Fault Tolerant control of nonlinear systems Iterative observer based approach

Abstract

This project deals with fault-tolerant control (FTC) for nonlinear systems using an observer- based estimation strategy. A complete framework is developed by combining Takagi-Sugeno fuzzy modeling with fast iterative k-step observers for real-time actuator fault estimation and compensation. Centralized and distributed observers are designed for both single-agent and multi-agent systems, with stability ensured using Lyapunov theory and Linear Matrix Inequal- ities. The method is applied to differential drive mobile robots, maintaining trajectory tracking despite actuator faults. The control architecture is implemented in ROS and tested in a re- alistic robotic simulation. Results confirm the method’s reliability and practical applicability. This work advances robust and adaptive control strategies for autonomous systems under fault conditions.