A novel linear quadratic observer and tracker for the linear sampled data regular system with a direct feedthrough term: A digital redesign approach

This paper develops a novel linear quadratic observer and tracker for the linear sampled data regular system with a direct transmission term from input to output. At the beginning, we use some existing techniques to decompose the singular system into an equivalent regular system which has a direct transmission term from input to output. Then, based on this equivalent regular system, a high-gain optimal linear quadratic analogue observer and tracker is proposed, so that it can effectively induce a high-quality performance on the trajectory-tracking design and the state estimation. Besides, the prediction-based digital redesign method is utilized to obtain a relatively low-gain and implementable digital tracker and observer from the theoretically well-designed high-gain analogue observer and tracker for the linear sampled data regular system with a direct transmission term from input to output. Several illustrated examples are given to demonstrate the performance and effectiveness of the proposed method.