CNN-Based Intelligent Fault-Tolerant Control Design for Turbofan Engines With Actuator Faults

Abstract

In this paper, the problem of fault-tolerant control is investigated for turbofan engines with actuator faults. The controller involvement has repressed the effects of actuator faults on the controlled outputs of turbofan engines, making fault-tolerant control difficult. To solve this problem, the internal gas-path data of turbofan engines is introduced to provide conducive fault information. Besides, the useful property of the convolution neural network (CNN) is explored and utilized in fault-tolerant control. Based on the analysis of actuator faults, by using the Lyapunov stability and $L_{2}$ -gain like theorems, a novel CNN-based intelligent fault-tolerant control system for turbofan engines is proposed, including a CNN-based fault diagnosis module and a nonlinear fault-tolerant control with corresponding reconfiguration unit. The CNN-based intelligent fault-tolerant control system has the advantages of reducing the accuracy requirements of the mathematical description of turbofan engines. Furthermore, the proposed system can diagnose actuator faults and reduce the adverse effects of actuator faults on turbofan engines. Finally, simulation results are presented to demonstrate the efficiency of the designed method.