Estimation of optimal moving target of mobile robots by using a topological map update
Woojin Lee, Sangseok Yun · 2021 21st International Conference on Control, Automation and Systems (ICCAS) · 2021
In this paper, we propose a methodology for estimating the optimal moving destination for the mobile robot to autonomously perform the objects delivery service in an indoor environment. The process for position estimation is to create a generalized Voronoi graph composed of nodes and links from a grid map, and then perform object recognition through an object recognition module using an RGB-D sensor to extract object information and location information. Then, the obstacle shape and distance information are calculated using the laser sensor. From the extracted object position, size, and obstacle information, the optimal candidate area in which the robot can move is estimated, and the optimal final position within the area using a hybrid technique of center of gravity and thinning is selected. This location information goes through the process of moving or generating nodes through geometric distance comparison with existing nodes on the grid map according to the rule-based node location selection method, so that the mobile robot updates the optimal node location for the final destination.