Robust Path Planning Against Terrain Classification Errors for Planetary Exploration Rovers
Year: 2021
Type: conference-paper
Abstract: Identifying terrain type is crucial for planetary exploration robots to move autonomously.Although terrain classification is performed using CNN (Convolutional Neural Network), the accuracy is not high due to the limitation of the data used for training.This path planning method considers the errors of terrain classification and generates a path that planetary rovers can move within the probabilistic risk specified by the operators.The simulation results show that the proposed method enables planetary rovers to generate a robust path against terrain classification errors.
Authors: 鈴木, 大和, 久保田, 孝, SUZUKI, Yamato
Institutions: Japan Aerospace Exploration Agency, The University of Tokyo
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7208230686
