Novel feedforward controller for straight-fiber-type artificial muscle based on an experimental identification model

• DOI: 10.1109/robosoft.2018.8404893
• Publication Date: 2018-07-09T23:06:33Z
• Authors: Ryuji Suzuki; Manabu Okui; Shingo Iikawa; Yasuyuki Yamada; Taro Nakamura
• Keywords: Feed forward; Open-loop controller; Feedforward neural network

This paper reports on an improvement to a feedforward controller for straight-fiber-type artificial muscle that can control the amount of contraction, stiffness, and contraction force by use experimental identification model. has higher rate than McKibben muscle. In previous study, we developed based mechanical However, this could not accurately stiffness or force. A feedback was necessary compensate lack accuracy, which increased system complexity. addition, calculations were so complex microcontroller keep up with sequential calculations. is practical when used in devices such as assist suit. paper, solve these problems, propose novel model whose are simpler ones. An enables improve accuracy identifying parameters Also, compare proposed one.