中文摘要：

大型复杂系统的故障隔离是系统维修的重要环节,现有的故障隔离算法存在平均故障隔离时间（MFIT）长,不能快速自动生成决策树等缺点。利用AO＊算法自动构造决策树,将离散粒子群算法应用于AO＊算法的每一个节点的测试选择,降低计算复杂度,使故障隔离策略智能化。实例表明：该算法将故障隔离时间缩短20%,并降低测试代价,提高故障隔离效率,为大型复杂系统故障隔离提供一种高效算法。

英文摘要：

Fault isolation （F1） of Large-scale complex system is an important part of system maintenance. Long mean fault isolation time （MFIT） and being incapable of quickly automatic building of FI decision-making tree are the disadvantages of existing methods. The AO＊ algorithm builds the decision-making tree automatically; discrete binary particle swarm optimization algorithm applied to the test node selection reduces the complexity and intelligent FI strategy. The example validates that this algorithm has shortened 20% MFIT, decreased the test cost and improved the efficiency of FI. It offers an efficiency algorithm to the fault isolation of large-scale complex system.