An Iterative Method for Inverse Optimal Control

摘要

This paper proposes an iterative method to solve inverse optimal control with data segments provided at every iteration. The unknown objective function is parameterized as a weighted sum of features with unknown weights. Each trajectory segment is a small snippet of optimal trajectory. The proposed method shows that each trajectory segment, if effective, can pose a linear constraint to the unknown weights, thus, the unknown weight vector is estimated by iteratively incorporating all informative segments. Effectiveness of the method is shown on a simulated 2-link robot arm and a 6-DoF maneuvering quadrotor system, in each of which only small demonstration segments are available.