Observer-based adaptive neural event-triggered control for nonlinear systems with multiple constraints and multiple faults

For nonlinear constrained systems with sensor and actuator faults (SAFs), a neural network-based adaptive event-triggered controller is proposed. Nonlinearities and SAFs commonly occur in actual engineering plants, which can lead to system instability, and severely degrade control performance. The state observer with a fault compensation coefficient is introduced in this paper to estimate the unknown system states caused by sensor faults. The presented method satisfies full-state constraints (FSCs) directly and eliminates intermediate controllers feasibility conditions through the use of nonlinear mappings. In order to save communication traffic, a NNs event-triggered controller is derived. Through Lyapunov theorem, the stability of the closed-loop system (CLS) can be guaranteed and the Zeno phenomenon is avoided. Simulation results from a numerical example and a continuous stirred tank reactor (CSTR) system are presented to demonstrate the effectiveness of the approach.