Model-Free Adaptive Iterative Learning Fault-Tolerant Control for Nonlinear Systems With Sensor Fault and Measurement Noise

To solve the challenge of fault-tolerant control (FTC) with both measurement noise and sensor fault, a model-free iterative learning adaptive fault-tolerant control strategy (IL-FTC) is proposed. An improved data-based featureless Kalman filter (IUKF) mechanism is devised for FTC. Then, an iterative estimation structure is designed using the dynamic linearization-based method to enhance the online estimation of some key parameters. Meanwhile, a new data-based fault detection (FD) algorithm is presented using an incremental broad learning system (IBLS) technique to detect sensor faults, avoiding the need for accurate model parameters for designing a fault detection observer. Based on this, controller reconfiguration is achieved by valid fault estimation. To improve the control performance of the system, a model-free adaptive controller (MFAC) that fully utilizes more historical information is developed. Finally, the superiority of the developed method is illustrated by a robotic arm simulation.