Digital Twin-based Body-in-white Panel Welding Completeness Detection And Configuration Method Research

With the rapid development and application of a new generation of information technology.The global manufacturing industry is accelerating towards digitalization,networking and intelligentization.Intelligent manufacturing is also receiving more and more attention from various countries.The digital twin,as an important means of realising intelligent manufacturing,is a digital representation of physical entities.By constructing a bridge between the physical world and the virtual world information mapping,it enables virtual-real interaction and iterative optimisation.As one of the key pillar industries of China's economy,the level of intelligence in the manufacturing of the automotive industry restricts the development of China's automotive industry.The intelligent production line of automobiles is one of the important scenarios for the application of digital twin technology and mixed reality.Body-in-white panels are the core components of a complete vehicle.The material completeness of the welded panels is one of the most important factors affecting the efficiency and quality of the body-in-white welding.Panel welding completeness means that some specific panels require material such as bolts,nuts,locating pins etc.to be welded to specific holes before they can be welded to other panels to form parts.When there are no missing materials,the panel is welded in place.This paper takes body-in-white panels as an object.The digital twin technology is applied to the detection and configuration of the welding completeness of the panels.It effectively improves the welding efficiency and quality of body-in-white.(1)A digital twin-based framework for detecting and configuring the welding completeness of body-in-white panels is constructed.Using body-in-white panels as the object,the target detection algorithm and mixed reality technology are studied and a framework for detecting and configuring the welding completeness of body-in-white panels is proposed.The relationship between the functions and layers of each part of the framework is analysed.(2)An improved Yolov4-based method for panel welding completeness detection is proposed.The Yolov4 target detection algorithm is improved for the detection of weld completeness of panels,a body-in-white panel weld material data set is constructed,and a neural network model is trained to accurately detect the weld material completeness of panels.(3)A digital twin and mixed reality based panel welding completeness configuration method is proposed.A digital twin knowledge base of body-in-white panels is constructed for the repair of defective panels for welding completeness.The Vuforia-based virtual reality registration method enables the fusion of reality and reality of the panel to generate repair solutions.(4)The system was developed to detect and configure the completeness of panels for soldering.A Holo Lens2 and unity3 d development environment was built to create humanmachine interaction and data transmission functions.The feasibility of the functional modules for panel welding completeness detection,configuration and reporting of abnormal information is verified with the body-in-white panel front surround components as the experimental object.Through comparative analysis,the effectiveness and rationality of the system technology in this paper is verified.