模型故障诊断中对广义约束传播技术的改进

IMPROVEMENT OF GENERALIZED CONSTRAINT PROPAGATION TECHNIQUE IN MODEL BASED DIAGNOSIS

在模型故障诊断中，不管采用那一种诊断方式，均首先需要根据测量结果和领域知识产生出最小冲突集合，广义约束传播技术是最有效的产生最小冲突集合的方法之一．本文使用测点元件影响集的概念，对广义约束传播技术进行了改进，改进后的广义约束传播技术可提高诊断的效率．

In model based diagnosis, to any diagnosis methods, it is the first step to generate the minimal conflict sets according to the measuring results and the domain knowledge, and the generalized constraint propagation technique is the most effective method in the process. In this paper, based on the conception of component set affecting measuring points(DSAMP) ,the generalized constraint propagation technique is improved and diagnosis efficiency is advanced.