Development of Stationary Bike Using Self-learning Fuzzy Neural Network

• Main Authors: Shang-YouYe, 葉尚祐
• Other Authors: Le-Ren Chang-Chien
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/dxhhuq

碩士 === 國立成功大學 === 電機工程學系 === 104 === In order to develop new application of stationary bikes, two innovative operation modes of stationary bikes are proposed in this thesis, which are tandem-bike-like exercise system and pedaling torque control system. For the tandem-bike-like exercise system, CAN bus is used for the electric coupling of two stationary bikes. In order to operate two separate stationary bikes together like real tandem bike, a tandem bike reference model and an online self-learning fuzzy neural network controller are introduced into the system to control the pedaling cadence (revolution speed). Regarding the pedaling torque control system, the pedaling torque of the rider is controlled to follow a given torque command. With the control of pedaling torque, the stationary bike can be used for training muscles in lower limbs. Besides, torque control can prevent the rider from injuries caused by excessive pedaling. In this system, the pedaling torque is detected and a controller is involved in the system to modulate the pedaling resistance so as to adjust the pedaling torque of the rider. According to the experimental results, the aforementioned system operates well as expected.