Yong Yu

In this paper, we propose a thumb rehabilitation device based on the concept of Repetitive Facilitation Exercise (RFE) therapy. Our device only uses rapid passive stretching to the extensor muscle to induce a stretch reflex as a facilitative stimulation. RFE, a proven rehabilitation method, promotes neural pathway recovery through repetitive voluntary movements facilitated by passive stimulation. We proposed a simple mechanism capable of performing flexion/extension and palmar adduction/abduction RFE training movements using only a single motor and a single link, without a tapping mechanism. Our results demonstrate that the proposed control method can induce a stretch reflex in the desired muscle through rapid adduction/flexion passive stretching and facilitate active abduction/extension motion.

Repetitive Facilitation Exercise (RFE) is known as an effective rehabilitation method that can elicit brain plasticity. In our previous study, we developed a hemiplegic upper limb rehabilitation system that can train the elbow and shoulder based on RFE theory. The system consists of a selective Arm-DoF constrainable mechanism and a multiple facilitative stimulus timing and duration control system. The selective Arm-DoF constrainable mechanism has three DoF, making it possible to selectively restrict a DoF of the shoulder or elbow. The multiple facilitative stimulation timing and duration control system only has an electrical stimulation source and a vibrator and no large actuators to perform the training exercise. In this study, we undertook an evaluation of the short-term clinical effects of a proposed elbow training program using the hemiplegic upper limb rehabilitation system. The training method consists of two sets, each involving 50 repetitions of the elbow extension exercise programmed in our system. We evaluated the short-term training effects through clinical trials for five stroke patients using the above method. In the evaluation, 10 repetitions of pre/post-tests without any facilitative stimulus were performed before and after the two sets of training. As a result, the maximum angular velocities or the average angular velocities of the elbow extensions improved significantly in most subjects. These results indicate that the quickness or the smoothness of elbow extension movements in the hemiplegic limb improved. Therefore, our clinical experiments support the short-term clinical benefits of the proposed method as well as the practicality of quantitative evaluation using our system.

The walking power assist leg (WPAL) is designed for normal human power augmentation during walking. WPAL tackles three fundamental problems. First, the architectural design is studied using ergonomics. Second, WPAL must understand the user motion intention. Then, the control algorithm is also investigated. In order to make WPAL work smoothly and provide power assist for user, we proposed a method based on multi-sensor data fusion to understand the user motion intention. And, a control method of pseudo-compliance is also proposed. We can adjust the coefficient of mass, damp and spring to change power assist supply for user during walking. Correlative experimental result shows the effectiveness of the control method.

In this paper, an arm wrestling robot (AWR) is developed for amusement and developing/strengthening human arm muscles. Everyone regardless of age, sex, weight, dexterity or experience can enjoy it. The AWR arm has 2-DOF, integrated with mechanical parts, elbow/wrist force sensor, motor, encoder and driver, can measure real-time elbow/wrist force of the player during Arm Wrestling game. In this paper, in order to design an anthropopathic, harmonious and fast-response system, intelligent control instead of traditional control method used in common robot movement control is developed for the movement control problems of the mechanical arm. Finally, some experiments on AWR are conducted objectively. The validity and feasibility of this control strategy are confirmed by analyzing experimental results.

In this paper, a wearable robot gives assistance or support to human locomotion is proposed, it will be of great value for certain activities or for special groups. As the wearable exoskeleton moves according to human intention, how to acquire human locomotion information accurately and quickly become so significantly. Firstly, we make a dynamic analysis of walking human body and build a correlative dynamic model, then introduces our wearable robot which depends on human-robot interaction force to capture human motion information real-time. Its design is strictly in accordance with dynamic model. We also make a detailed introduction of the acquisition method for mechanical information from lower limbs, detection device and its principle of operation, a LAN(Local Area Network) based on CAN(Controller Area Network) bus used to acquire motion information of human lower limbs is constructed. At last, a series of experiments are carried out, by analysis of experimental data, we can clearly recognize and depict a whole process of human motion, this acquisition method is verified feasible and effective.

In this paper, we intend to introduce a concept of power assist system---walking power assist leg (WPAL), which is designed for human power augmentation. WPAL is designed according as Ergonomics and the premise is that the system doesn't affect user walking freely. To make the system work smoothly and provide power assist for user, we proposed a method based on human-robot interaction force to understand the user motion intention, and a control method of pseudo- compliance. We can adjust the coefficient of mass, damp and spring to change power assist supply for user during walking. Correlative experimental result shows the effectiveness of the control method.