未知初始条件下滞后过程的模型参数估计方法

Method to Estimation of Time-Delay System with Unknown Initial Conditions

提出了一种参数估计过程中消去未知初始条件的方法。在对象输入输出微分方程两边乘以时间变量,依据分部积分原理,经过多重积分,估计方程中不再含有未知初始条件。通过优化幅值均差函数辨识对象的滞后时间,然后采用最小二乘法辨识得到其他模型参数。仿真结果表明算法具有较高精度。

A new method by introducing time variable for eliminating the unknown initial conditions in the way to build estimating equation using integration tool is proposed in this paper. With the delay time identified by optimizing the average magnitude difference function, the other model parameters can be estimated by solving the estimating equation with Least-Squares method. It is sufficient for estimation with fine accuracy demonstrated by the simulation.