Papers
14
Total Citations
753
H-Index
8
About
Guangzhu Peng is a prominent robotics researcher whose work centers on adaptive control, human-robot interaction, and neural network-based learning systems for robotic manipulators. His most significant contributions lie in the development of admittance control frameworks that enable robots to compliantly and safely interact with unknown, dynamic environments — a critical challenge in modern robotics applications ranging from industrial automation to collaborative human-robot systems. Peng's most impactful work focuses on force sensorless control, eliminating the need for costly and fragile force sensors by leveraging neural networks to estimate external forces and environmental dynamics. His 2018 paper on neural network-enhanced adaptive admittance control has accumulated 194 citations, with closely related sensorless admittance control studies from 2019 each surpassing 160 citations — a testament to the field's recognition of his approaches. He has further advanced these methods to address practical hardware constraints such as actuator saturation and input deadzone, while incorporating reinforcement learning for optimal performance in time-varying environments. More recently, Peng has extended his research toward human motion intention estimation, finite-time prescribed performance control, and robot skill learning, broadening the applicability of his frameworks to real-world collaborative tasks. His body of work collectively represents a rigorous and highly cited effort to make robotic interaction smarter, safer, and more adaptable.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10Teleoperation control of Baxter robot based on human motion capture4 citations · 2016