The Research Based On Improved Association Rules Of Equipment Fault Diagnosis

The purpose of this study is through the data mining technology to achieve equipment fault management, and through the data mining association rules algorithm of continuous improvement, proposed can be more accurate judgement of equipment malfunction algorithm to reduce losses caused by equipment fault diagnosis of misdiagnosis or delay. According to the research content characteristic, uses the model analysis, comparative analysis method, qualitative analysis and quantitative analysis methods. This thesis mainly from the following several aspects to research:Firstly, on the foundation of the research background and significance, respectively from three aspects of the maintenance ways, equipment fault diagnosis technology and equipment fault diagnosis based on data mining equipment fault diagnosis to analyze the domestic and foreign research of the current situation, and expounds the equipment maintenance and fault diagnosis theory knowledge and data mining related theory which established the theoretical basis of this article.Secondly, on the basis of analyzing the association rules, this thesis puts forward the bottleneck that association rules used in equipment fault diagnosis, and raises of thoughts and weighted association rules algorithm. Thirdly, on the basis of analyzing the weighted association rules and algorithm, in view of the actual situation, put forward the restricting factors of its application in the equipment fault diagnosis, and illustrates the important roles of its important influencing factors, proposed the variable weights association rules, then describes variable weights association rules algorithm thoughts and model.Finally, according to the front proposed variable weights association rules idea and model, this thesis respectively makes weighted association rules algorithm and variable weights association rules algorithm applied to the same example, then analyze and appraise the two algorithm, which explains that variable weights association rules algorithm can more accurately to equipment failure to judge.