Reliability Assessment for Complex Systems by Fusing Multi-Stage Test Information

To solve the problem that the reliability assessment of complex systems was difficult due to the lack of reliability information in the development stage, a hybrid Bayesian assessment method based on KL divergence was proposed in this paper. The probability distribution of the failure rate at each test stage was calculated using reliability test data from the previous development stages, which were transferred as the prior distribution in the final test stage, and then Bayesian theory was used to calculate the posterior distribution of the system's failure rate. Considering the difference between the posterior distribution of the failure rate and the probability distribution of the failure rate in the current test stage, the mixed probability distribution of the failure rate was constructed by introducing the similarity factor based on the KL divergence method, and then the integrated posterior distribution of failure rate was calculated by normalizing the mixed probability distribution of each failure rate. Based on the relationship between failure rate and reliability and MTBF, the probability distributions of the system's failure rate, reliability, and MTBF were established, and the Bayesian confidence limits of the system's reliability measures were estimated.