In recent years, Japanese society has been aging, engendering a labor shortage of young workers. For a robot that is often in contact with people and which must provide safety and flexibility in nursing and welfare, the development of a soft, lightweight, and compact actuator has been sought. Particularly robots that are intended for use in fields of medical care and welfare should be safe when functioning around humans because they often come into contact with people. Therefore, a

more »

... balloon actuator (balloon actuator) has been developed for a robot hand to be used in such environments. For this study, we developed a stroke control system for the balloon actuator using a predictive functional control (PFC). The PFC is one of model predictive control (MPC) schemes, which predicts the future outputs of the actual plant over the prediction horizon and computes the control effort over the control horizon at every sampling instance. This paper reports the PFC control performance of the balloon actuator. We compared the control performance for the actuator with that of the PFC and a PID control system. Journal of Advanced Mechanical Design, Systems, and Manufacturing ratio. Moreover, it is compact and lightweight. Then the balloon actuator was applied to a robot hand [9] [10]. The robot hand must control a finger joint angle to manipulate an object with dexterity. However, for pneumatic rubber actuators including the balloon actuator, it is difficult to maintain exact control because these actuators have nonlinear properties that change their characteristics. Therefore, robust control systems such as sliding mode control and H-infinity control have been attempted for pneumatic rubber actuators [11] [12] . However, it is complex and difficult for many people to understand the robust control theory. From this reason, it has not penetrated the industry to any great degree. In contrast, PID control systems have been commonly used in industry. However, it takes much time and effort to find adequate gains that achieve good control performance for plants that have nonlinear properties, although the PID control system is a simple structure that is easy to use. In this study, predictive functional control (PFC) [13] [14] , which has high robustness and easily tuned control parameters to achieve good control performance, is applied to a balloon actuator. The PFC has one of model predictive control (MPC) schemes that have been originally developed for use in industry. Generally, the calculated load of the MPC is high because it demands a solution of a nonlinear optimization problem every sample time. Therefore, the MPC has been often applied to chemical plants, at which the sample time is long. However, calculated load of the PFC is low because its control law consists only of a basis function. Furthermore, it is easier to tune control parameters because there is only one parameter. Moreover, because the PFC control system has an internal model, it has high robustness for varying properties and disturbances. Herein, the PFC control performance for the balloon actuator is evaluated. Moreover, comparison between the PFC and the PID Control System for the actuator is conducted.