Power assisting control with visual interaction for robotic wheelchair

This paper describes an approach to power assisting control with visual interaction. The power assisting control enables the human-friendly support while manually driving the wheelchair. However, in conventional control, the additional implementation of the operational assistance is difficult because of considering only the relationship between the human pushing force and wheelchair dynamics. In the paper, a novel strategy is presented in order to couple with the dynamics of the visual environment on the power assisting controller. In the proposed approach, both encoder-based and vision-based reaction force observer are designed for obtaining the environmental change in the field of view. The vision-based observer using the optical flow field of visual scene can estimate the external force including the remote inertial force due to the moving obstacle, and it enables the flexible design of the several operational functions without complicated visual recognition while power assisting control. The fundamental functions are evaluated by several experimental results.