基于双容液位系统的神经网络反步滑模与PCH协同控制  

Synergistic Optimal Control of Neural Network Backstepping Sliding Mode and PCH for Two Tank Liquid Level System

作　　者：张克 于海生[1,2] 孟祥祥 ZHANG Ke;YU Haisheng;MENG Xiangxiang(College of Automation,Qingdao University,Qingdao 266071,China;Shandong Province Key Laboratory of Industrial Control Technology,Qingdao 266071,China)

机构地区：[1]青岛大学自动化学院,青岛266071 [2]山东省工业控制技术重点实验室,青岛266071

出　　处：《自动化与仪表》2023年第8期104-109,共6页Automation & Instrumentation

基　　金：国家自然科学基金项目(62273189);山东省自然科学基金项目(ZR2021MF005)。

摘　　要：针对工业生产过程的实际需求,该文研究了具有参数不确定性和外部未知干扰的双容液位系统的控制问题。首先,设计了一种反步滑模控制策略(BSMC),以提高控制系统的速度;其次,利用神经网络(NN)逼近系统的非线性函数,以求解系统参数的不确定性和未建模的动态扰动;第三,设计了一种端口控制的哈密顿(PCH)控制策略,以保证控制系统的稳态控制精度;最后,提出了一种协同优化控制方法,它结合了这两种控制方法,充分发挥了神经网络反步滑模(NNBSMC)算法的快速性和PCH算法的高稳态精度。仿真结果表明,利用神经网络反步滑模控制和端口控制哈密顿协同优化控制(NNBSMC+PCH)可以使系统兼顾两种控制器的优点。Considering the actual demand of industrial production process,this paper studies the control of dual liquid level system with parameter uncertainty and external unknown interference.First,a backward sliding mode control strategy(BSMC)is designed to improve the speed of the control system.Secondly,the neural network(NN)is used to approximate the nonlinear function of the system to solve the uncertainty of the system parameters and the unmodeled dynamic perturbations.Thirdly,we designed a portcontrolled Hamiltonian(PCH)control strategy to ensure the steadystate control accuracy of the control system.Finally,a cooperative optimization control method is proposed,which combines these two control methods and gives full play to the speed of neural network backward sliding mode(NNBSMC)algorithm and the high steady state accuracy of PCH algorithm.The simulation results show that the backward sliding mode and port control(NNBSMC+PCH)can enable the system to consider the advantages of two controllers.

关 键 词：双容液位系统 反步滑模控制 端口受控哈密顿控制 协同优化控制 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]