五连杆足式机器人腿部机构多目标优化算法

为了减小腿部转动惯量,提高承载能力,提出了一种电驱式2自由度平面五杆机构的腿部结构。然而杆件尺寸会对机器人性能产生影响,为了得到最优五杆尺寸参数,对腿部运动学和动力学进行分析,建立关节电机峰值力矩、峰值角速度、单个步态周期内总能耗与五杆尺寸参数间的函数关系,构建面向机器人五杆尺寸参数设计的多目标优化模型,采用层次分析法确定各目标决策属性权重,将多目标优化问题转换为单目标优化问题,利用遗传算法得到机器人总体性能最优解。在ADAMS中建立优化前后的机器人腿部模型,进行行走仿真试验,并与理论计算结果进行对比,再根据优化结果进行电机选型,使机器人自重减少7.8%,有利于负载能力和续航能力的提高,从而验证了本文所提算法的正确性和有效性。

Multi-objective Optimization Algorithm of Robot Leg Based on Planar Five-bar Mechanism

In view of the shortcomings of cable-driven and chain transmission mechanism with motor rotation direction changing frequently and large load, a novel electric-drive robot leg based on planar five-bar mechanism with dual freedoms was proposed to reduce the inertia of robot leg and enhance load capacity. Considering the close relationship between the dynamic performance of robot and the dimension parameters of planar five bar mechanism, the kinematic and dynamic analyses of this mechanism was made, and the functional relationship between peak torque, peak angular velocity of joint motors, energy consumption in a gait period and dimension parameters were setup to get the optimal dimension parameters. In this way, a multi objective optimization model of the dimension parameters was got. By determining the weight of the objectives based on analytic hierarchy process (AHP) method, the multi-objective optimization problem was transformed into single objective optimization problem. Then the genetic algorithm was applied to obtain the optimal solution. Finally, the virtual prototypes of robot leg before and after optimization were built via ADAMS to conduct the walking simulation test, and the comparison between simulation results and computation results was made. The motors were selected subsequently according to the optimization results, which made the robot weight decreased by 7.8%, which was beneficial to the improvement of load capacity and battery life. These results verified the correctness and validity of the proposed algorithm which can also be applicable to the selection of brushless DC motor.