基于遗传算法的故障样本优化选取方法

为降低测试性验证试验费用,提出基于遗传算法的故障样本优化选取方法。方法通过故障-测试关联分析和故障-故障等价分析,确定初始故障样本集中各元素对应的等价集,并对初始故障样本集进行扩展,在此基础上,建立了故障样本选取优化求解模型。在不降低样本注入数量和测试特性的条件下,以试验费用最小为优化目标,给出了基于改进遗传算法的样本优化选取方法。算例应用结果表明,该方法设计的故障样本选取方法能有效降低测试性验证试验费用。

Method of failure sample optimization selection based on genetic algorithm

Aiming to the cost of testability verification experiment, a method of failure sample optimization selection based on the genetic algorithm is proposed. Through the analyses of fault test correlation and fault-fault equivalent, the alternative failure sample concentration equivalent set of each element is determined and the extension alternative failure sample set is established. On this basis the solving model of failure sample optimization selection is set up. Without reducing the sample injection quantity and the characteristics of the test conditions, the coding model of sample optimization selection is built by using the generalized chromosome. A method of failure sample selection and sequence injection is put forward based on the genetic algorithm, which takes the minimum cost as the optimization goal. Finally, an example results show that this method can effectively reduce the test cost.