随机连续系统的连续时间最小二乘辨识的数值实现  

Numerical Realization of the Continuous-time Least-squaresIdentification for Stochastic Continuous Systems

作　　者：赵明旺[1] 

机构地区：[1]武汉冶金科技大学自动化系

出　　处：《控制与决策》1996年第6期654-658,共5页Control and Decision

基　　金：冶金工业部理论研究基金

摘　　要：讨论了随机连续系统的连续时间最小二乘（ＣＴＬＳ）辨识的数值实现及仿真。首先回顾了随机连续系统的ＣＴＬＳ辨识法和理论分析结果，然后基于数值积分技术和求解常微分方程的数值解的欧拉法和龙格－库塔法，给出了ＣＴＬＳ法的两种数值实现方法。仿真结果显示出此方法的有效性。The numerical realization and simulation of the continuous-time least-squares (CTLS) identification for stochastic continuous systems are studied. First the CTLS identification method for stochastic continuous systems and its convergence results are reviewed, then two numerical realization methods for the CTLS identification based on the numerical integral technique, and Euler method and Runge-Kutta method for ordinary differential equation are given. The simulation results show the effectiveness of these methods.

关 键 词：随机连续系统 参数估计 最小二乘法 系统辨识 

分 类 号：O231[理学—运筹学与控制论]