基于动态熵搜索的故障隔离策略研究

Research on Fault Isolation Strategy Based on Dynamic Entropy Search

为了进一步缩短平均故障隔离时间,在直接熵法和间接熵法的基础上,提出了动态熵搜索的故障隔离策略.该方法以系统最大动态熵作为测试选择的依据,并以动态熵为核心遍历搜索而建立故障隔离决策树.算例验证表明:与直接熵法和间接熵法相比较,该方法所需的测试少,且减少了平均故障隔离时间.

In order to shorten the mean fault isolation time,the fault isolation strategy based on dynamic entropy search is proposed in this paper on the basis of entropy-directed algorithm and entropy-indirected algorithm.The strategy takes the maximum dynamic-entropy of the system for account to select tests and establishes fault isolation decision tree using ergodic search on the centre of dynamic entropy.It is showed by algorithm example that the strategy in this paper needs less tests and mean fault isolation time than entropy-directed algorithm and entropy-indirected algorithm.