A unified CPG-based and multi-agent control framework for low-cost quadruped robots
Likai Wu, Meina Zhang, Wenzhe Wang · 2026
Read A unified CPG-based and multi-agent control framework for low-cost quadruped robots by Likai Wu, Meina Zhang, Wenzhe Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0356505.txt.
Description
Likai Wu, Meina Zhang, and Wenzhe Wang combined two control ideas to make a cheap four-legged robot walk well. The first idea is biological rhythm generators that set each leg's beat. The second is a team of learning agents, one per leg, that nudge the rhythm to fit the ground. Their 2026 framework deliberately keeps the robot's onboard brain tiny. All four legs share one small policy network, run from simple body sensors.
In simulation and on a real low-cost robot, the system crossed obstacles, ramps, stairs, and rough ground. It did so more stably and efficiently than baseline controllers. Cameras, torque sensors, and motion capture stayed in the lab, not on the robot. The design points toward inspection and rescue robots that schools and small firms can actually afford.
Catalog record
| CREATOR | Likai Wu, Meina Zhang, Wenzhe Wang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 57427 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |