Research On The Estimation Methods Of Stator Flux For Direct Torque Control

AC drive has gradually replaced the DC drive and occcupied the dominant position in the field of electric drives. Speed sensorless direct torque control technique can not only reduce the complexity of control systems, but also improve system's reliability, so it has been achieved great development. In this paper, is of using the modern control theory, the stator flux observation methods of DTC system are studied in-depth in the thesis. The major research works are as follows:The observation of stator flux is accurate or not, which has a significant impact on the speed sensorless direct torque control system. And it is an essential link of high-performance speed sensorless direct torque control system. It has taken analysis and comparison of traditional open-loop stator flux observation methods, improved integration algorithms, and adaptive flux observer. The impact on the voltage-current model and adaptive flux observer of stator resistance variations has given detailed analysis. Considering the above analysis, this paper decides to adopt the adaptive flux observer to observe the stator flux, and use the pole placement method and linear matrix inequalities iterative (LMI) approach to obtain the gain matrix, use Lyapunov stability theory to guarantee the adaptive flux observer stability.Speed sensorless control technology of induction motor is an important research direction of DTC system. Adaptive speed observer is the most commonly used method of speed sensorless control system, and the performance of speed observer depends on the choice of gain matrix and the adaptive law of speed. At present, most adaptive speed observers are using the method of pole placement to determine the gain matrix. Through analysis of the stability of the observer, we know that the speed adaptive observer is of using pole placement which is using the motor's pole as the reference may lead to instability of the whole system. The adaptive speed observer that can remain stable over its full speed region is proposed in the dissertation.The resistance value of stator windings will change with temperature in the operation process of induction motor. If the used resistance value in the control system and actual value does not match, it will have a greater error of speed and stator flux, and may even lead to deteroriation of the whole system's performance. Hence the on-line identification of stator resistance is necessary. Based on the analysis of the limitation of DTC stator resistance estimation with the pole-placement method, the stator resistance identification law on condition that the Lyapunov inequality strictly holds is proposed. Then, the online stator resistance estimation can be achieved by adding the law to the adaptive speed observer based on LMI. The system performance and the estimation precison of flux and speed are improved by feeding back the estimation value of stator resistance to DTC.The speed sensorless induction motor DTC simulation platform is built in the Matlab/Simulink7.0 environment, which introduces the space vector modulation technique to replace the traditional swithed scheme of DTC. Simulations and comparisons are made for the strategies and schemes proposed in this thesis paper. Simulation results show that speed sensorless DTC strategies proposed in this thesis paper can achieve better performance.