摘要
为提高直驱力矩电机在变外载荷工况下的位置精度,使用基于狼群算法(WPA)优化支持向量回归(SVR)神经网络,建立直驱力矩电机位置误差预测模型。搭建试验台,将磁粉制动器与力矩电机用联轴器连接,利用磁粉制动器对直驱力矩电机施加变化外载荷。以可变外载荷数据和直驱力矩电机运行时电流数据作为输入,以直驱力矩电机位置误差作为输出。实验结果表明:所建立的WPA—SVR预测模型精度高于SVR预测模型,能够有效预测直驱力矩电机变负载下位置误差。
In order to improve the position precision of direct drive torque motor under variable external load condition,a position error prediction model of direct drive torque motor is established by using support vector regression(SVR)neural network optimized by wolf pack algorithm(WPA).An testbed is built,the magnetic powder brake is connected with coupling for torque motor,and the magnetic powder brake is used to apply variable external load to the direct drive torque motor.The variable external load data and the current data when the direct drive torque motor is running are taken as the input,and the position error of the direct drive torque motor is taken as the output.The experimental results show that the precision of WPA—SVR prediction model is higher than that of SVR prediction model,and can effectively predict the position error of direct drive torque motor under variable load.
作者
徐祐民
陈秀梅
彭宝营
王鹏家
XU Youmin;CHEN Xiumei;PENG Baoying;WANG Pengjia(School of Electromechanical Engineering,Beijing Information Science&Technology University,Beijing 100192,China)
出处
《传感器与微系统》
CSCD
北大核心
2024年第1期69-71,83,共4页
Transducer and Microsystem Technologies
基金
北京市科技计划资助项目(Z191100002019004)
北京市教委科技计划一般项目(KM202011232012)
国家自然科学基金资助项目(51575056)。
关键词
变负载
位置误差
直驱力矩电机
狼群算法
支持向量回归
variable load
position error
direct drive torque motor
wolf pack algorithm(WPA)
support vector regression(SVR)
