The Sensorless Control System Of Air-conditioning Compressor Permanent Magnetic Motor In Electric Vehicles

In recent years, energy crisis and environmental pollution become more and more serious. Electric vehicles which are environmental clean and efficient are developing at a very fast speed, it’s very popular and necessary to research the electric vehicles which using the PMSM as a compressor. The vector control of PMSM has a fine performance such as high power density, high efficiency and wide speed range. Accurate position and speed data of the rotor are necessary in the motor control, usually, sensors such as a resolver or an encoder can provide this information. However, mechanic position sensors will bring a lot of disadvantages, these sensors are hard to be installed, sensitive to environmental constraints and expensive. Therefore the elimination of the speed sensor has been considered as a very attractive prospect.Firstly, this dissertation builds the mathematical model of PMSM, analyses the vector control, introduces the SVPWM modulation technology and the principle of PI controller and provides the method to test motor parameter. Secondly, all kinds of sensorless control theories are studied theoretically, especially the high-frequency signal injection, sliding mode observer, Luenberger observer and open loop control. These position detection algorithms are compared. Thirdly the phase-locked loop(PLL) which is successfully used to estimate the rotor speed is designed, the optimum parameter is chosen to make the system steady.Finally, considering the specific load characteristic and non-salient-pole property of the compressor, the open loop control in low speed and Luenberger observer in high is chosen as the research topic and between the two methods a hybrid control is designed to achieve the smooth switching. A Matlab simulation platform is developed and the experimental platform based on the STM32 is implemented. Also the overall structure of software algorithm and hardware circuit systems is designed. This dissertation introduces the main hardware units, such as the power supply, sampling circuit and driving module, presents the software’s main structural framework, Luenberger observer model algorithm and sampling point setting method.According to the control strategy the experiment and the simulation is presented. The simulations results and experiments results validate that the combined observer can estimate the rotor position and speed precisely and the system has the feasibility, also it improves that the system is stable and robust. the dissertation realizes the sensorless control of PMSM.