A User's Steps Considered Motion Control Approach of an Intelligent Walking Training Robot

Abstract

To guide the patients to follow the designed training trajectory precisely, an omnidirectional intelligent walking training robot (IWTR) is being developed in our laboratory. However, in the clinical experiments, the users usually walk with stumbling gait and at times fall behind the designed walking training trajectory of IWTR due to the incomplete walking ability. If the IWTR keeps tracking original trajectory continuously without considering user's walking steps, it will cause dangers for the users when they stop or fall behind the designed trajectory. To guarantee the patients' safety during the walking rehabilitation, a user's steps considered motion control approach of IWTR was studied in this research. A vision-based measurement of feet motion was proposed to detect user's feet position and orientation. With the measured feet information, the user's body center position and orientation were estimated relative to IWTR. A safe distance was defined to ensure the user's position in the safe walk space of IWTR during the walk training. The experimental results showed the effectiveness of the proposed method.