Research On The Fault Diagnosis Technology Of Aerospace Propulsion Testing

Aerospace propulsion system testing, which is known for its expensiveness and hazardousness, is a vital key aspect in process of the research and development of aerospace engineering. It is an enormous systematic engineering with very complicated and comprehensive technology. With the development of the aerospace industry, aerospace engineering needs more sophisticated technology of propulsion system testing. It's crucial to improve the tasting system's ability to resolve the emergencies.Focusing on aerospace propulsion testing system, this paper is to deal with the foult diagnosis technology issues in respect of fault detecting and fault differentiation. The main purpose is to create a comprehensive and practical fault diagnosis platform that is applicable to aerospace propulsion system testing to realize efficient and reliable online fault diagnosis real-timely and discern systematic fault types,as well as fault setting. The platform can also be applied to provide policies of fault intervention and control the status of the testing system based on the results of detecting and diagnosis.The study of thesis is on basis of the method of fault detection of statistical theory. This method with high real-time performance is simple and reliable. The fault detection bases on that if the data have statistical characteristic in the normal station, but not the precise mathematical model of the detected system. The method of health degree is used in test stand system fault detection. It is on basis of the fuzzy set theoretical research how to measure quantitatively the health state of the test stand, and then predicts the fault detection. With the aid of abundant history test data and the expert experiences, the thesis uses the expert systems method to inference, analyze and judge, then get the formation of the fault type, cause and position to diagnosis the work attitude of the test system. The fuzzy comprehensive evaluation is studied in this paper which is based on data aggregation. According to different degree causality between every fault's cause and symptom, the method diagnoses the possible cause of system fault on the basis of comprehensively considering all the symptoms. It uses fuzzy transformation and maximum subordination principle. These methods will be the composition of core link in the fault detection of aerospace power system and diagnosis platform.Employed in research experiments of the aerospace propulsion system ultimately, the platform is to improve the reliability, safety and fault-tolerand ability of the testing system and to avoid or reduce economic loss of the testing fault in the process of research and development of aerospace models so that it can provide technology securities of the aerospace propulsion system research.