基于CPSO-RBF神经网络算法的预测控制方法研究  

Research on predictive control method based on cpso-rbf neural network algorithm

作　　者：杨秋菊[1] YANG Qiu-ju(Suzhou Vocational and Technical College,Department of computer information,Suzhou 234101,Anhui,China)

机构地区：[1]宿州职业技术学院计算机信息系,安徽宿州234101

出　　处：《贵阳学院学报（自然科学版）》2021年第3期13-17,共5页Journal of Guiyang University：Natural Sciences

基　　金：安徽省教育厅质量工程项目“网站开发与网页设计教学团队”(项目编号:2018jxtd051);宿州职业技术学院项目“photoshop图形图像处理智慧课堂试点”(项目编号:szg2019zlgc07)。

摘　　要：本研究在粒子群(PSO)算法中引入混沌思想构建出混沌粒子群算法(CPSO),提出基于CPSO改进的径向基函数(RBF)神经网络模型,以解决非线性控制问题。结果表明,模型的均方根误差RMSE与平均绝对误差AAE均明显优于其他算法,可高精度的辨识非线性系统;模型对节约药剂优势明显,超调量较小且可缩短调节时间,实现成本最低的效果,反映出本研究提出的模型的样本训练效果好,在整体上考虑了水流速与进口pH值,控制了系统的输入与扰动。In this study,chaos is introduced into particle swarm optimization(PSO)algorithm,chaotic particle swarm optimization(CPSO)algorithm is constructed,and an improved radial basis function(RBF)neural network model based on CPSO is proposed to solve the nonlinear control problem.The results show that the root mean square error RMSE and average absolute error AAE of the model are obviously better than other algorithms,and can identify nonlinear systems with high accuracy;The model has obvious advantages in saving reagent,small overshoot,shortening adjustment time and realizing the effect of minimum cost,which reflects that the sample training effect of the model proposed in this study is good,considering the water flow velocity and inlet pH value as a whole,and controlling the input and disturbance of the system.

关 键 词：预测控制 非线性系统 径向基函数神经网络 混沌粒子群算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]