Structural state mapping and visual analysis of industrial robots based on digital twins

Abstract

With the rapid development of intelligent manufacturing, digital twin technology is increasingly widely used in industrial robot technology. However, traditional finite element simulation is computationally intensive, making it difficult to meet the needs of real-time interactive systems. To resolve the contradiction between computational accuracy and real-time performance, this paper proposes a lightweight digital twin framework based on “offline simulation + online mapping” to achieve efficient structural response prediction and visualization. First, this study establishes a 3D model of the industrial robot and imports it into Abaqus software for multi-condition static analysis, generating multiple sets of offline databases. Then, an inverse distance weighted interpolation algorithm is introduced to estimate the results under uncalculated conditions based on existing data. Finally, Unity3D is used to achieve visualization. Experimental results show that the proposed method improves computational efficiency and accuracy. The proposed method compresses the finite element solution time from tens of seconds to milliseconds, meeting the requirements of real-time interaction. Accuracy verification shows that the inverse distance weighted interpolation algorithm maintains a maximum relative error of less than 3% at unsampled posture nodes. Furthermore, data analysis reveals a significant “spatial clustering” characteristic in the maximum equivalent stress, mainly concentrated in the base, wrist joint, and elbow joint regions. These results demonstrate that the proposed system not only achieves high-fidelity real-time visualization and monitoring of structural response but also provides quantitative theoretical guidance for structural optimization and low-cost sensor network deployment in practical engineering.