摘要
针对存在时延以及丢包的多包传输直流伺服电机网络控制系统(networked control system,NCS),提出一种利用滑动窗口策略多核LS-SVM丢包在线补偿的神经网络PID趋近律滑模控制器。将系统模型进行等价变换,建立无时延多包传输离散系统模型;利用滑动窗口多核LS-SVM对多包传输的数据丢包进行在线预测补偿,建立系统补偿模型。提出神经网络PID趋近律滑模控制器设计方法,通过神经网络非线性映射实现对PID趋近律参数的在线调整。利用Truetime对该方法进行仿真,结果表明,该策略可以提升丢包补偿的精度,滑模控制能够在较快响应速度的条件下减小系统抖振,对直流伺服电机网络控制系统实现了较好的跟踪控制。
The multiple-packet transmission DC servo motor network control system has time delay and packet loss.To solve this problem,we propose a neural network PID reaching law sliding mode controller with sliding window strategy and multiple kernel LS-SVM packet loss online compensation.The system model was transformed equivalently,and the discrete system model of multiple-packet transmission was established without delay.The sliding window multiple kernel LS-SVM was further used to predict and compensate the data packet loss of multiple-packet transmission on line,and the system compensation model was established.We proposed a sliding mode controller design method based on neural network PID reaching law.The online adjustment of PID reaching law parameters was realized by nonlinear mapping of neural network.The method was simulated by Truetime.The results show that the strategy can improve the accuracy of packet loss compensation.The proposed sliding mode control can reduce the system chattering at faster response speed,and achieve better tracking control for the network control system of DC servo motor.
作者
陈莹
唐友亮
张锦
於锋
Chen Ying;Tang Youliang;Zhang Jin;Yu Feng(School of Mechanical and Electrical Engineering,Suqian College,Suqian 223800,Jiangsu,China;College of Electrical Engineering,Nantong University,Nantong 226019,Jiangsu,China)
出处
《计算机应用与软件》
北大核心
2019年第11期70-77,共8页
Computer Applications and Software
基金
国家自然科学基金项目(51507087)
江苏省高等学校自然科学研究项目(17KJB470013)
宿迁市科技计划项目(Z2018220)
关键词
直流伺服电机
网络控制系统
最小二乘支持向量机
神经网络
滑模控制
DC servo motor
Network control system
Least square support vector machine
Neural network
Sliding mode control
