A detection and configuration method for welding completeness in the automotive body-in-white panel based on digital twin
Hao Li, Bing Li, Gen Liu, Xiaoyu Wen, Haoqi Wang, Xiaocong Wang, Shuai Zhang, Zhongshang Zhai, Wenchao Yang
Abstract
To address the problems of poor welding completeness and inefficient configuration for defective automotive body-in-white panels, we propose a method for detecting and configuring the welding completeness of automotive body-in-white panels based on digital twin (DT) and mixed reality (MR). The method uses DT to build an MR-oriented DT framework for the detections and configuration of body-in-white panel welding completeness. We propose a method to build a DT knowledge base for panels, a Yolov4-based welding completeness detection method, and a MR-based configuration method for the welding completeness in panels. Our team develop a panel welding completeness detection and configuration system to fully validate the effectiveness of the method.