训练模型。模型应用与仿真实验验证结果表明:在不同负载力工况下估算控制准确度为98%以上,RMSE为3×10^(-4)以下;以不同位移控制目标对电磁执行器进行控制,准确度为98%以上,RMSE为3×10^(-4)以下。

Aiming at the problems of reduced compact structure,high cost and reduced reliability due to existence of displacement sensors in application process of electromagnetic linear actuators,a displacement estimation method based on least squares support vector machine(LSSVM)is proposed.The multi-objective particle swarm optimization(PSO)algorithm is used to optimize the regularization parameters and kernel width parameters using precision and root mean square error(RMSE)as the optimization goals.The model is trained under specific working conditions.The model application and simulation experimental verification results show that the estimated control precision under different load force conditions is above 98%,the RMSE is below 3×10^(-4);using model to control the electromagnetic actuator with different displacement control targets,the precision is above 98%,and the RMSE is below 3×10^(-4).