Research On Fault Diagnosis Method Of Rolling Bearing Based On Wireless Sensor Network

Today,when mechanization is becoming more and more popular,most of the mechanical failures are caused by damage to rolling bearings.Therefore,real-time monitoring of the running state of rolling bearings is essential.At present,vibration signals of rolling bearings are generally collected through wired communication.The wired communication network has a stable transmission and is not easily affected by the environment,but it requires a lot of human and financial resources for assembly and maintenance.These additional inputs may even exceed the price of the sensor itself.The emergence of wireless sensor networks solves the above-mentioned problems of wired networks.In addition,because wireless sensor networks have many advantages such as easy assembly,strong adaptability,and low cost of production and application,they are widely used in various fields.This paper combines wireless sensor network and fault diagnosis technology,compares and optimizes the two fault diagnosis models,and proposes a rolling bearing fault diagnosis method based on wireless sensor networks and convolutional neural network.The wireless transmission of rolling bearing vibration data is realized between the terminal node and the central node,and the computer is used for fault diagnosis and classification.The specific work content is as follows:(1)According to the current mainstream rolling bearing fault diagnosis methods,two fault diagnosis models are constructed based on the deep learning theory,and the computer platform is used for simulation testing to find the optimal solution and determine the optimal diagnosis model.(2)Based on the NXP JN5168 development board,a hardware platform for fault diagnosis of rolling bearings based on wireless sensor networks is built.The hardware platform is composed of three parts: terminal nodes,central nodes,and computers.The terminal node uses the IEEE 802.15.4-based Zig Bee protocol to wirelessly transmit the rolling bearing vibration signal data to the central node,and then the central node sends it to the computer through the USB interface.Finally,completes the diagnosis and classification of the rolling bearing vibration signal in the computer.(3)In the computer,the software and hardware interface programs are written,including data receiving functions,callback functions,etc.,to ensure the coordination and cooperation of the two stages of data transmission and fault diagnosis.In addition,some software programs based on wireless sensor networks have been developed,including vibration signal data transmission programs on the terminal nodes,timer functions,and wireless network communication programs,and wireless network and serial communication programs on the central node.(4)Using the rolling bearing vibration signal collected by Case Western Reserve University Bearing Data Center Website,the proposed fault diagnosis system was experimentally verified.The experimental results show that the system can accurately distinguish different types of faults,and the average accuracy of multiple experimental results can reach 99.71%.At the same time,the system is trained and tested using data sets with different training-test sample ratios,which verifies that the system is not relying on a large number of training samples,it can also deal with the lack of fault samples in actual production.