The speed estimation is core question of the speed sensorless direct torque control system. The torque ripple often exists in the direct torque system, because of angle discontinuity between voltage vector and magnetic flux and stepping of voltage vector generated by inverter. The torque ripple is more obvious when speed changing or load torque ripple. The torque ripple can cause speed estimation error. The error reflects the reference torque through PI regulator, therefore aggravates the torque ripple. So, how to depress influence of torque ripple on speed estimation has important research significance and application value.The dissertation starting from the mathematical model of induction motor, expounds the basic principles and structure of direct torque control and establishes simulate platform by the software named Matlab/Simulik. Then, qualitative analysis several traditional methods of speed estimation, point out their advantages and disadvantages. The paper presents a novel adaptive speed estimation method based on motor motion equation according to the problem of peed estimated error caused by torque ripple.The speed estimation scheme based on motion equation has two main sections. One is adaptive state observer, and stator flux and rotor flux is used as variables of it. The iterative LMI algorithm based on Lyapunov stability theory is used to ensure the stability of state observer. The other is speed adaptive mechanism. It adopts the speed adaptive law induced by motion equation which can directly reflect the motor operation characteristic. So it can improve rapidity of speed identification and to reduce the influence of torque ripple speed estimation. The whole system stability proved by affine quadratic stability decision method of robust control theory.The simulation result shows that: the speed is estimated very precisely and had more robust for the torque ripple. The system is capable of working from very low speed to high speed and exhibits good dynamic and static performance. |