Implementation of Odometry with EKF in Hector SLAM Methods

Map building for plain spatial soundings, such as a long and straight corridor in simultaneous localization and mapping (SLAM) is a challenging problem because of lacks of distinguishable landmarks. Such an environment is highly possible to induce erroneous mapping results, such as alias problems. This paper presents a scan matching algorithm with odometer prediction using Extended Kalman Filter (EKF) and an optimal path planning based on regression subgoals. The scan matching process can relax the problems of local minima by means of an effective correction in the odometrical information. By iterating odometrical corrections in each step of running motion model, the matching result can be better than one only believes in individual information from scanning or odometry. Meanwhile, an optimal path planning utilizing an A * algorithm with a regression method is introduced to ensure a mobile robot be able to move elaborately around the corner and speed up along a straight line. Experiments in an indoor environment have been conducted to verify the effectiveness and validation of the proposed techniques.