Generative Design of a Calf Structure for a Humanoid Robot Based on Gait Simulation

As a rapidly developing robot type, humanoid robots still have the problem of excessive mass of structural components. Most of the existing lightweight methods usually add dangerous load conditions in the robot movement to optimize, but ignore the load conditions at other moments in the movement process. In this paper, a new lightweight design method for humanoid robot is proposed by combining gait simulation and generative design. The walking motion simulation of the robot is conducted to obtain the load conditions of the robot calf during the movement. Then, load conditions at multiple moments are applied for generative design, and the optimized model of the calf is obtained. Furthermore, the calf model designed by this method is compared with the topology optimization model and the generative design models under dangerous conditions. Finally, the knee joint operation test is carried out. The results show that the calf model got from the generative design method based on gait simulation has better lightweight effect and mechanical performance than the other models. Therefore, the lightweight design method of humanoid robots combining generative design and gait simulation is effective and promising.