A Path Prediction Method for Human-Accompanying Mobile Robot Based on Neural Network

摘要

This paper presents a path prediction method for human-accompanying mobile robot such as robotic wheelchair, domestic robot and tour guide robot. An accompanying human is detected using an in-vehicle laser range sensor (LRS). A new filter gets a smoothed track from raw LRS data of human footprints. Back propagation neural network predicts future positions of the accompanying human from the human track. Based on the future positions, a cubic spline generates a future path of the accompanying human. The experimental result validates the feasibility of the path prediction method.