The Research On Fault Tolerant Control For Aero-Engine Sensors Based On MPC

Sensor faults may affect the safety of aero-engine control system,and even lead to the failure of the control system.In order to improve the safety of the system and reduce the impact of faults on the system,it is necessary to implement Fault Tolerant Control(FTC).The requirements for the fault-tolerant capability of the system are set out in the Article 33.28 of CCAR33.Model Predictive Control(MPC)is an advanced control algorithm.It is of great significance to incorporate FTC for sensor fault into MPC design for MPC controller airworthiness compliance verification.For sensor faults,FTC based on MPC is carried out,which takes turbofan engine as the research object.The specific research contents are as follows:Firstly,in order to avoid the max-min switching in traditional limit management and reduce the structure complexity of the system,a steady-state MPC controller is designed using the CMPASS engine model developed by NASA,which lays a theoretical and simulation foundation for the later FTC based on MPC.Secondly,using steady-state MPC controller,the influence of single sensor fault on the system is analyzed,and the necessity of FTC for fan speed N_f sensor is expounded.The simulation results show that both hard and soft faults of the N_f sensor can affect the system,and the greater the faults,the greater the impact.Thirdly,in order to eliminate or reduce the impact,the predecessor's FTCl of MPC method based on predictive model switching is improved for the hard fault of N_f sensor.First,a predictive model library including a normal model and some models with different sensor faults is established;Then,the monitoring decision unit selects the predictive model that best matches the sensor measurement value,and sends a switching instruction to the switching unit,thereby achieving online fault tolerance.If unknown faults occur in the system,the interpolation unit reconstructs the new predictive model that is incorporated into the predictive model library.The simulation results show that this method can not only tolerate known faults,but also have good fault tolerance effect for unknown faults.Fourthly,for the soft fault of N_f sensor,the FTC method of MPC based on trajectory transformation is proposed.First,Fault Detection and Diagnosis(FDD)is used to diagnose the sensor fault information;Then,according to the fault information,the system adjusts the reference trajectory online;Finally,the controller adjusts the control amount so that the output tracks the new reference trajectory,so as to realize the active FTC for the N_f sensor fault.The simulation results show that the method has good fault tolerance effect on both soft and hard faults,and it has better fault tolerance effect for soft faults than hard faults.We can also see that the Fault-tolerant Rate(FR)is inversely proportional to the fault size.