Papers
7
Total Citations
83
H-Index
5
About
Dongqing Wang is a leading researcher in multiagent reinforcement learning (MARL), robot audition, and robotic manipulation. His most impactful contribution is a collaborative multiagent reinforcement learning method based on policy gradient potential (42 citations), which has become foundational for gradient-based MARL algorithms addressing convergence challenges in multi-agent systems. Wang also pioneered solutions to the cocktail party problem through audio-visual sound source localization and tracking for mobile robots (16 citations), enabling robots to isolate and track specific speakers in noisy, multi-source environments. His work extends to robotic kinematics, where he solved the inverse kinematic problem for general 6R serial manipulators using unit dual quaternions and the Dixon resultant (8 citations), and to simultaneous localization and mapping (SLAM) with a multi-innovation forgetting factor-based EKF-SLAM method (6 citations). Additionally, Wang has contributed to adaptive manipulator control using RBF neural networks (5 citations) and myoelectric pattern recognition for stroke rehabilitation using wavelet packet transforms (3 citations). His research bridges theoretical advances in MARL convergence with practical robotic systems for perception, control, and human-robot interaction, demonstrating significant impact across autonomous systems and assistive robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5Adaptive manipulator control based on RBF network approximation5 citations · 2017
- 6
- 7Control System Design and Trajectory Planning for SCARA Robots3 citations · 2014