Model Predictive Control Of Permanent Magnet Sychronous Motor And The Research Of Speed Sensorless Control

The permanent magnet synchronous motor (PMSM) has the characteristics of small size, high efficiency and high power density, which is widely used in aerospace, intelligent robots,servo machine tool and other fields. Model predictive control (MPC) is a new control strategy, which has simple principle and is easy to handle nonlinear constraints compared with the classical vector control and direct torque control. In this paper, MPC in permanent magnet synchronous motor is studied deeply, and their performance is compared; in order to further expand the scope of application of permanent magnet synchronous motor, this paper also analyze speed sensorless technology and the combination of MPC and sensorless technology, and the effectiveness of the methods are validated through simulation and experimental.In order to compare the performance, this paper firstly introduces the vector control, and the current vector control loop design is optimized. The control method is validated by a 2.2 kW permanent magnet synchronous motor.Then the paper focuses on the model predictive control. According to the classification of model predictive control, this paper is divided into two parts, one is model predictive torque control (MPTC); the two is the model predictive current control (MPCC). In MPTC, single-vector model predictive torque control presents larger torque and flux ripple. In order to improve the steady-state performance of the system, this paper proposes an optimized model predictive control based on the original double-vector method, which can effectively reduce the ripple of torque and flux. However, due to the selection of the second vector is not optimal, so the vector selection has improved space. In this paper, an generalized double vector model predictive control is proposed, which expands the selection range of the second voltage vector, and further improve the steady-state performance of the system. In the MPCC, to reduce the amount of computation, a fast vector selection method is proposed. Considering, the parameter dependence of MPC, this paper proposes a model free predictive current control. And same with MPTC, in order to improve the steady state performance, double vector control is introduced. On the basis of previous summary, generalized two-vector model predictive torque control is proposed in this paper generalized vector dual model predictive torque control, and combining fast vector selection, on the foundation of improving the calculation speed, steady state performance is improved.At last, this paper studys the speed sensorless control technology of permanent magnet synchronous motor. A sliding mode observer based on adaptive filtering is proposed to realize the rotor speed and rotor position, which is based on the characteristics of PMSM and previous experience. In order to improve the dynamic performance, a flux observer based on the current model and voltage model is proposed to observe the speed and angle, and the correctness of the theory is verified by simulation.All the research in this paper is based on the MATLAB automatic code generation software platform. This method is fast and efficient, and it is a kind of research method with high value.