The design and control of an autopilot system for motorcycles with CVT system

• Main Authors: Yu-hu Wu, 吳裕虎
• Other Authors: Yaojung Shiao
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/72262733408206933505

碩士 === 國立臺北科技大學 === 車輛工程系碩士班 === 92 === Abstract Title: The Design and Control of an Autopilot System for Motorcycles with CVT System Pages:133 School: National Taipei University of Technology Department: Institute of Vehicle Engineering Time: June, 2004 Degree: Master Researcher: Yu-hu Wu Advisor: Prof. Yaojung Shiao Keyword: CVT, Fuzzy Logic, Genetic Algorithm This research designed an autopilot system for motorcycles. The autopilot system mainly performs motorcycle speed control by controlling throttle and brake. Therefore, motorcycle tests can be performed by this system automatically instead of by qualified testers. And then testing errors can be reduced to some accepted range. The driving pattern used in this research is CNS 11386 for motorcycles driving in urban areas. A set of motorcycle dynamic models was built first in this research for simulation and controller design. These models are focused on the dynamics of CVT transmission system. These CVT dynamic models can predict dynamic performances of CVT subsystems more precisely than other often used CVT models. Detailed parameters are needed to know in these models. Besides, Matlab/Simulink is used in this research to simulate CVT system. Fuzzy Logic was applied in this research to design controllers for throttle and brake. In order to get better design of Fuzzy membership functions and gains, genetic algorithm is also used here. Therefore, optimal control parameters can be found for the controllers, and controlled performance is guaranteed. With the aids of genetic algorithm, the designed Fuzzy controller can control motorcycle speed in specified range with small errors, and make actuators run smoothly.