A Study On Direct Torque Control System Speed Identification Of Speed-Sensorless

In the recent 30 years, the direct torque control of asynchronous motor obtained enormous development by its simple and brief system structure, the fine static state and the dynamic property. It compute and control motor's torque under the stator coordinate system, and uses the stator magnetic field orientation control, carries on the optimizing control directly to invertor's on-off state. The rotational speed regarding alternating current machine's indirect vector control as well as other speed closed-loop control is the essential information. Using the velocity generator to obtain the rotational speed not only to increase control system's cost, but also exist installed difficulty, and reduced system's reliability and robustness. Theoretically ,we may obtain the rotational speed through survey electrical machinery's terminal voltage and the electric current, thus realizes the sensorless control. This article has conducted the research to the non-velocity generator speed estimate algorithm, solves the problem which the above brings.The thesis was focused on the design of the DTC system, the stator flux observation and the speed estimator and the simulation of sensorless DTC system.Asynchronous machine's mathematical model is the foundation to its analytical study. Firstly, established the asynchronous machine's dynamic mathematical model,in static coordinate system and given the equivalent circuit, given invertor's model under the voltage space vector , analyzed the influence of voltage space vector to the stator flux linkage and the electric motor torque, and gave the torque hysteresis comparator and flux hysteresis comparator control diagram, which can choice the switch condition signal by the table of voltage space ventor. Secondly, analyzed the present traditional asynchronous machine stator flux linkage observation method, joins low pass filter's flux linkage sight, to have the phase and the peak-to-peak value compensation flux linkage sight makes the improvement to the traditional algorithm. Thirdly, focus on MRAS,uses refers to auto-adapted method based on the neural network model to improve the traditional speed sight.Finally, established the asynchronous machine simulation model through the matlab simulation software, uses the improvement the stator flux linkage sight to improve the simulation to the direct torque control system. And joined the improvement the model to refer to the auto-adapted module to carry on the simulation to the system, confirmed the method is validity.