Model Predictive Task-Priority Control using Control Lyapunov Functions

Abstract

Redundant robotic systems allow for the simultaneous execution of multiple tasks. It is often desirable to have priority between these tasks, such that lower priority tasks do not interfere with higher priority tasks. Many existing methods for task-priority control, while providing stability and strict priority between tasks, do not take into account the overall performance of the closed-loop system. This paper presents an optimization-based framework for dynamic task-priority control of redundant robotic systems, providing strict priority between tasks while improving the performance of the closed-loop system relative to a user-defined performance metric. The method is based on using a nominal dynamic task-priority control law together with a hierarchical control Lyapunov function based model-predictive control method. The proposed method is validated in simulation on a redundant robotic system, where it is shown to provide improved performance over existing dynamic task-priority control methods.