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.
出处
《控制与决策》
EI
CSCD
北大核心
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
关键词
随机连续系统
参数估计
最小二乘法
系统辨识
stochastic continous systems, parameter estimation, least squares method, numerical solution of ordinary differential equation, Wiener process