Bearing Cloud-edge Collaborative Monitoring System

The monitoring of key components of important equipment has always been an important part of ensuring the safety of equipment.Bearings,as important parts of various mechanical equipment,should also be designed with corresponding stable and reliable monitoring systems to meet demand.Due to the large amount of collected data,traditional monitoring systems are prone to data congestion and energy consumption problems when transmitting data.In addition,the traditional monitoring system is usually a one-way data flow,and the versatility is poor,and the function is relatively single.A feasible solution is to perform data dimensionality reduction and algorithm logic layering through a multi-processor and multi-network structure,and transfer the results to the cloud network after completing the collection processing and fault diagnosis in the edge network system.As a remote control terminal,the cloud network can not only store and display data,but also update the parameters and function configuration of edge devices.Cloud-edge collaborates to complete the monitoring function of the overall bearing.This article first designed the overall system architecture according to the needs of the system,determined the basic hardware selection and related network protocols;designed the basic network topology and the command interaction protocol for terminal equipment and relay equipment,according to the terminal equipment Corresponding embedded software logic is designed with the function of the relay equipment,which realizes the data collection and dimensionality reduction processing functions of the terminal equipment,the relay pre-alarm and fault diagnosis functions,and the network data upload function.The fault diagnosis method is mainly used to identify the bearing status and fault degree.After determining the model parameters on the PC side,it is implanted in the edge device,and the feature with strong classification ability is selected as the model input from the multi-dimensional features,and the cross-validated BP neural network classification is adopted The algorithm obtains the final diagnosis result through multi-model fusion decision,which has a high accuracy rate in the experiment.The cloud is mainly composed of remote servers and databases deployed in Alibaba Cloud.By building a server logic architecture,edge network data storage,OTA remote control,data visualization,and related interactive functions are realized.Designed a complete bearing cloud-edge collaborative monitoring system,The main innovation is:using multi-processing cores and multi-layer distributed algorithm structure to improve processing efficiency,transplanting fault diagnosis to the edge network to improve real-time performance;proposing feature extraction suitable for monitoring systems And the fault diagnosis algorithm improves the system reliability;the unique relay pre-alarm algorithm logic is proposed to improve the reliability of the alarm;the cloud control system and the command interaction protocol are designed,so that the system can be flexibly configured according to the needs,and the general purpose of the system is improved.Sex;designed a related remote server architecture,providing a good interactive interface and data visualization functions.Finally,the experiment was designed and verified the function of the monitoring system.Finally,summarize the content of this article as a whole,and look forward to future research.