高级检索+

基于模拟退火算法的机械臂刚度辨识构型优化与实验

Method and Experiment of Configuration Optimization for Manipulator Stiffness Identification Based on Simulated Annealing Algorithm

  • 摘要: 针对机械臂关节刚度辨识的测量构型提出了新的优化方法和实验设计思路。首先,综合考虑机械臂形位以及载荷矢量对刚度辨识的影响,采用参数κ■(A)作为测量构型评价指标。在此基础上,通过合适的模拟退火优化算法得到基于κ■(A)的最优测量构型。基于设计的多向加载装置实现了载荷的优化加载。实验结果表明,与典型评价指标κ■(J)相比,κ■(A)的最优化测量构型能更好地克服多种测量误差影响,补偿后末端位置精度提高29.59%,最大位置误差降低32.71%。可应用于实际工业环境中的串联机械臂刚度标定。

     

    Abstract: Positioning accuracy is of great significance for industrial applications. Nevertheless, in actual machining operations, deformation will generate on the end-effector of industrial manipulators under external loads due to the flexibility of actuated joints. In industrial environment, the stiffness identification accuracy of serial manipulators is affected by various measurement errors. However, there is little research on dealing with the inevitable error perturbation. An optimization method and experimental design were proposed for measurement configuration of manipulator stiffness identification. Firstly, κ-1F(A) was adopted as the evaluation criterion of measurement configuration considering comprehensively the influence of manipulator posture and wrench on stiffness identification. On this basis, optimal configurations based on κ■(A) were obtained by the appropriate simulated annealing algorithm. The optimized loading was achieved on the basis of the designed multi-directional loading setup. The experimental results showed that compared with the typical evaluation criterion κ■(J), the optimal configurations based on κ■(A) can better overcome the impact of various measurement errors. After displacement compensation, the end position accuracy was increased by 29.59%, and the maximum end position error was reduced by 32.71% compared with the typical criterion set. The proposed method can be subsequently applied to the stiffness identification of serial manipulators in industrial environment.

     

/

返回文章
返回