Observer design for Linear Hybrid Systems with unknown inputs and Petri-net discrete dynamics

摘要

This work investigates the observability analysis and the observer synthesis for Linear Hybrid Systems (LHS's) considering both known and unknown inputs and constrained discrete dynamics. In this framework, none assumption is made over the unknown inputs (that can represent faults or disturbances), moreover, the discrete dynamics are represented by a Petri net, whose behaviour constraints the possible switchings. For this, the concept of eventual observability is considered as the possibility of uniquely determining both the discrete and the continuous states after a finite number of switchings. In this way, the information provided by the continuous and the discrete outputs of the LHS is combined to determine the discrete state after a finite number of switchings. Next, based on the knowledge of the visited discrete states, a continuous observer estimates the continuous state. Under this approach, neither the observability of the linear systems nor the observability of the underlying discrete event system are required for estimating the complete state of the LHS. (c) 2020 Elsevier Ltd. All rights reserved.