An improvement of D∗ lite algorithm for planetary rover mission planning

This paper presents a mission planner for planetary rover based on a new algorithm. The paths returned by the algorithm consider the stationary solar charge, the communication with the work station on earth and the complicated environment which can affect daily mobility activities of the planetary rover. Most grid-based planners use discrete state transitions that constrain an agent motion to a small set of possible headings. As a result, the optimal grid path is actually a suboptimal path. Our approach uses Euclidean distance to estimate the cost and then converse to the time of work pattern, after scanning the visibility of two points. Consequently, it is well suited to planning smooth trajectories for the planetary rover.