The document discusses several approaches to designing interval observers for discrete-time linear switched systems affected by disturbances and measurement noises. The first approach details conditions for ensuring stability and nonnegativity of estimation errors using a common Lyapunov function. The second approach uses coordinate transformations to relax restrictive stability conditions. Another method designs interval observers under constrained switching signals using input-to-state stability and linear matrix inequalities with multiple Lyapunov functions. A final approach introduces new matrices to handle design difficulties with coordinate transformations and computes gains using linear matrix inequalities based on a common or multiple Lyapunov functions.
The document discusses several approaches to designing interval observers for discrete-time linear switched systems affected by disturbances and measurement noises. The first approach details conditions for ensuring stability and nonnegativity of estimation errors using a common Lyapunov function. The second approach uses coordinate transformations to relax restrictive stability conditions. Another method designs interval observers under constrained switching signals using input-to-state stability and linear matrix inequalities with multiple Lyapunov functions. A final approach introduces new matrices to handle design difficulties with coordinate transformations and computes gains using linear matrix inequalities based on a common or multiple Lyapunov functions.
The document discusses several approaches to designing interval observers for discrete-time linear switched systems affected by disturbances and measurement noises. The first approach details conditions for ensuring stability and nonnegativity of estimation errors using a common Lyapunov function. The second approach uses coordinate transformations to relax restrictive stability conditions. Another method designs interval observers under constrained switching signals using input-to-state stability and linear matrix inequalities with multiple Lyapunov functions. A final approach introduces new matrices to handle design difficulties with coordinate transformations and computes gains using linear matrix inequalities based on a common or multiple Lyapunov functions.
Interval observer for discrete-time linear switched systems
In this part, we detail different steps to design an interval observer for discrete time switched systems affected by disturbances and measurement noises which are assumed unknown but, bounded with known bounds. Two approaches are presented. The first one, which is restrictive, details the conditions ensuring the nonnegativity (cooperativity) and the stability of estimations errors using a common Lyapunov function and the nonnegativity (cooperativity) in the original coordinates. The second approach is based on transformation of coordinates allows to relax restrictive conditions of stability and nonnegativity of estimations errors.
II. Design of an interval observer for discrete time linear switched
systems under constrained switched signal We explain different steps of design of an interval observer for discrete time switched systems affected by disturbances and measurement noises. The design methodology is based on the Input-to-State Stability (ISS). The gains are computed by solving Linear Matrix Inequalities (LMI) formulated based on multiple quadratic Lyapunov functions under average dwell time switching signals. To ensure the nonnegativity of the estimation errors a change of coordinates is made.
III. A new interval observer design for discrete-time linear switched
systems (T-N-L method) The design of interval observers requires an extra constraint of nonnegativity compared with Luenberguer observers. To overcome the design difficulty, several methods based on coordinate transformation are evoked in literature. Such methods have shown its limitation. It is not often possible to synthesis simultaneous the coordinate transformation matrix and the observer gain matrix to fulfill both the positivity property and the stability of the estimation errors. A new interval observer structure is introduced based on the introduction of new matrices to handle the design difficulty of coordinate transformation. The interval observer gains are computed by solving Linear Matrix Inequalities (LMI) based on:
1. A common Lyapunov function.
2. A multiple quadratic Lyapunov functions under average dwell time switching signals.
IV. Comparative studies
To discuss the effectiveness of the presented methods A numerical example is given