基于最大后验概率候选集更换法的多故障诊断策略研究

Multiple Fault Diagnosis Strategy Based on Maximum Posterior Probability Candidate Fault Set Replace Method

给出了一种基于最大后验概率候选集更换法的多故障诊断策略，并深入阐述了基于最大后验概率候选集更换法的形式化描述，分析了对单故障诊断和多故障诊断的计算模型，提出了基本的故障诊断算法——改进二进制粒子群算法，并对抽象实例进行验证，结果表明，采用改进BPSO算法能有效地求解基于最大后验概率候选集更换法的多故障诊断问题。

In this paper, a multiple fault diagnosis strategy based on most posterior probability candidate fault set replace method is put forward. Firstly the formalization of most posterior probability candidate fault set replace method is studied;then the two computational models are analyzed：single fault diagnosis and multiple fault diagnosis, and a multiple fault diagnosis algorithm of Binary Particle Swarm Optimization is presented. At last, according to an illustration,it is proved that the improved BPSO does to be practical, and effective for multiple fault diagnosis based on maximum probability candidate fault set replace method.