Design Of Software System For Motor Vibration Analysis Based On Cloud Computing

The motor is a very important power equipment in the production activities.Establishing a reliable and practical motor vibration analysis system is the key to ensure the healthy and smooth operation of the motor equipment.At present,the motor vibration analysis system has a single vibration analysis method,is subject to local computing power,and has poor information synchronization.In this paper,based on the cloud computing platform,a set of vibration analysis software system is designed for the above problems.It mainly includes the design of vibration analysis method combining spectrum and acceleration envelope spectrum,the research of intelligent vibration analysis method based on convolutional neural network,the research of denoising and compression of motor vibration signals,and network transmission design and concurrency design of software systems.The specific work is as follows:This paper first solves the problem of preprocessing and compression transmission of system vibration signals.The motor vibration signal has a high sampling frequency and a large amount of data.If the network transmission is directly performed without processing,not only the network bandwidth is required to be high,but also the storage pressure on the cloud is large.A vibration signal data compression method lossless compression based on wavelet decomposition combined with bz2 lossless compression is proposed.The information loss is small and the compression ratio is high.It has a good effect on the vibration signal of the test bench motor.It helps greatly improve the real-time performance of system vibration signal data transmission.The analysis of the motor vibration signal uses a combination of two analytical methods.Based on the signal analysis method,the spectrum analysis and acceleration envelope spectrum analysis of the motor vibration signal can be combined with the fault characteristic frequency to perform early fault diagnosis,and provide a large number of motor status information for experienced fault diagnosis personnel.For non-professionals with no relevant experience,accurate vibration analysis results can be obtained,and another intelligent vibration analysis method using empirical mode decomposition and deep neural network is proposed.Using Tensorflow to achieve the model,the accuracy of the model for motor fault identification after the motor vibration data training reaches 100%.Finally,the motor vibration analysis system is deployed on the cloud computing platform.Considering the practicability and security of the software system,the Socket network connection method and the I/O multiplexing method are adopted to reliably transmit the packet to prevent packet loss and achieve high concurrency.After systematic testing,the software system has high reliability and good stability,which can meet the standards of engineering application.