DPSO算法在故障诊断测试集优化中的应用

摘要

In order to realize real-time fault diagnosis for complex systems, the multi-objective discrete particle swarm optimization (DPSO) algorithm is introduced to optimize the test set to implement effective fault monitoring and fault isolation. On the basis of establishing correlation model of the system, in accordance with various indexes in system real-time monitoring and diagnosis, such as the numbers of faults detected and isolated, the number of tests, and the test cost, etc. , the multi-objective fitness functions of the optimal fault detection test set and fault diagnosis test set are designed respectively. The examples of practical application indicate that the test set can be optimized rapidly for enhancing efficiency of real-time diagnosis, and the algorithm can be used as a comprehensive test evaluation method in improving design for test (DFT).%针对复杂系统的实时故障诊断问题,引入一种多目标离散粒子群优化(DPSO)算法对测试集进行优化,以实现有效的故障监测与隔离.在建立系统相关性模型的基础上,针对系统实时监视和实时诊断中故障检测数、故障隔离数、测试个数和测试成本等指标,分别设计了优化故障检测测试集和故障诊断测试集的多目标适应度函数.应用实例表明,该算法能快速地对测试集进行优化,提高实时诊断的效率,并可作为改进系统测试性设计中的一种测试综合评估方法.