【摘要】 针对目前智能制造越来越复杂，在构建数字孪生系统过程中其数字孪生体几何模型在系统实时仿真过程时需进行大量的渲染分析计算，占用大量资源，从而导致在实际应用中无法保证数字孪生系统的时效性的问题，提出通过折叠边高斯曲率改进以二次误差作为度量代价的边折叠模型简化算法，实现了对焊接车间设备几何模型轻量化和优化处理。通过仿真实验对比分析，得出改进QEM(quadic error metrics)算法相较于原始QEM算法的模型简化效果更优，改进QEM算法能够保证模型几何细节特征尽可能保留且模型未出现错面、破损的同时，简化模型的数据量进而减少模型渲染对系统资源的占用，同时保证系统的仿真要求和时效性要求。更多还原

【Abstract】 With the increasing complexity of intelligent manufacturing workshops, the development of digital twin systems faces growing computational challenges. During real-time simulation, the geometric models used in digital twins require intensive rendering, analysis, and computation, leading to high resource consumption and compromising system responsiveness. To address this issue, this study proposes an improved edge collapse model simplification algorithm, which incorporates folded-edge Gaussian curvature as a weighting factor within the quadric error metric(QEM) framework. The approach is applied to optimise and lighten geometric models of welding workshop equipment. Comparative simulation experiments demonstrate that the improved QEM algorithm achieves more effective model simplification than the original method, preserving geometric detail while avoiding surface faceting or structural distortion. Furthermore, the reduction in data complexity significantly lowers rendering resource demands, ensuring both the accuracy and timeliness required for practical digital twin applications.更多还原