Guangran Cheng
Papers
2
Total Citations
5
H-Index
2
About
Guangran Cheng is a rising researcher in robotics and artificial intelligence, with a focus on learning-based control and visual navigation. Their work bridges the gap between optimal control theory and modern reinforcement learning, particularly for complex robotic systems. In a notable 2019 paper, Cheng proposed an optimal control scheme for robotic manipulators with input saturation and dynamical disturbances, using a single critic neural network and a nonquadratic function to handle constrained inputs—a practical contribution to real-world robot control. More recently, in 2024, Cheng introduced DGMem, a visual navigation policy that learns without any labels by leveraging dynamic graph memory, an innovative approach that reduces the need for costly supervision in autonomous navigation. While still early in their career, with papers accumulating citations, Cheng’s work demonstrates a clear trajectory toward scalable, label-efficient learning for robotics. Their research is particularly relevant for students and engineers interested in combining control theory with data-driven methods to build more autonomous and adaptive robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2