Terrain Based D∗ Algorithm for Path Planning

Abstract

Path planning algorithms provide autonomy in mobile robots to reach targets even in unknown environments. Path planning using grid based techniques such as A∗ and D∗, are cost function based, which is primarily a function of the distance to be travelled to reach the target. Robots targeted for outdoor environments should consider the terrain features also during path planning. In this paper, a modified approach of D∗ path planning algorithm is proposed. In addition to distance to be travelled, terrain slope estimate is also used in cost function computation to plan the path. The algorithm was simulated and tested with different terrain slopes. Results with different test scenarios are also brought out.