Development of Vehicle Controls based on the Gravity-Shifting Signals

• Main Authors: Chun-lung Wu, 吳俊龍
• Other Authors: Chung-Neng Huang
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/09490286173708742497

碩士 === 國立臺南大學 === 機電系統工程研究所碩士班 === 98 === In the last years, most motion controls for vehicles have been developed for people and improving the traditional systems with special needs. For example, by using joystick to control assistant robot system, for disabled persons, has been developed. However, the traditional methods, using joystick, steering wheel, or keyboard, between humans and machines still can not match the requirement of those with severe disabilities. In order to satisfy this requirement, the new method is developed like Brain-Computer Interfaces (BCI), visual information and voice recognition. For example, using voice recognition to control some wheelchair or applications by means of basic voice commands, writing a text in “speech and spell” applications etc... On the other hand, the other options, likely based on visual information for detecting gesture or eye movements with an on-board computer, are inconvenient and difficult to control for the elderly or the handicapped. Besides, they are high cost. The purpose of this study is to design and develop a new controller for solving above problems. First, the steering motion control by using gravity shift signals is proposed. Here, the angle of body slope can be measured by tilts as the input signals for control. The control concept is possible to be developed in the use of military, entertainment, and society welfare etc... Second, in order to simplify the transmission mechanism, a wheel motor, also called hub-in motor, which is directly mounted inside the wheel so that the transmission gears and differentials are eliminated with associated energy loss. In addition, it is difficult to build up the dynamic model of wheel motor, and install rotational speed sensor, the neural-network (NN) is adopted to the instead of the dynamic model of wheel motor for sensorless motor control. To detect angles by using tilt sensor, but outputs of tilt sensor are different according to the variation of environment. In [27], for the fuzzy controllers, with stable and reliable output performance, are suitable for real-time speed control, it is used as an input signal processor of this system. The implements of system setup include designing the A/D converter, motor driver circuit and the main body of the system, a tricycle with two front wheels and one rear wheel (2F/1R). Most of the methods concerning with sensorless estimation need to use the Digital Signal Processor (DSP), and large number of mathematic computation. For above reasons, the microcontroller for sensorless estimation method is proposed in this study to reduce system complexity and costs. The effectiveness of this proposal has been confirmed through simulation study and experimental implementation. The mean absolute percentage error is about 1.05%.