Zhufei Leng

Papers

Total Citations

H-Index

About

Zhufei Leng is a researcher advancing the frontier of reinforcement learning, with a primary focus on intelligent path planning and decision-making algorithms. Their most-cited work, "A path planning method based on noisy D3QN algorithm with N-step updates" (2025, 3 citations), tackles critical challenges in deep reinforcement learning by addressing unstable Q-value estimation and insufficient exploration during early training stages. Leng proposed an innovative N-step and Noisy Dueling Double Deep Q Network (D3QN) algorithm, where multi-step cumulative rewards replace traditional single-step updates to enhance learning stability and efficiency. This contribution is particularly significant for autonomous navigation and robotic control systems, where reliable path planning under uncertainty is essential. While early in their career, Leng’s work demonstrates a strong commitment to improving the robustness and convergence of deep RL models. Their research holds promise for real-world applications in robotics, autonomous vehicles, and dynamic environment navigation, marking them as an emerging voice in the reinforcement learning community.