2 자유도 전동기 설계 및 센서리스 제어

Abstract

Double degree of freedom motor which is the main subject in this dissertation is a one of the Mlulti degree of freedom position control actuator that is capable of realizing the movement of a human joint system. Considering the narrow mounting space of its application, such as robot joint systems or adaptive front lighting systems (AFLS), reducing the total volume of the double degree of freedom motor and its control system is essential. The main contributors for the bulky system are the sensors. Widely used sensors, such as rotary encoders and resolvers, mounted on the motor shaft to detect rotor position significantly increase the size of the motor system. To make matters worse, the double degree of freedom motor requires at least two of these encoders which lead to an even greater volume increase. One of the effective means to resolve this issue is to apply sensorless control technology to get rid of these encoders. The term sensorless refers to the control method that does not require rotary encoders or resolvers to obtain position information. In general, the sensorless control engineering falls into two main categories which are the back EMF method, and the high frequency signal injection method. Considering the tilting movement of the double degree of freedom motor, the high frequency signal injection method which is effective regardless of rotor speed was chosen to be applied into this system. The designing of the double degree of freedom motor was carried out in a way that is appropriate to implement high frequency signal injection sensorless control as well as minimize cogging torque to allow high precision position control performance. The basic mechanism of high frequency signal injection method utilizes the inductance change corresponding to the rotor angle. Therefore, the rotor and stator of the multi degree of freedom motor was designed to have different inductances at different tilting angles. This dissertation not only proved the feasibility of the sensorless control via finite element analysis simulation results but also designed the controller for the DDOF Motor and proposed an algorithm which estimates the position without sensors. Finally, the TMS320F28335DSP experimental results, which prove the validity of the proposed sensorless control method, are included in the latter part of the thesis.