工业机器人动力学参数的改进遗传算法辨识

Improved Genetic Algorithm Identification for Dynamic Parameters of Industrial Robots

为了准确辨识出六自由度工业机器人的动力学参数,提出一种基于改进遗传算法的参数辨识方法。构建牛顿-欧拉机器人动力学模型,明确反映各关节力矩与动力学参数的函数关系;通过改进遗传算法获取优化激励轨迹,并对机器人进行动力学参数的整体辨识,减少关节间耦合作用影响,避免多次识别实验环境不一致而产生的误差。最后采用最小二乘法计算机器人的动力学参数,解决因初始值选择不合理而导致辨识精度受限的问题。实验结果表明:此方法得到的最优激励轨迹能够满足约束条件,缩短优化时间,有效提高动态参数辨识的准确性和有效性。

In order to accurately identify the dynamic parameters of industrial robots with six degrees of freedom,a parameter identification approach based on an enhanced genetic algorithm was proposed.The dynamic model of the Newton-Euler robot was constructed,and the functional relationship between the torque and dynamic parameters of each joint was clarified.Through the improvement of the genetic algorithm,the optimal excitation trajectory of the robot was obtained,and the whole dynamic parameters of the robot were determined.The coupling effect between nodes was reduced,and the error caused by inconsistent identification of the experimental environment was avoided.Finally,the least squares method was used to calculate the kinetic parameters of the robot to solve the problem of limited recognition accuracy due to the unreasonable selection of initial values.The experimental results show that the optimal excitation trajectory obtained by the method can satisfy the constraints,reduce the optimized time and effectively improve the accuracy and effectiveness of dynamic parameter identification.