Research on the Positioning of Quality Problem of EMU Parts

At present, the positioning of the quality problem of EMU parts in our country is still in the stage of manual investigation, which is lack of big data as the technical support, so it not only tests the quality and experience of the relevant personnel but also costs lots of manpower and time. From the EMU parts quality data, it is of great significance to position the quality problems and prediction through big data technology. In order to solve these problems, this article proposed a method for the analysis of EMU quality based on the Chameleon clustering algorithm. Compared with traditional Chameleon algorithms and improved M- Chameleon algorithms, it not only avoids the difficulty of manually entering k and fixed thresholds, but also reduces many of the computational time. Based on the Hadoop, the parallelization of the algorithm is realized to further improve the clustering quality of the algorithm and the algorithm's big data analysis ability. It uses the original quality data of EMU parts for the test and analysis. The experimental results show that the MR-Chameleon algorithm has a shorter running time and a higher accuracy of clustering. It realizes the positioning and prediction of the quality problems of the quality problems of the EMU parts.