基于故障扩散强度的故障样本选取方法

传统的测试性验证试验样本选取方法没有考虑测试对象的各故障模式之间的复杂性与不确定性关系。针对这一问题，建立了一种基于模糊概率Petri网的复杂系统故障扩散模型，该模型完备地描述了测试对象故障模式之间的传播扩散关系，并引入故障扩散强度概念，按分步扩散的原则对故障扩散过程进行描述，给出了每个故障的扩散强度量值。提出了基于故障率和故障扩散强度的故障样本选择方法，案例验证结果表明，基于该方法选择的故障样本集更能覆盖测试对象的故障模式集。

A Method of Failure Sample Selection Based on Failure Pervasion Intensity

The complex and uncertain relationship among ianures is ignored, in failure sample selection based on traditional testability demonstration experiment method. A lailure pervasion model was founded-fuzzy probability Petri net (FPPN) model which depicts the propagation and pervasion relation among failures, then iailure pervasion intensity was introduced, the process of lailure pervasion was depicted based on k-step fault pervasion principle and the pervasion intensity was expressed by a num?ber. The method of sample selection based on iailure pervasion intensity and failure rate was intro?duced. The demonstrated result by an example shows that the sample set selected based on the method can represent the failure set adequately.