Research On Hydrogen Safety Monitoring And Fault Diagnosis Of Fuel Cell Engine System

The fuel cell engine system is a nonlinear system with many input and output variables.In order to diagnose the faults accurately and ensure the normal work and safe operation of the fuel cell engine system,it is necessary to monitor and diagnose the faults in fuel cell engine system.In this paper,the fuel cell engine system is taken as the research object,the hydrogen safety monitoring system is designed,and the PFCM-ABC-SVM(Possiblistic Fuzzy C-Means clustering-Artificial Bee Colony-Support Vector Machine)algorithm is used to carry out the fault diagnosis research on the air supply system of the fuel cell engine.According to the structure of the fuel cell engine system and the requirement of the hydrogen safety monitoring system,the overall scheme design of the hydrogen safety monitoring system is completed.The key components of the hydrogen safety monitoring system are selected,and the hardware circuit design of the system is completed.The circuit schematic diagram and PCB diagram of the hydrogen safety monitoring system are designed by Altium Designer,and the prototype of hydrogen safety monitoring system is developed.The driver software of hydrogen safety monitoring system is developed by DAVE software,including STM module,CAN communication module and AD conversion module.The fault level of hydrogen safety monitoring system is divided,and the fault diagnosis strategy is developed by Matlab/Stateflow software.In TASKING software,the driver software configured by DAVE software and the fault diagnosis strategy of hydrogen safety monitoring system developed by Matlab/Stateflow software are integrated to realize the integration of main functions and the exchange of variables.The hydrogen safety monitoring system is tested for hydrogen leakage fault,over temperature and over pressure fault,and CAN communication fault to verify the function of the hydrogen safety monitoring system.The leakage of air supply manifold and other faults in the air supply system of fuel cell engine is simulated.And the variance of Gaussian noise is set to 1,0.5,0.2 and 0.1,respectively.Eight sets of diagnostic variables,such as fuel cell current,fuel cell voltage and air compressor outlet pressure,are obtained under normal and fault conditions.The initial sample set is established and the data is normalized.The PFCM-ABC-SVM algorithm is used for fault diagnosis,get the accuracy of training set and testing set,and compare the fault diagnosis results with other algorithms.Theresults show that the PFCM-ABC-SVM algorithm can effectively diagnose the leakage of air supply manifold and other faults in the air supply system of fuel cell engine.