Research Of Fault Diagnosis Of Auto Gearbox Based On Explanation Learning

Based on the explanation of the learning strategy decision-tree inference engine, which is a kind of learning strategies to explain a priori knowledge guidance for generalization theory, decision-tree ID3algorithm structure decision-tree method as a foundation, establish a general rules for proof tree method, in the way of deduce to describe inference engine.According to the symbols of the socialist ideas, knowledge is the foundation of all intelligent behavior, the premise of a computer intelligent must make it have knowledge. Prior domain knowledge is a way to establish validated before the domain knowledge. To be easily describe the facts and rules, form the domain knowledge system. Application to forward reasoning type to produce system, which are symbols of the existence of rule base, establish a domain knowledge rule library.Based on the explanation of the study (EBL) with clear that the domain knowledge to build up the explanation of training example, which is usually to prove the theory in the way of logic. Through the examples explain structure, can describe the areas of knowledge in the specific rules clearly, and achieve a priori knowledge on the basis of intelligence. Among them, the structure is explained that exists in the form of a tree. The decision-tree learning process is the process that structure a decision-tree actually. The learning premise is must have a group of training examples, the study results are made from these training examples that made of a decision-tree.The decision-tree is based on symbols of the decision table type builted, in true data collection process, a large amount of data are continuous variables. Use practical data to establish the decision-tree, continuous variables should be change into the discrete variables, namely discretization. This article introduces a kind of have the anti-interference ability and the generalization ability of the algorithm based on quantile parameter estimation method.Base on the work above of all, take the auto transmission fault diagnosis for example, to establish with the advantages of the inference engine above, and develop the related algorithm of C language program,and get the application.