Research On Electric Valve Operation And Maintenance System Based On Deep Learning And Cloud Platform

The development of new technology,mobile Internet,cloud computing,big data technology and other technologies have greatly accelerated the trend of industrial intelligence.China has also proposed the Made in China 2025 for the industrial field,in which intelligent manufacturing such as intelligent equipment and intelligent factories based on information physical systems is the top priority,providing a major opportunity for the transformation,upgrading,innovation and development of China’s manufacturing industry.Today,industrial sites are gradually tending to unmanned,information-based In this context,based on deep learning,cloud platform,embedded,Android and other related technologies,this paper uses neural networks and other methods to identify the internal leakage of electric valves under different working conditions.At the same time,a set of Internet of Things system for electric valves based on cloud platform is designed with C/S architecture,which is used for remote data visualization and remote valve opening control in the field of electric valves.This system provides a feasible scheme for the field of electric valve operation and maintenance.The internal leakage fault diagnosis method of the valve mostly adopts the internal leakage detector and other detection devices for detection.The detector and other devices have defects such as high detection error and inconvenient detection.With the development of artificial intelligence technology,the method of deep learning is used for detection.Its advantages such as high accuracy and greatly reducing dependence on expert experience make it gradually become the mainstream in the field of mechanical fault diagnosis.In this paper,1DCNN-CBAM convolution neural network is developed to identify the inner leakage of electric ball valve.Aiming at the problems of high noise in industrial field,such as low recognition rate,a 1DRDN DAE electric ball valve internal leakage diagnosis model was designed.Experiments show that both models can achieve high recognition rate under different working conditions.As an important component in the industrial field,valves are overhauled and maintained as important problems in the valve field.According to the requirements of the electric valve Io T system and the existing Io T-related technologies,this paper designs the overall architecture of the system,which consists of an intelligent gateway,a cloud server,and an Android client.The smart gateway is built using the Rasberry Pi4 B platform and designed a fault push terminal on the offline computer.Based on MQTT protocol and Modbus protocol,program modules such as data acquisition,protocol conversion,and data push/reception are developed.Build a web server based on the Gin framework,realize data persistence through MySQL database and InfluxDB time series database,and build Nginx reverse HTTP protocol proxy to achieve interface access.The Android client adopts the MVVM framework to build,and the App has functions such as device status detection,device data monitoring,device map/monitoring/control,device addition/deletion,device fault push,user permission management,etc.,and the system supports multi-user and multi-device.According to the actual needs of enterprises,combined with the current deep learning and Internet of Things and other related technologies,this paper designs the system,and tests the identification rate of the internal leakage fault diagnosis system,and the functional and performance tests of the Io T operation and maintenance system,and the test results prove that the system has good performance and feasibility.Figure [76] Table [17] Reference [76]...