Alriksson, Peter (Lund Univ.), Rantzer, Anders (Lund Univ.)

Observer Synthesis for Switched Discrete-Time Linear Systems Using Relaxed Dynamic Programming

Scheduled for presentation during the Regular Session "Switched Systems I" (MoA05), Monday, July 24, 2006, 10:50−11:15, Room F

17th International Symposium on Mathematical Theory of Networks and Systems, July 24-28, 2006, Kyoto, Japan

This information is tentative and subject to change. Compiled on April 25, 2024

Abstract

In this paper, state estimation for Switched Discrete-Time Linear Systems is performed using relaxed dynamic programming. Taking the Bayesian point of view, the estimation problem is transformed into an infinite dimensional optimization problem. The optimization problem is then solved using relaxed dynamic programming. The estimate of both the mode and the continuous state can then be computed from the value-function. From an unknown initial state the estimation error goes to zero as more measurements are collected.