Optimal path planning using a continuous anisotropic model for navigation on irregular terrains

• Published in: Intelligent service robotics Vol. 16; no. 1; pp. 19 - 32
• Main Authors: Sánchez-Ibáñez, J. Ricardo, Pérez-Del-Pulgar, Carlos J., Serón, Javier, García-Cerezo, Alfonso
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.03.2023; Springer Nature B.V
• Subjects: Algorithms; Anisotropy; Artificial Intelligence; Control; Cost function; Dynamical Systems; Energy consumption; Engineering; Mars; Mechatronics; Original Research Paper; Path planning; Planning; Robotics; Robotics and Automation; Robots; Terrain; User Interfaces and Human Computer Interaction; Vibration; Guidance; Anisotropic; Path planning; Rough surfaces; Mobile Robots
• ISSN: 1861-2776; 1861-2784
• DOI: 10.1007/s11370-022-00450-6

Mobile robots usually need to minimize energy when they are traversing uneven terrains. To reach a location of interest, one strategy consists of making the robot follow the path that demands the least possible amount of energy. Yet, its calculation is not trivial with irregular surfaces. Gravity makes the energy consumption of a robot change according to its heading. Such a variation is subject to the terramechanic characteristics of the surface. This paper introduces a cost function that addresses this variation when traversing slopes. This function presents direction-dependency (anisotropic) and returns the cost for all directions (continuous).. Moreover, it is compatible with the Ordered Upwind Method, which allows finding globally optimal paths in a deterministic way. Besides, the segments of these paths are not restricted to the shape of a grid. Finally, this paper also introduces the description and discussion of a simulation experiment. It served to analyse what kinds of terrain motivate the use of anisotropy. The Ordered Upwind Method was executed on a virtual crater with different terrain parameter configurations, using both isotropic (direction-non-dependent) and anisotropic cost functions. The results evince how in certain situations the use of an anisotropic cost function instead of an isotropic one produces a path that reduces the accumulated cost by up to 20%.