Research And Application Of Distributed SVM Algorithm Based On Hadoop Platform

With the rapid development of high-speed trains, high-speed train security issues gradually attracted people’s attention. High-speed train vibration monitoring data for the analysis of the performance of the train service provided the conditions. However, how timely and accurate fault mining properties from these massive data for fault diagnosis, given that existing problems.At the same time, with the rapid development of Internet technology in recent years, large-scale data has been produced constantly,and how to make good use of these data has become the the research hotspot.The support vector machine (SVM) is a very effective supervised method of machine learning, which is widely used in classification and egression analysis. When the traditional SVM is processing huge amounts of data, time and space complexity of the algorithm is relative high.This made the training particularly solw. In order to solve these two problems, this paper proposes an algorithm of SVM in distributed way. Because this algorithm has a strong ability to express the data features, it is applied separately to the high-speed train fault diagnosis.Firstly, we analyzed distributed support vector machine (SVM). According to the CascadeSVM algorithm, we Propose a new training model. Through a simulation experiment, we make comparisons with traditional SVM, SMO, and CascadeSVM. Considering the recognition accuracy and training efficiency, Both proved that the distributed SVM algorithm, can achieve good results when dealing with large data.Then the SVM algorithm combined with Hadoop platform, parallel distributed the SVM algorithm is constructed. And the selected criteria MNIST datasct experimental results show overall digital recognition rate of 98% and speedup increased to 3, which illustrate distributed SVM algorithm has good performance results in the recognition accuracy, parallelization efficiency.Finally we analyzed the characteristics of the high-speed train vibration data in time and frequency domain. IMFs features were extracted by using the algorithm of EEMD. And we used distributed SVM algorithm for high-speed train vibration data to extract deep fault feature and classify these failures. Experimental results show that, by the statistical results of the better channel, the train fault identification recognition rate of 96%, and the fault location identification rate of 89%.