改进的粒子群算法在磨煤机PID神经网络控制中的应用  

Research on the Improved Particle Swarm Optimization Algorithm in the Coal Mill PID Neural Network Control

作　　者：穆海芳[1] 韩君[1] 李明[1] MU Hai-fang;HAN Jun;LI Ming(Suzhou University, Suzhou 234000, China)

机构地区：[1]宿州学院机械与电子工程学院

出　　处：《金陵科技学院学报》2019年第2期16-20,共5页Journal of Jinling Institute of Technology

基　　金：宿州学院校级重点科研项目(2016yzd09)

摘　　要：磨煤机制粉系统的输入量和输出量之间相互耦合,而且具有非线性、时滞性大等特点,因此使用常规的控制方法难以达到良好的效果。设计了磨煤机制粉系统的多变量PID神经元网络控制系统,提出了改进的粒子群算法优化多变量PID神经网络参数,然后采用误差反传算法调整网络权值,避免了网络陷入局部最优解。磨煤机制粉系统的仿真实验表明,该方法解决了系统的耦合、时滞性问题,同时减小了系统的超调量,避免了系统的震荡,具有良好的稳态性能和动态性能。The input and output of the pulverizing system are coupled with each other, which have the characteristics of nonlinear, large time delay.So using conventional control method is difficult to achieve good effect.This paper designed the coal mill pulverizing system of multi-variable PID neural network control system, and proposed an improved particle swarm algorithm to optimize the multi-variable PID neural network parameters.Then, back propagation algorithm is used to adjust the network weights, and to avoid the network into a local optimal solution.The simulation experiment of coal pulverizing system shows that the method can solve the coupling and time delay problems of the system.And at the same time it reduces the overshoot of the system and avoid system shock, which leads to good steady-state performance and dynamic performance.

关 键 词：磨煤机 PID 神经网络 粒子群 

分 类 号：TD453[矿业工程—矿山机电] TP183[自动化与计算机技术—控制理论与控制工程]