碳纤维复合材料-金属混合机械臂的改进动力学辨识方法  

Improved dynamics identification for carbon fiber composite-metal robotic arms

作　　者：洪博锴 孟婥[1] 张豪 孙以泽[1] HONG Bokai;MENG Zhuo;ZHANG Hao;SUN Yize(College of Mechanical Engineering,Donghua University,Shanghai,China)

机构地区：[1]东华大学机械工程学院,上海

出　　处：《东华大学学报(自然科学版)》2025年第1期119-130,共12页Journal of Donghua University(Natural Science)

基　　金：江苏省重点研发计划(BE2023070)。

摘　　要：针对三维编织的机械臂与理论设计的参数存在偏差,导致机械臂在运行时产生扰动力矩的问题,采用改进蜣螂算法(improved dung beetle optimizer,IDBO)和加权最小二乘法的改进动力学参数辨识方法,分步辨识非线性斯特里贝克(Stribeck)摩擦力矩、线性惯性力矩、科氏向心力矩和重力矩。为设计激励轨迹,引入周期性傅里叶级数和五次多项式,采用IDBO设计最优激励轨迹,获取机械臂在最优激励轨迹下的实时状态,确定其动力学参数并建立动力学模型。试验以碳纤维复合材料和45钢混合的大臂机器人为对象。结果表明:考虑非线性摩擦力矩的改进动力学参数辨识方法有效提高了动力学模型精度,与仅考虑线性摩擦力矩的常规方法相比,关节1~6的预测力矩误差均方根(RMS)减小25.6%~47.9%。Aiming at the problem that the three-dimensional woven robotic arm deviates from the theoretically designed parameters,which leads to the generation of perturbation moments in the operation of the robotic arm.An improved dynamic parameter identification method using improved dung beetle optimizer(IDBO)and weighted least squares is used to identify the nonlinear Stribeck friction moment,linearized moment of inertia,Koch centripetal moment,and gravitational moment in a stepwise manner.Periodic Fourier series and fifth-degree polynomials are introduced to design the excitation trajectory,and IDBO is used to design the optimal excitation trajectory,obtain the real-time state of the robotic arm under the optimal excitation trajectory,and determine the dynamics parameters of the robotic arm and establish the dynamics model.The results show that the improved kinetic parameter identification method considering the nonlinear friction moment effectively improves the accuracy of the kinetic model,and the root mean square(RMS)of the prediction moments of joints 1 to 6 is reduced by 25.6%-47.9%compared with that of the conventional kinetic parameter identification method considering only the linear friction moment.

关 键 词：机械臂 碳纤维复材-金属混合 动力学参数辨识 改进蜣螂算法 激励轨迹 

分 类 号：TP241.2[自动化与计算机技术—检测技术与自动化装置]