Multi-Agent Feedback Motion Planning using Probably Approximately Correct Nonlinear Model Predictive Control

2025-01-21

AI Summary

Coordinating formation control and obstacle avoidance for multi-robot systems in dynamic, uncertain environments—characterized by stochastic dynamics and sensor noise—remains challenging. Method: This paper introduces, for the first time, Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC) into distributed multi-agent feedback planning. It proposes a novel terminal cost function incorporating gyroscope-inspired obstacle avoidance modeling, enabling joint robust inference against model mismatch and measurement noise. The approach relies on distributed receding-horizon optimization and cooperative feedback control, requiring no global information. Contribution/Results: Simulation results demonstrate that the method outperforms centralized NMPC under severe measurement noise, achieving significantly improved formation stability and obstacle avoidance reliability. Moreover, it is scalable to high-dimensional nonlinear systems, offering a principled framework for robust, decentralized multi-robot coordination under uncertainty.

Technology Category

• Intelligent Robots: State Estimation
• Multiagent Systems: Multiagent Planning
• Planning, Routing, and Scheduling: Planning under Uncertainty

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

For many tasks, multi-robot teams often provide greater efficiency, robustness, and resiliency. However, multi-robot collaboration in real-world scenarios poses a number of major challenges, especially when dynamic robots must balance competing objectives like formation control and obstacle avoidance in the presence of stochastic dynamics and sensor uncertainty. In this paper, we propose a distributed, multi-agent receding-horizon feedback motion planning approach using Probably Approximately Correct Nonlinear Model Predictive Control (PAC-NMPC) that is able to reason about both model and measurement uncertainty to achieve robust multi-agent formation control while navigating cluttered obstacle fields and avoiding inter-robot collisions. Our approach relies not only on the underlying PAC-NMPC algorithm but also on a terminal cost-function derived from gyroscopic obstacle avoidance. Through numerical simulation, we show that our distributed approach performs on par with a centralized formulation, that it offers improved performance in the case of significant measurement noise, and that it can scale to more complex dynamical systems.

Problem

• Multi-Robot Coordination
• Formation Control
• Sensor Uncertainty

Innovation

• Distributed Multi-Robot Planning
• PAC-NMPC
• Gyroscope-Assisted Control