A Constrained-Optimization Approach to the Execution of Prioritized Stacks of Learned Multi-Robot Tasks

TL;DR

This paper introduces a constrained-optimization framework for executing prioritized multi-robot tasks encoded by value functions, enabling coordinated task execution with explicit priority management.

Contribution

It proposes a novel convex optimization approach using control Lyapunov functions to enforce task priorities and execution constraints for learned robot tasks.

Findings

Effective in simulation with multi-robot teams

Handles prioritized task execution robustly

Integrates reinforcement learning with control constraints

Abstract

This paper presents a constrained-optimization formulation for the prioritized execution of learned robot tasks. The framework lends itself to the execution of tasks encoded by value functions, such as tasks learned using the reinforcement learning paradigm. The tasks are encoded as constraints of a convex optimization program by using control Lyapunov functions. Moreover, an additional constraint is enforced in order to specify relative priorities between the tasks. The proposed approach is showcased in simulation using a team of mobile robots executing coordinated multi-robot tasks.