Exploiting the Natural Dynamics of Series Elastic Robots by Actuator-Centered Sequential Linear Programming

Series elastic robots are best able to follow trajectories which obey the limitations of their actuators, since they cannot instantly change their joint forces. In fact, their actuators can even allow them to improve over the performance of a robot with an ideal force source actuator by storing and releasing energy. In this paper, we formulate the trajectory optimization problem for series elastic robots in a new way based on sequential linear programming. Our framework is unique in the separation of the actuator dynamics from the rest of the dynamics, and in the use of a tunable pseudo-mass parameter that improves the discretization accuracy of our approach. The actuator dynamics are truly linear, which allows them to be excluded from trust-region mechanics. This causes our algorithm to have similar run times with and without the actuator dynamics. We test the accuracy of our discretization strategy using conservation of energy. We then demonstrate our optimization algorithm by tuning a jump behavior for a single leg robot in simulation, showing that compliance allows a higher jump and takes a similar amount of computation time.