| Direct torque control (DTC) system for induction motor has got wide attention with the advantages such as its new thoughts, simple control scheme and excellent performance. In this thesis the author made a study of asynchronous motor model, designed a hybrid simulation system of DTC and analyzed the characteristics of the DTC system and the existing problems detailedly.Due to the large torque ripple of induction motor in DTC system, especially in low speed, this thesis proposed a discrete time space vector modulation and designed a sliding-mode torque and flux controllers based on sliding-mode theory to overcome this drawback. On the other hand, the preciseness of stator flux estimation is the key to DTC technique, the conventional stator observation models are detailly analyzed in this thesis, and the author proposed a sliding-mode stator flux observation, which integrates the advantages of the normal models and has high robustness again the motor parameters varieties in DTC system.The study on speed sensorless is an important research field. The methods for speed identification based on Kalman filter and model reference adaptive system are deeply investigated in this thesis, a new PI adaptive method of speed identification based on MRAS is put forward. And the simulation will testify the effectiveness. Finally a speed sensorless DTC system is carried out.The DTC system is a high-performance AC driving system. But conventional speed controller in DTC system is PED adjuster, which can't make the system have enough quick response, the ability to restrict the noise and strong robustness to theparameters of the system. So the author designed a speed regulator based on sliding-mode control theory, which make the system has strong robustness to the parameters, excellent dynamic and steady-state performance. Then a torque and flux regulator based sliding-mode control theory was designed to reduce the torque ripple in conventional DTC system. Simulation results show the method is effective.
|
PDF Full Text Request