Research On PCB Equipment Control And Defect Detection System Based On Industrial Internet Of Things

With the development of the Internet of things in the industrial field,the production mode of the factory is gradually changing to intelligent,but there are still many manufacturing industries still maintain the traditional production mode.Because of the lack of connectivity and "island" equipment layout,the equipment management is confused,the response delay after the equipment failure,and the real-time monitoring and fault warning can not be realized,so the equipment can only be shut down for diagnosis after the equipment failure.In addition,with the continuous improvement of printed circuit board(PCB)density,complexity and production quantity,it is easy to cause PCB appearance defects due to improper parameter setting,machine downtime and other factors in production.The traditional PCB defect detection technology has been unable to meet the needs of the current industrial production level.To solve the above problems,this paper takes PCB industry as the research object,designs a set of device control and defect detection system based on industrial Internet of things.Firstly,according to the communication protocol,the standard communication model IMes Bus communication information model is customized.Combined with the software development kit(SDK)and socket,a set of equipment control system(ECS)based on industrial Internet of things is designed.Each process equipment can be connected to the two ports open to the server through socket or SDK through the underlying TCP / IP protocol.According to the IMes Bus communication information model,it can complete the analysis of the communication messages of different businesses' equipment,and realize the communication and data flow between the standalone equipment of independent equipment manufacturers and other related equipment,such as intelligent central control system,handheld computer,card reader,scanner,automatic guided vehicle,etc,It enables the interconnection of various process equipment in the PCB production workshop,which can monitor the production status of the equipment in real time and deal with the abnormal conditions of the equipment in the production process in time.Secondly,this paper applies the object detection algorithm in deep learning to PCB defect detection,selects the yolov3 algorithm which is suitable for small object detection,establishes PCB appearance defect detection model,locates the defect location,identifies the defect category,and improves many disadvantages of traditional detection methods such as artificial eye detection,so as to achieve the purpose of intelligent detection and accurate detection.Finally,the device control system is deployed and tested in the PCB Factory,which fully meets the needs of the factory and achieves the expected effect.The defect detection model is tested on the PCB defect data set,and the experiment shows that the detection effect of the model is obviously better than that of the model based on YOLOv2 and Faster R-CNN algorithm,and also better than the traditional detection methods such as manual visual inspection.The effectiveness of the algorithm model is verified by horizontal comparison.