Abstract: During the development process, the sample size of aerospace valves is small, and the sample rarely fails or even has no failure. In order to solve this reliability evaluation problem, it is necessary to make full use of various existing test data. The tests used for quantitative reliability analysis of valves mainly include temperature test, vibration test and action test. On the basis of preprocessing the temperature and vibration test data, the test data is fused into a unified prior distribution. Then taking the action test data as field data, the reliability evaluation of valves is completed according to the Bayes theory. The research shows that, compared with single-type test evaluation results and large sample evaluation results, the multi-source information fusion method is accurate, stable and credible.

Key words: aerospacevalve, reliabilityevaluation, Bayestheory, multi-sourceinformationfusion