Yilin Wang

Guilin University of Aerospace Technology

Papers

Total Citations

H-Index

About

Yilin Wang is a rising researcher at the forefront of intelligent robotics, specializing in deep reinforcement learning for autonomous navigation. Their work focuses on overcoming fundamental challenges in robot path planning, particularly the instability of Q-value estimation and inefficient exploration during early training stages. Wang's most impactful contribution, "Improved Double Deep Q Network Algorithm Based on Average Q-Value Estimation and Reward Redistribution for Robot Path Planning" (2024, 9 citations), integrates deep neural networks with reinforcement learning to surpass Q-learning's limitations in continuous spaces. Building on this, their recent "A path planning method based on noisy D3QN algorithm with N-step updates" (2025, 3 citations) introduces an innovative N-step update strategy that replaces single-step rewards with multi-step cumulative rewards, combined with noisy exploration layers, to stabilize learning and enhance exploration efficiency. These advances are critical for real-world robotic applications requiring adaptive, collision-free navigation in complex environments. Wang's work demonstrates a clear trajectory of methodological refinement, from improving value estimation to tackling exploration-exploitation trade-offs, positioning them as a promising contributor to the growing field of learning-based robot control.