개인용 이동수단의 모바일 역진자 실시간 자세제어에 관한 연구

Abstract

In these days, a number of systems become complicated according to customers' demands. At the same time, it is now getting difficult to meet a variety of demands with the existing control theories as the non-linearity in the systems increases. The purpose of this research is to model the nonlinear attributes and design the fast and precise controller, which can maintain stability when it reaches the target. The controller should also be able to come back to the target from the unstable conditions caused by random external input. In this thesis, two-wheeled transportation has been designed as follows. The inverted pendulum control theory was utilized to design 6 axis-acceleration and gyro sensor, which are integrated as one. A motor controller for direction setting using a joy stick and operation was also manufactured. Because transportation system is limited by filtering and mathematization of sensor signal, system is unstable according to in-external condition of that. The PID value should be adjusted to stabilize the system according to the attributes of the real-time transportation. The stable driving method utilizes non-linear control algorithm to design a controller, which finds the PID gain value by analyzing the drive speed for the robot’s drive stability and information obtained from the steering sensor. The attitude control on the two-wheeled robot through the analysis suggests a possibility of applying this onto other robots’ attitude control and various systems. Since the robot is operated vertically on two wheels, there must be a consideration for errors to drive it on vertical attitude. These errors are measured from the angles, drive algorithm and sensor signals. The two-wheeled transportation was handled stably through the sensor-signal filtering and controller algorithm design. In addition, the sensor only uses 6-axis acceleration and gyro sensor in order to get precise an gles. In order to maintain the vertical attitude of the two-wheeled robot, a system should be designed to compensate for the changes of angles and correspond to it properly. Through the experiment, the stability of the personal mobility has been affirmed through the analysis on the sensor state according to the PID control gain.