Design And Implementation Of An Intelligent Analysis System For Time Series Failure Data

Time series analysis is a commonly used method for processing dynamic data,which can reflect the development trend of current objects.The time series fault data reflects the degree and status of the fault change of the object over time.Through the analysis of the time series fault data,the characteristic change law can be discovered,so as to determine whether the current data source is faulty and the loss of the fault degree.This paper first uses the vibration signal generated by the bearing of the rail vehicle as the data source,combined with the characteristics of the time series fault data,and proposes a fault classification model Multi-scale-VMD-PSO-DDBN-BP(MVPDD)based on the time series fault data.This model first selects an appropriate step size to perform multi-scale processing on the signal,and then decomposes the processed signal into multiple IMFs components through variational mode,and then inputs the decomposed IMFs components in parallel to the dualscale deep belief network.Among them,the output of the dual-scale deep belief network is superimposed as the overall output.The particle swarm optimization algorithm and the deep belief network are combined to determine the number of neurons in each layer of the neural network.Finally,the BP neural network is used to reverse the entire model.To fine-tune,use the Softmax classifier to classify the signal.Compared with the experimental results of other models,the classification model MVPDD proposed in this paper has higher accuracy and effectiveness.The traditional vehicle bearing fault detection method mainly relies on the work experience and technical level of the inspector,which reduces the fault detection accuracy to a certain extent,and the detection is time-consuming and labor-intensive.This paper proposes an online real-time and efficient remote bearing fault diagnosis system,which uses a variety of detection methods including MVPDD model and traditional detection methods,adopts the idea of microservices,and uses the Spring Cloud framework and big data related technologies to build a distributed system.The guarantee system has the characteristics of high reliability and easy expansion.Inspectors can remotely and real-time check whether the equipment is malfunctioning through this system,which not only improves the inspection efficiency,but also reduces the hidden dangers of the enterprise.