Intercepting an Agile Target with Net-Carrying Drones using Competitive Multi-Agent Reinforcement Learning

2026-07-07

AI Summary

This study addresses the problem of cooperative interception of highly maneuverable targets by multiple agile drones equipped with capture nets, formulated as a competitive multi-agent reinforcement learning task. The proposed approach employs the MAPPO algorithm integrated with Prioritized Fictitious Self-Play (PFSP) to train policies that directly output low-level control commands—namely collective thrust and body-frame angular rates—enabling end-to-end learning in a high-fidelity simulation environment. This method effectively mitigates non-stationarity and catastrophic forgetting, significantly outperforming heuristic baselines in terms of higher capture success rate, shorter interception time, and lower collision frequency. Ablation studies confirm the critical contributions of PFSP and low-level control, while emergent cooperative tactics among pursuers are observed during training.

Abstract

This article presents a solution to intercept an agile drone by a team of agile drone carrying catching nets. We formulate the problem as a competitive Multi-Agent Reinforcement Learning (MARL) task. To address the problem of nonstationarity and catastrophic forgetting of agents overfitting to the current opponent strategy, we train the pursuers and the evader using Multi-Agent Proximal Policy Optimization (MAPPO) with Prioritized Fictitious Self Play (PFSP). We train the agents in a high-fidelity simulator using low-level control commands, collective thrust and body rates (CTBR), to achieve agile flights for both the pursuers and the evader. We compare the performance of the trained policies in terms of catch rate, time to catch and crash rates, against heuristic baselines and show that our solution outperforms them. Ablation studies show that PFSP lead to more robust policies that can adapt to different opponent strategies, and that a low-level control commands are crucial for learning performing strategies in the pursuit-evasion task. Finally, a qualitative analysis of the learned behaviours highlights the emergence of cooperative tactics among the pursuers.

Problem

• Agile Target Interception
• Net-Carrying Drones
• Multi-Agent Reinforcement Learning
• Pursuit-Evasion
• Competitive MARL

Innovation

• Competitive Multi-Agent Reinforcement Learning
• Prioritized Fictitious Self-Play
• Low-level Control
• Agile Drone Interception
• Cooperative Tactics