摘要
针对机械式负载模拟器体积大、成本高及电液式负载模拟器污染大、维修困难等问题,设计了一款电动式直线电机加载测试台。通过Simulink仿真对比了二自由度PID和改进型BP-PID两种控制系统的控制效果。结果表明,改进型BP-PID控制系统不依赖于参数整定,而是通过神经网络多次训练样本自整定出所需参数,稳定性更高、跟踪性更好,完全满足加载测试系统指标要求。
An electric linear motor loading test bench was designed to solve the problems existing in mechanical load simulator, such as big volume, high costs and pollutions, maintenance difficulties in electro-hydraulic load simulator. Then, the control effectivenesses of improved BP-PID and 2-DOF PID control systems were compared through Simulink simulation . Results show that the improved BP-PID control system does not depend on the parameter setting, the required parameters are self-set-ting through the neural network repeatedly training samples. Therefore, the BP-PID control system has greater stability and better tracking property, and may fully meet the requirements of the loading test system metrics.
出处
《中国机械工程》
EI
CAS
CSCD
北大核心
2017年第10期1221-1226,共6页
China Mechanical Engineering
基金
湖北省自然科学基金资助项目(2014CFA528
2016CFB653)
湖北省教育厅资助重点项目(D20151406)
