Auxiliary Diagnosis Method of Equipment Fault Based on Case Representation and Semantic Matching
In order to solve the requirement of equipment fault rapid diagnosis,a semantic similarity matching algorithm based on sentence vector is constructed by combining case-based reasoning technology with text semantic similarity algorithm.According to the characteristics of equipment fault information,an object-oriented case representation method is presented.On the basis of optimizing text similarity calculation method of equipment fault phenomenon and reasonable weight distribution of equipment fault information,a comprehensive calculation method of equipment fault case similarity based on sentence vector(comprehensive method)is proposed.For 1000 fault cases,the experimental analysis is carried out with or without increasing the interference term.The results show that:This paper proposes that the synthesis method can achieve the best fault case matching effect with or without interference items,which can not only give the correct ranking of recommended cases according to the similarity of cases,but also have a good degree of discrimination.
case representationsimilarity calculationcase matchingfault diagnosis
NETL
NSTL
万方数据
