Adaptive and robust wheel-legged biped robot for semistructured community tasks

Abstract

This article presents the development and optimization of a wheel-legged biped robot designed for community inspection tasks. The robot combines the efficiency of wheeled robots with the adaptability of legged systems, making it suitable for navigating diverse terrains in semistructured environments. The main innovations of this research include the development of an integrated control system that combines balance control, state estimation, and terrain adaptation using multisensor fusion. The robot’s mechanical structure is designed to withstand jumps and falls, while its electronic hardware and software architecture ensure real-time control and robust performance. Experiments demonstrate the robot’s capabilities including tracking velocity as high as 2 m/s, resisting severe disturbance or slippage, and traversing steps over 10 cm automatically. The results show that the proposed control methods and hardware design effectively ensure the robot’s robustness and adaptability, validating its potential for practical community inspection applications. This work contributes to the growing field of autonomous robots in public service and provides a foundation for future research and development.