Adaptive fault-tolerant observer-based control for multi-input multi-output interconnected systems with bandwidth-limited communication

Abstract

The intensification of distributed manufacturing processes often increases the level of process interconnectivity, which subsequently leads to more intricate and complex process dynamics. Furthermore, wireless communication between processes and their controllers requires signal quantization. Motivated by these, this work presents an adaptive finite-time fault-tolerant observer-based quantized controller for a class of nonlinear state-delayed interconnected switched multi-input multi-output systems subject to arbitrary switching, quantized control and sensor signals, actuator and sensor faults, and nonlinear actuator behavior. The controller is tailored for interconnected distributed manufacturing processes with limited bandwidth communication channels for signal transmission. Sensor-fault-model parameters are estimated using a new adaptation law. Distributed observers are designed based on sensor-fault parameter estimates, the quantized faulty sensor data, and the quantized control signal. Nonlinear actuator behaviors including saturation, back-lash, and hysteresis are considered. The unknown nonlinear behavior of each subsystem is approximated using radial basis function neural networks. Using the Lyapunov-Razumikhin approach and an appropriate common Lyapunov function, the stability of the closed-loop system and the convergence of the tracking error to a desired neighborhood of the origin within a finite time are proved. The effectiveness of the proposed controller is validated through a simulation study conducted on a process system comprising integrated chemical reactors. The study highlights the controller’s ability to manage the interconnected dynamics of the system, ensuring stable and efficient operation under varying conditions.