研究薄膜厚度系统控制器优化问题,由于神经网络初始权值难以确定,使PID神经网络对控制器参数的自适应、自学习能力变差,最终导致控制效果不理想.为了解决这一问题,提出一种混合的粒子群算法,用来优化神经网络初始权值,进而实现控制器的优化,并应用于薄膜厚度控制系统.仿真结果表明:与PID神经网络控制器相比,优化后的控制器更好的实现了多变量控制系统的解耦控制,提高了控制器参数的自适应自学习能力,控制效果明显,并且系统的鲁棒性较好.%The paper proposed a hybrid particle swarm optimization (pso) algorithm,which was used to optimize neural network initial weights and achieve the optimization of controller.The algorithm was applied to film thickness control system.The simulation results show that compared with PID neural network controller,the optimized controller can realize the decoupling control of multivariable control system better and improve the adaptive self learning ability of the controller parameters.The control effect is obvious,and the system robustness is better.
展开▼
