Bootstrap分析在管道泄漏状态库构造中的应用

管道泄漏诊断中泄漏状态特征库是十分重要而必需研究解决的问题。采用Bootstrap方法，通过在小样本情况下建立管道泄漏状态特征库，解决了管道泄漏诊断中无法采集到大量泄漏故障样本来建立状态特征库的难题。并把假设检验理论和模糊分类算法结合起来，设计一种Bootstrap波形筛选器，对所产生的统计模拟样本进行筛选。室内实验表明，Bootstrap筛选器能较好地选择泄漏特征波形，同时所构造的管道泄漏状态库提高了泄漏负压波状态特征的辨识能力。实践证明，该方法在管道泄漏诊断中是简单实用和可行的。

APPLICATION OF BOOTSTRAP METHOD IN PIPELINE LEAKAGE DETECTION

In pipeline leakage detection,knowledge base design should be based on massive data collection,which cannot be implemented easily in practice.Therefore,the bootstrap method is introduced to establish the knowledge base using small sampling.To select the leakage states pattern of negative pressure wave,a fuzzy cluster machine based on hypothesis theory is presented.According to the fuzzy machine's steady behavior,appropriated patterns corresponding to their states are optimized and selected.Based on these selected state patterns,bootstrap method is used to design their state knowledge base.As an example,the leakage signals of the experimental pipeline are processed and patterns are selected.Practically,it is verified that the method is simple,feasible and effective.