Frontier-Based Exploration Planning for Amphibious Robot with Gaussian Process Occupancy Map
Bo Li, Feng Pan, Zhang Shi, Xiaofei Song, Rongxin Cui · 2023
The autonomous observation of amphibious environments necessitates the robot to possess the capability of simultaneous exploration and mapping. In this article, we present an efficient exploration planning framework for amphibious robot. This method utilizes the Gaussian Process Occupancy Map (GPOM) as a representation of the environment to improve map quality, and Rapidly-exploring Random Trees (RRT) as a global frontier detector for rapid detection of frontiers. Our proposed method has been evaluated through extensive simulations and real-world experiments utilizing a wheel-propeller integrated amphibious robot. The experimental results demonstrate that our approach outperforms other methods in terms of exploration efficiency.