信息融合的机械臂数字孪生模型迭代实验

Experimental Research on Digital Twin Model Iteration of Manipulator Based on Information Fusion

数字孪生(Digital Twin,DT)充分利用物理模型、传感器更新、运行历史等数据,集成多学科、多物理量、多尺度的仿真过程,实现虚拟空间和实体空间的双向交互映射。针对生产设备的状态监控和状态预测问题,以机械臂DT模型为例,提出了一种基于数据级、特征级和决策级的多层次信息融合DT模型迭代方法,实现了基于多层次信息融合模型迭代的机械臂状态估计。利用迭代模型库设计了机械臂轨迹预测实验,通过不同迭代周期、不同融合层次的模型迭代效果分析对比,验证了该方法的有效性。

With the proposal of smart manufacturing strategy,modern statistical-analysis methods based on big data have attracted widespread attention.Digital twin(DT)makes full use of data such as physical models,transmitted sensor data,operating history,and integrates multi-disciplinary,multi-source,multi-scale simulation processes,and realizes the bidirectional interactive mapping between virtual space and entity space.To monitor and prediction the state of production equipment,a multi-level information fusion based manipulator DT model iteration method including data level,feature level and decision level is proposed to implement the state monitor and prediction of manipulator.A series of manipulator trajectory prediction experiments are designed using the iterative model library.The effectiveness of the method is verified through the analysis and comparison of model iteration performance on different iteration cycles or fusion levels.