Welding Information Remote Monitoring And Quality Evaluation System

At present,welding enterprises mainly complete the collection,monitoring and eval-uation of welding quality by manual work,which can not meet the needs of modern weld-ing technology with high cost and low efficiency.With the proposal of ”industry 4.0” in Germany,which is known as the integration of information and physical systems,China has put forward the development strategy of ”made in China 2025”.Industrial production will enter the stage of informatization,digitization and intellectualization,and the realiza-tion of intelligent production and manufacturing will be regarded as the focus of develop-ment at this stage.With the wide application of various intelligent sensors and wireless communication technologies in the industrial field,as well as the deep development of industrial technology,China's industrial production will enter the stage of informatiza-tion,digitization and intellectualization The development trend of degree learning theory in the direction of image detection is to add ”remote information monitoring and quality evaluation system” on the original basis to realize the data acquisition,remote monitoring and welding quality evaluation of welding process,which has the potential to enhance the economic competitiveness of welding manufacturing industry.In this paper,according to the trend of digital development industry and the demand of digital workshop,the welding information remote monitoring and quality evaluation system is designed.Using embedded development technology,the main control unit based on STM32F103VET6 is introduced from chip selection,chip characteristics and other as-pects.Through the hardware connection and software design of interrupt control module,temperature acquisition module,position information acquisition module and voltage ac-quisition module,the important functions including real-time temperature,longitude and latitude,altitude and voltage are realized in the welding production process Real time ac-quisition of welding information.The host computer communicates with STM32 through serial port.MFC platform in VS2010 is used for the overall design of man-machine in-terface system.Press the key to call the camera to capture the welding image in real time and upload it to the cloud storage platform.Taking TCP as the access mode and 4G wire-less communication network as the data transmission channel,the welding information collected by each sensor module of STM32 is uploaded to OneNET,the cloud platform of Internet of things of China Mobile.The UI monitoring interface is designed on the cloud platform for real-time monitoring of welding information.Restful API is used to obtain resources from OneNET cloud platform,develop background information management system,develop web application,design MySQL background database structure,real-ize welding information storage,query and display of historical data.According to the welding defects,the welding quality evaluation standard is established,and the convo-lution neural network(CNN)is constructed based on Tensorflow framework to evaluate the welding quality of the collected welding images.Finally,the hardware construction and software development test of each module and the whole system are carried out.The welding information remote monitoring and quality evaluation system can meet the design requirements,reduce the labor cost,standardize the welding production management,and improve the market competitiveness of welding enterprises.