Fighter Jet Navigation and Combat using Deep Reinforcement Learning with Explainable AI

2025-02-19

AI Summary

To address the challenges of multi-target collaborative decision-making and low model credibility in air combat simulation, this paper constructs a dynamic adversarial environment using Pygame and proposes a deep reinforcement learning framework integrated with counterfactual analysis. Methodologically, we adopt an enhanced DQN architecture, design a sparse reward function, and implement hyperparameter optimization to enable end-to-end joint training for navigation, target acquisition, and aerial engagement (i.e., attack or evasion). Our key contribution is a novel explainability mechanism based on counterfactual analysis: by quantifying the Q-value discrepancy between actual and counterfactual actions, we identify critical decision drivers. Experimental results demonstrate an 82.6% mission completion rate; the agent exhibits real-time responsiveness, robustness under high-dynamics and partially observable adversarial conditions, and transparent decision-making—establishing a verifiable technical pathway for trustworthy military AI deployment.

Technology Category

• Multiagent Systems: Adversarial Agents
• Game Theory and Economic Paradigms: Adversarial Learning
• Search and Optimization: Adversarial Search

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This paper presents the development of an Artificial Intelligence (AI) based fighter jet agent within a customized Pygame simulation environment, designed to solve multi-objective tasks via deep reinforcement learning (DRL). The jet's primary objectives include efficiently navigating the environment, reaching a target, and selectively engaging or evading an enemy. A reward function balances these goals while optimized hyperparameters enhance learning efficiency. Results show more than 80% task completion rate, demonstrating effective decision-making. To enhance transparency, the jet's action choices are analyzed by comparing the rewards of the actual chosen action (factual action) with those of alternate actions (counterfactual actions), providing insights into the decision-making rationale. This study illustrates DRL's potential for multi-objective problem-solving with explainable AI. Project page is available at: href{https://github.com/swatikar95/Autonomous-Fighter-Jet-Navigation-and-Combat}{Project GitHub Link}.

Problem

• Deep Reinforcement Learning for fighter jet navigation
• Explainable AI in combat decision-making
• Multi-objective task optimization in simulation

Innovation

• Deep Reinforcement Learning for jet navigation
• Explainable AI analyzes counterfactual action rewards
• Custom Pygame simulation optimizes multi-objective tasks