基于模型预测的网络闭环控制研究

Research on Networked Closed-loop Control Based on Model Prediction

研究了远程传输时延不确定情况下闭环控制系统控制稳定性问题,给出了一种基于神经网络的模型预测控制法,仿真实验表明当控制网络存在时延时,随着控制系统时延值的大小和系统其他参数变化,不断的动态调整系统控制参数PID值可以使对工作环境未知的被控对象的控制变得稳定,改善远程闭环控制系统性能,以达到用户期望结果.

The paper first examines the stability of the closed-loop control system under the condition of uncertain remote transmission delay, and then describes a model predictive control method based on neural network. The simulation results show that when the control network has delay,with the change in the control system delay value and other parameters, the constant dynamic adjustment of the value of system control parameter PID can fnake stable the control of the controlled object which is unknown to the work environment, improve the performance of the remote closed-loop control system, thus meeting the user＇s expectation.