PSONN在可编程控制系统PID参数优化中的应用

摘要

The modern industry control systems are of the characteristics of nonlinear and time-varying, and the it is difficult using traditional PID controller parameter optimization method to establish the precise mathematical model for the PLC control system, and the control effects are not ideal. The using the nonlinear and self-learning ability of RBF neural network, the paper proposed a PID parameter optimization algorithm based on particle swarm neural net-work , and applied it in PLC control system. The test results show that particle swarm neural network can improve the PID control parameter optimization speed, and obtain the global optimal PID parameters, thus improve the reliability of PLC control system and robustness, and has certain theoretical and practical value.%研究可编程控制系统优化问题,可编程控制系统具有非线性、时变性等特点,传统PID控制器优化方法难以建立精确的数学模型,使得系统参数设定困难,导致可编程控制系统的控制效果不理想.为了解决传统的PID算法所带来的问题,利用RBF神经网络非线性、自学习能力,提出一种基于粒子群神经网络的PID参数优化算法.将粒子群和神经网络相结合,形成了一种智能控制算法,并将应用于可编程控制系统.测试结果表明,粒子群神经网络提高了PID控制参数优化速度,提高了可编程控制系统可了靠性和鲁棒性,具有一定的理论和实用价值.