A recursive identification method for continuous time-varying linear systems

Abstract

A method for online identification of continuous time-varying linear systems is proposed based on the block pulse functions. From the relations between entries of the block pulse operational matrices, simple regression equations corresponding to the original differential equation models with time-varying parameters can be obtained, and algorithms developed in discrete-time model identification can be applied directly to estimate the time-varying parameters of the continuous models without much modification. Compared with other methods for solving the same identification problem, no analog devices, no large computations in the data preparation stage, and no initial values are involved in this new method. Therefore, the estimation procedure is much simpler. Owing to the block pulse regression equations, the proposed recursive method is suitable for online identification of continuous time-varying linear systems from their sampled input and output data by means of digital computers.<>