PILOC: A Pheromone Inverse Guidance Mechanism and Local-Communication Framework for Dynamic Target Search of Multi-Agent in Unknown Environments

2025-07-09

AI Summary

To address the challenges of unpredictable target dynamics, high environmental uncertainty, and absence of global prior knowledge in multi-agent cooperative search within unknown dynamic environments, this paper proposes PILOC—a decentralized framework relying solely on local perception and limited inter-agent communication, thereby eliminating dependence on global environmental modeling. PILOC innovatively incorporates a pheromone mechanism into the observation space of deep reinforcement learning to enable indirect, environmental-cue-driven coordination. Furthermore, it introduces a pheromone backward-guidance strategy to enhance target localization efficiency and system robustness. Experimental results demonstrate that PILOC significantly outperforms state-of-the-art methods under dynamic target migration and communication-constrained conditions: search success rate improves by 23.6%, and communication overhead decreases by 41.2%. The framework exhibits superior adaptability and decentralized collaborative capability.

Technology Category

• Search and Optimization: Learning to Search
• Multiagent Systems: Coordination and Collaboration
• Intelligent Robots: Localization, Mapping, and Navigation

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• User Modeling, Personalization and Recommendation: ML for personalized search and recommendations

Abstract

Multi-Agent Search and Rescue (MASAR) plays a vital role in disaster response, exploration, and reconnaissance. However, dynamic and unknown environments pose significant challenges due to target unpredictability and environmental uncertainty. To tackle these issues, we propose PILOC, a framework that operates without global prior knowledge, leveraging local perception and communication. It introduces a pheromone inverse guidance mechanism to enable efficient coordination and dynamic target localization. PILOC promotes decentralized cooperation through local communication, significantly reducing reliance on global channels. Unlike conventional heuristics, the pheromone mechanism is embedded into the observation space of Deep Reinforcement Learning (DRL), supporting indirect agent coordination based on environmental cues. We further integrate this strategy into a DRL-based multi-agent architecture and conduct extensive experiments. Results show that combining local communication with pheromone-based guidance significantly boosts search efficiency, adaptability, and system robustness. Compared to existing methods, PILOC performs better under dynamic and communication-constrained scenarios, offering promising directions for future MASAR applications.

Problem

• Dynamic target search in unknown environments with multi-agent systems
• Reducing reliance on global communication for decentralized cooperation
• Integrating pheromone guidance with DRL for efficient agent coordination

Innovation

• Pheromone inverse guidance for dynamic target localization
• Local communication reduces global channel reliance
• DRL-integrated pheromone mechanism enhances coordination