Achievement of localization system for humanoid robots with virtual horizontal scan relative to improved odometry fusing internal sensors and visual information

摘要: To achieve tasks in unknown environments with high reliability, highly accurate localization during task execution is necessary for humanoid robots. In this paper, we discuss a system which can be applied to robot when executing the real world. During such tasks, robots typically do not possess referential constant horizontal plane turn used as part of fast and cost efficient methods. We solve problem by first computing an improved odometry estimate through fusing visual odometry, feedforward commands from gait generator orientation inertia sensors. This generate 3D point cloud accumulation successive laser scans then properly sliced create height virtual scan. Finally, slice observation base fed 2D SLAM method. The fusion process uses velocity error model greater accuracy, parameters are measured on robot. evaluate our world experiment using JAXON show how practical solution complex processes.