An Approach to Guidance Motion by Gait-Training Equipment in Semipassive Walking

Recently, population aging and declining birth rate have become very serious, and people who lack walking capability have truly increased in Japan. To address this issue, gait training has become a requirement for them. Although some training systems have been researched in the past, a training system that entails guidance of walking has not been researched well. This paper focuses on guidance of walking, and it is aimed to develop a training system that utilizes the activity and passivity of the trainee. In addition, rhythmic walking, which takes walking stability and trainee safety into account, is aimed for. In this paper, an improved electric wheelchair is used as gait-training equipment (GTE); then, the trainee is towed by the GTE, and walking is guided. In this paper, a targeted walking cycle is predefined, and the GTE must distinguish the cycle in order to guide walking. Therefore, this paper uses the center of gravity (COG) of the trainee as the index. Thus, the GTE tows the trainee when the swing leg moves more forward than the stance leg; then, rhythmic walking can be guided. In addition, this paper proposes a method to control the GTE by a proportional-differential controller based on virtual compliance that is composed of mass, spring, and damper. The gait training can be safely achieved by deciding the desired position and velocity of the GTE from the tow force through virtual compliance. In this paper, the validity of the proposed method is verified by numerical simulation and experiment. The validity of the proposed system is verified by using a four-link robot whose geometric expression is known as the trainee; then, the application of the proposed method will be extended to a human trainee by only adjusting the parameters.