Pose estimation-based path planning for a tracked mobile robot traversing uneven terrains

Most path planners for ground mobile robots for traversing uneven terrains use single probabilistic measures that characterize the terrains alone considering at most the robot dimension to search for optimal paths. Though these measures can approximately give a sense of the terrain ruggedness or the traversal difficulty, they overlook the details on the interaction between the robot body and the terrain, which arguably represent better the actual traversability of the robot over uneven terrains. The present work represents such interaction as the robot’s tip-over stability for the quasi-static case, which in turn is computed by first estimating the robot’s pose. The problem of robot pose estimation is formulated as a linear complementarity problem (LCP) and solved using the Lemke’s method, which guarantees a fast convergence. Three RRT-like algorithms are employed over various uneven terrains considering the robot pose estimation problem, with the purpose to compare their planning performances (such as the planning time and the associated cost values) as functions of the parameters of both path planning and the robot pose estimation for a tracked mobile robot employed in an ongoing project named FRAUDO.