Reinforcement learning enhances autonomous target tracking accuracy and robustness

Researchers have developed a deep reinforcement learning approach for autonomous bearings-only tracking of moving targets. The system formulates the observer maneuver problem as a belief Markov decision process, using a Cubature Kalman Filter to represent the belief state. A reward function balances minimizing estimation error with maintaining filter consistency, and the policy was trained using a deep Q-network over 50,000 episodes. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel RL-based control policy for improved target tracking accuracy and robustness in bearings-only scenarios.

RANK_REASON This is a research paper detailing a novel application of reinforcement learning to a specific tracking problem. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.AI →

• arXiv
• Markov decision process
• Cubature Kalman Filter
• deep Q-network
• DQN

• paper
• other