改进金豺狼算法的柔性关节机器人运动学参数辨识

Flexible Joint Robot Dynamic Parameter Identification Based on Gauss Mutation Golden Jackal Optimization

为提高柔性关节机器人运动学参数辨识效果,首次将金豺狼优化算法(GJO)用于运动学参数辨识中,并针对GJO存在的不足,提出了高斯变异金豺狼优化算法(GMGJO)。GMGJO是在GJO的基础上,引入了高斯变异策略,增强了种群多样性和跳出局部最优能力,3个基准测试函数验证了GMGJO在寻优精度、寻优速度和寻优稳定性上比GJO更优。利用GMGJO进行了柔性关节机器人运动学参数辨识的仿真分析和实验验证,结果表明,GMGJO在收敛速度和辨识精度上较其余几种方法更具优势。

In order to improve the identification effect of kinematic parameters of flexible joint robot,the golden jackal optimization(GJO) was used for the first time,and the gaussian mutation golden jackal optimization(GMGJO) was proposed to solve the shortcomings of GJO.On the basis of GJO,GMGJO introduced gaussian mutation strategy,which enhances the population diversity and the ability to jump out of local optimum.Three benchmark functions verify that GMGJO is better than GJO in terms of search accuracy,search speed and search stability.The simulation analysis and experimental verification of kinematic parameter identification of flexible joint robot are carried out by using GMGJO.The results show that GMGJO is superior to other methods in terms of convergence speed and identification accuracy.