| The sensorless control technique of the permanent magnet synchronous motor(PMSM) has been the hotspot of the motion control research area in recent 30 years. A lot of valuable researches on rotor position estimation method are carried out by researchers in China and abroad, obtaining abundant achievements. However, such research achievements have not been adopted in practical application fields until now. In this paper, based on the comments on the PMSM sensorless control technique, fractional order control and sliding mode variable structure control(SMVSC), the scientific problems needing to be solved urgently in this research area are refined and listed as follows.(1) The bottleneck problem of the PMSM sensorless control technique is the starting of the motor. In the PMSM sensorless control system, the rotor position is provided by the rotor position estimation device, according to which the stator magnetic field is configured to be synchronized with the rotor magnetic field. Applying all these estimation techniques, the rotor position is calculated according to the sampled stator current and voltage. However, if the motor does not rotate, no such signals can be sampled and the rotor positon cannot be calculated. Therefore, the rotor position can be estimated only when the motor is rotating.(2) The rotor position estimation is the key issue in the PMSM sensorless control system. The estimation performance should be measured according to proper evaluation index. However, the current researches are focusing on analyzing the estimation accuracy, while another important index, the estimation range, is rarely mentioned. Therefore, in order to improve the estimation performance, more evaluation index should be considered, and new theories and methods should be applied to develop new estimation technique.(3) The controller design is the major issue of different kinds of control systems. In order to satisfy the rising control demand, applying different kinds of advanced control theories to solve the problems in practical application is the eternal goals and tasks of cybernetics.The scientific problems mentioned above are studied in this paper. The creative works mainly includes:(1) A control strategy combining separately-control pattern and self-control pattern is proposed, based on the deeply study and analysis of the internal operation mechanism of PMSM. When the motor speed is low, the separately-control pattern is applied and the stator vector is configured to force the motor to start and accelerate with constant acceleration; Once the motor speed reaches the stage when the rotor position can be accurately estimated, the motor speed is kept constant and the control pattern is switched from separately-control to self-control. The hysteresis function is applied to be the switching function, improving the fault-tolerant ability of the switch operation. In the constant speed process, the power angle is under closed-loop control, reducing the risk of losing synchronism during control pattern switching. When the self-control pattern is applied, according to the estimated rotor position, the stator magnetic field is configured to rotate synchronously with the rotor magnetic field.(2) The rotor position estimation technique is used to replace the rotor position detector and provide the rotor position information. Based on the deeply analysis of resolution, the major performance index of the position sensor, taking the sensorless sliding-mode control system of PMSM as an example, two performance indexes of sliding-mode observer are proposed: observation accuracy and observation dead zone. On this basis, the fractional order sliding-mode observer is designed, tuned and proved to be converged. Simulations and experiments are performed to compare the proposed observer and the augmented Kalman filtering algorithm. Simulation and experimental results show that the fractional order sliding-mode observer achieves better estimation performance.(3) The fractional order sliding-mode controller is designed using the frequency-domain method, composing the overall fractional order control system of PMSM with the fractional order sliding-mode observer and the fractional order model of PMSM. Simulations and experiments are performed to compare the fractional order controller and the integer order controller. Simulation and experimental results show that the fractional order sliding-mode controller achieves better control performance.In order to cover the practical operation range of the motor, the set values of the motor speed in the simulations and experiments in this paper are 200, 1000 and 1500rad/min, respectively. In this paper, the starting problem of the PMSM sensorless control system is solved. Besides, both the rotor position estimation performance and the speed control performance are improved. |