Development of a Movable Neck Supporter with Adaptive and Learning Mechanism

点击翻译按钮获取中文摘要

Recently, problems of a nursing care and a quality of life for aged people have been raised. Therefore, the medical and health care devices are required to not only supplement a physical disability but also contribute improvement the quality of life for the user. It is considered as one of the solutions that the control and information processing technology is introduced adequately, to realize the health care device supporting the user autonomously. By the way, some of the handicapped with a high degree of disability such as the amyotrophic lateral sclerosis(ALS) are not held their head up, because of the neck muscle weakness caused by the disease.Therefore, in this work, the development of a movable neck supporter(MNS) that assists the twist and nod motion of the user's head is proposed for the handicapped with the neck muscle weakness. The MNS infers the user intention using the eye control device and assists the head motion using several pneumatic actuators, to realize expanding the field of vision and making the gestures. By the way, to control the actuator that assists the head motion is difficult. Because the degree of lesion is different corresponding to the user, and his condition is always changing at the same user. Therefore, to realize an appropriate assist motion, the controller of the actuator is required that always adjusts the user condition. Moreover, because of the life sustaining organs are concentrated in the neck, an adequate security is required. Hence, the controller is also required high reliability. Thus, the CMAC-PID control scheme that has been already proposed by the authors is installed to the MNS, because it based on the highly reliable PID controller and the control parameters are tuned by the Cerebellar Model Articulation Controller(CMAC) that is a kind of the neural networks.