Trajectory Planning of a Humanoid One-legged Robot with Tendon Elastic Actuation during Squatting after Landing

In order to improve the energy storage effect of the humanoid one-legged robot during squatting after landing, this paper adopts tendon elastic actuation (TEA) and proposes a trajectory planning method using elastic elements for energy storage. The objective is to plan an appropriate joint trajectory and convert part of the mechanical energy into the elastic potential energy of the TEA unit, so as to reduce the negative work done by the actuator. Firstly, the TEA unit is modeled, and the relationship between the elastic potential energy change and the drive torque of the TEA unit is obtained through analysis. Secondly, a capture experiment is carried out on the landing motion of humans, and the constraint conditions of joint trajectory are obtained after data preprocessing. While satisfying the joint trajectory constraints, the energy storage requirements are transformed into joint drive torque requirements and angular acceleration requirements for trajectory planning. Finally, the simulation experiment shows that, compared with rigid actuation (RA), the advantage of TEA is that it reduces the negative peak power and accumulated negative work of the actuator, and stores 34.7% of the energy into the elastic element of the system during squatting after landing. The method adopted as well as the results obtained can provide significant assistance for the application of TEA.