Fault Diagnosis Analysis of Electromechanical Equipment Based on Parallel Integrated Machine Learning Algorithms Under Restricted Conditions of Integrated Data Features

In today's industrial context of an increasing number of electromechanical equipment, this paper presents a series of studies on fault diagnosis of electromechanical equipment through parallel integrated machine learning algorithms. Firstly, the quality of data features of electromechanical equipment is analyzed by normality test and correlation analysis, and the analysis results show that the feature correlation of the studied fault diagnosis dataset is not high. Then, two machine learning algorithms were designed to carry out the fault diagnosis analysis of electromechanical equipment under the constrained conditions of data features, one using the original BP neural network algorithm, and the other a parallel integrated learning algorithm based on the BP neural network. The average accuracy of the test results shows that the integrated learning BP neural network algorithm is significantly better than the original BP neural network algorithm, but due to the limitation of the quality of the dataset, the prediction accuracy of both is not high. In order to verify how the algorithm performance of the integrated-learning BP neural network algorithm designed in this paper performs under normal datasets, the algorithm performance is verified with a set of public datasets for automotive evaluation, and the final average accuracy is 91.68593, which verifies the superior performance of the integrated-learning BP neural network algorithm.