Papers
2
Total Citations
13
H-Index
2
About
Peihao Zhu is a robotics researcher focused on advancing intelligent manipulation and autonomous navigation through deep reinforcement learning. His work centers on two key challenges: improving robotic grasping accuracy in complex environments and enhancing path planning efficiency for mobile manipulators. In his most cited work (2023, 11 citations), Zhu proposed an object recognition grasping approach that integrates Proximal Policy Optimization (PPO) with YOLOv5, addressing limitations in traditional grasping methods such as low accuracy and single-scenario applicability. This fusion of computer vision and reinforcement learning enables robots to recognize and grasp objects more reliably in dynamic settings. Additionally, Zhu tackled slow convergence in early-stage training by developing a path planning algorithm that combines obstacle avoidance with memory functions, inspired by the Artificial Potential Field method. His work on optimizing the Deep Deterministic Policy Gradient (DDPG) algorithm demonstrates a commitment to making autonomous systems learn faster and more efficiently. Through these contributions, Zhu is helping bridge the gap between simulation and real-world robotic performance, with implications for manufacturing, logistics, and service robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2