基于动态线性化的非线性系统的自适应预测函数控制  

Adaptive Predictive Functional Control of a Class of Nonlinear Systems Based on Dynamic Linearization

作　　者：张彬[1] 张卫东[1] 

机构地区：[1]上海交通大学自动化系,上海200240

出　　处：《控制与决策》2005年第11期1274-1278,共5页Control and Decision

基　　金：国家自然科学基金项目(60274032);高校博士点基金项目(20030248040)

摘　　要：基于pseudo-partia l-derivative(PPD)的概念动态线性化非线性系统,利用集结方法处理未来预测时刻的PPD,实现了非线性系统的自适应预测函数控制.所给算法的预测模型只与当前时刻的测量数据有关,不依赖于对象的具体结构.该算法能够提供有界的输入输出,并能无偏差跟踪给定值.最后通过大滞后对象和强非线性pH中和滴定实验验证了该方法的有效性,并说明了其具有很强的鲁棒性和抗干扰能力.A concept of partial derivative called pseudo-partial-derivative （PPD） is used to dynamically linearize a nonlinear system, and an aggregation method is applied to deal with the future predictive PPD. Then, an adaptive predictive functional control algorithm is presented. The design is based directly on PPD derived online from the input-output data, and the given algorithm also provides the bounded input-output sequence and tracks setpoint without steady-state error. Simulations for the time-delay plant and the pH neutralization experiment of the chemical reaction process show that the proposed method is efficient for the system parameters perturbation and external disturbance.

关 键 词：预测函数控制 非线性 集结 动态线性化 基函数 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]