Research On Fault Tolerant Control For Brake Execution System Of Electromechanical Brake Vehicle

With the development of the X-by-wire technology, vehicle brake-by-wire systememerges. Electro-mechanical brake system (EMB) is a new brake system. Compared withthe traditional brake system, the mechanical connection between brake pedal and actuator iscancelled and the electric signal is used as a transfer command medium to drive motor toachieve the brake. The EMB system which represents the future development direction ofbraking system is more efficient and environmental. Because the braking system is relatedto the safety of vehicle directly, the new actuators must have greater reliability thanconventional ones so that they can be extended. The structure and control method of theEMB actuator which is the actuator of the EMB system are different from traditionalbraking systems, so the failure form has changed. In this paper, fault tolerant control forEMB actuator system has been researched.Based on the project which named “Research on vehicle intelligent brake system”(No.111080) founded by HuoYingdong Education Foundation, and “Research on vehicledynamic control based on braking by wire”(No.50805063) founded by National NaturalScience Foundation of China. The objective of this paper is to carry on a failure research forEMB actuator system used in the laboratory which is controled by tricyclic control strategy.A failure diagnosis and failure tolerant control strategy by neural network is raised. Theinfluence that the actuator failure has on vehicle stability is analyzed. A control strategy ofanti-deviation in the condition that the brake is in error is studied. And verify the validity ofthe fault-tolerant algorithm with EMB vehicle dynamics simulation software. The main contents of this dissertation include:(1)Based on the full understanding of the composition and working principle of theEMB system, this paper focuses on the research of the EMB actuator system, analyzes theweakness of the EMB actuator system, and summarises problems in the test. The influencethat the actuator failure has on vehicle stability is analyzed. The failure judgment methodand working status judgment basis of the actuator is proposed. the braking stability controlmethod in the condition that the brake is in error is raised.(2) The BP neural network is studied. The working principle and deficiency areanalyzed and then the BP neural network is improved. The sensor failure diagnosis andfailure tolerant control strategy of the EMB actuator system based on neural network isproposed. The EMB actuator observer is established by BP neural network.According to the comparison with the sensor signal, judging whether the sensor is inerror. The BP neural network is adopted to establish a fault signal reconstruction model.When the sensor is in error, reconstruction sensor will output the signal into the controller toensure the stability of the controller.(3) The vehicle anti-deviation controller in the EMB actuator failure mode is designed.The yaw rate is used as control variables, and then the yaw rate PID control is adopted tocalculate the compensation moment. A compensation moment distribution rule isestablished. The purpose to reduce braking deviation is achieved by adjusting the brakingforce of the wheels which have no fault.(4) Based on the existing EMB car model, the EMB actuator model and actuator faultdiagnosis and tolerant control model are build and integrated. Simulaton analysis for theEMB actuator sensors fault and verify the effectiveness of the EMB actuator sensorsdiagnosis and tolerant algorithm. Seting one EMB actuator in error, simulation EMBautomotive brake stability control.