Mobility Assistive Robots for Disabled Patients

摘要

This chapter presents the navigation technologies of mobility assistive robots for disabled patients. The autonomous navigations of conventional robotic wheelchairs are not easy to be implemented because the inconvenience of setting target positions for each segmental path, and unavailable global localization and map of environment. Therefore, a feasible robotic wheelchair is achieved by controlling the wheelchair using the joystick, and the hybrid function converts joystick command as the virtual target for the fuzzy based navigation functions. In this manner, the proposed navigation functions become important assistive function to justify user's command and to perform more safe and comfortable wheelchair driving mechanism. On the other hand, a PSoC based distributed computing architecture is implemented to control the robotic wheelchairs. The proposed architecture is proposed to reduce the costs, size, and power consumptions and to increase the reliability of the popular PC based navigation system. Experiment results validated the PSoC based navigation functions. Based on practical experiments, the computational error is within 8 % and the time elapsed for computation and communication traffic is within 0.14 second. More specially, the proposed architecture is more feasible than the PC based navigation system. In the future, the navigation algorithm will be optimized to increase the computational efficiency and accuracy as well. At the same, in-hospital test will be planned to verify the clinical effectiveness.