Multi-Agent Control Synthesis from Global Temporal Logic Tasks with Synchronous Satisfaction Requirements

TL;DR

This paper develops a control synthesis framework for multi-agent systems to satisfy complex global temporal logic tasks with strict synchronization requirements, using MILP encoding for efficient planning.

Contribution

It introduces a novel synchronous robustness metric and an MILP encoding method that efficiently ensures high synchronization and minimal control effort in multi-agent control.

Findings

The MILP encoding captures synchronous robustness with linear complexity.

Simulations validate the effectiveness of the control strategy.

The approach handles heterogeneous agent capabilities and complex task dependencies.

Abstract

This paper addresses the multi-agent control problem under global temporal logic tasks, considering agents with heterogeneous capabilities. These global tasks involve not only absolute and relative temporal and spatial constraints, but also group behaviors, including task completion times, agent capabilities, and task interdependencies such as the need for synchronous execution. The global tasks are formally formulated into global signal temporal logic (STL) formulae, and a synchronous robustness metric is designed to evaluate the synchronization quality with real values. A mixed-integer linear programming (MILP) encoding method is further proposed to realize task-satisfied motion planning with high synchronicity and minimum control efforts. The encoding method uses a logarithmic number of binary variables to fully capture synchronous robustness, leading to only linear computational complexity. Simulations are conducted to demonstrate the efficiency of the proposed control strategy.