Minimal Energy Control on Trajectory Generation

Abstract

Minimal energy control using artificial intelligence techniques is developed in this paper. A traditional feedforward neural network is used as the controller. Through learning, the controller can generate trajectory along a pre-defined path. The learning strategy is called recurrent averaging learning. It takes the average of initial states and final states after a cycle of training and sets this value as the new initial and final states for next training cycle. By including the energy criterion in the cost function, this technique can generate a minimal-energy walking gait and still follow the reference trajectory.