Coordination Of Multiple Robots For Mapping In Indoor Environment

Map Building is a very important aspect in mobile robot navigation. Multi-robot cooperation is a challenging subject which flourished these years. For the navigation technology for single mobile robot has been perfect day by day. Finishing a complex task by a group of robots can save time obviously. This article is mainly in search of multi-robot cooperative mapping;We obtain the sonar sensor model and error model from experiments. Then construct the systems of single robot mapping and the multi-robot coordination. To sum up, this article has two main advantages.1 We choose the online incremental algorithm for line extraction, and in considering of the error of sonar data, a new method of sonar grouping is used to cluster the points. Point features are extracted by the advanced triangulation base fusion. We improve on the matching rule for TBF algorithm. Put forward a new method that can prove two sonar arcs are echoed from one edge or corner. This can filter the data from the war effectively.2 We bring forward the dynamic allocate area algorithm for multi-robot geometrical map building. Triangular area is the main sub-area that robot obtain them during the exploring. This system includes the area allocation method, task allocation algorithm, robot action design, sub area utility design and global map update. Sub area utility design the peak and border utility of triangular area in order to plan the robot action. Task allocation design the cost and gain function for robot, robot action design contain four main actions and two sub actions, and demonstrate the process of these actions. An emulational experiment proves this system is efficiency and applied at the last. This method brings a new idea for multi-robot cooperative mapping in complex indoor environment.