Simulation of fault-tolerant control strategy of BP neural network automatic transmission based on MATLAB

Abstract

The application of wire control technology in the automobile is the development trend of the automobile industry. Wire control technology uses the cooperation between sensor, controller and actuator to complete the work of the traditional hydraulic mechanical structure on the automobile, and has the advantages of high-speed response, flexible installation, easy maintenance and so on that the traditional mechanical hydraulic structure does not have. The application of drive-by-wire technology in transmission has obvious advantages. The research of drive-by-wire automatic transmission for electric vehicles has obvious practical application significance. According to the needs of transmission development engineering, this paper studies the electronic control system of automatic drive-by-wire transmission for electric cars. In Simulink/Stateflow, the design of shift logic, shift rule, fault diagnosis and fault-tolerant control strategy based on BP neural network is completed and simulation tests are carried out. Improve the fault tolerance rate to improve the stability of the control system of automatic transmission by wire, so as to avoid the occurrence of traffic accidents.