Localization for robot mowers covering unmarked operational area

Abstract

The accurate localization is significant for both the accurate terrain acquisition and the successful area covering. Our research aims at the operational area without any manmade marks, the robot utilizes its localization system to establish the digital boundaries of the unmarked operational area. According to the specialties of outdoor environment and the practical mowing requirements, the localization system with combined sensors and the localization algorithm based on neural network are designed. The results of experiment show that the designed neural network can make the accuracy of the localization well improved. With the measurements of the ultrasonic sensors, an effective error-correction method based on the database of environmental features knowledge is proposed for the robot mower to correct its position errors in the real-time coverage operation. The localization information is reliable to ensure the intelligent behaviors of the robot mower. The technical presentations in this paper can facilitate the development of the environmental robotics.