Heuristic optimization for global scan matching with focused sonar on a mobile robot

A robot moving in an environment may lose track of its position. The global scan matching algorithm described in this paper aims to recover from this situation for arbitrary large position errors. The algorithm compares focused sonar scans and uses heuristic optimization to provide a rough estimate of the robot position. This position estimate may then be refined using a closed-form expression. Two solutions have been shown bringing to a trade-off in performance and computation time