基于神经网络模型的扩展优化自校正预测控制  

Extended optimizing adaptive predictive control based on neural network model

作　　者：董光平[1] 方敏[1] 王经维[1] 

机构地区：[1]合肥工业大学电气与自动化工程学院,安徽合肥230009

出　　处：《合肥工业大学学报（自然科学版）》2002年第6期1163-1166,共4页Journal of Hefei University of Technology：Natural Science

摘　　要：利用前馈神经网络权初值优化的快速BP算法建立对象的非线性预测模型,采用分段线性化的技术建立动态线性模型,基于该线性模型进行滚动优化,同时用非线性预测模型对其进行补偿,实现对具有时延的非线性系统的预测控制,较好地解决了非线性系统存在时变、模型失配等情况下的控制问题。仿真实验表明由它构成的控制系统具有很好的动态响应和较强的鲁棒性。The feedforward neural network of fast BP algorithm with initial weight optimization is used to establish the nonlinear predictive model of the object, and the dynamic linear model is obtained by using the piecewise linearizing method. Roll optimization is carried out based on the linear model and the nonlinear model is adopted to make compensation at the same time, so that the predictive control of the nonlinear system with time delay is successfully realized,and the control problems in the nonlinear system under conditions of timevarying and modelmismatching are solved. The simulation experiments show that the control system displays good dynamic response and strong robustness.

关 键 词：神经网络模型 优化 自校正预测控制 动态线性模型 非线性系统 控制系统 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置] TP271.[自动化与计算机技术—控制科学与工程]