The Research On Parameter Identification And Control Method Of Direct Torque Control For Induction Motor

In recent years, with the continuous development of intelligent control theory, microprocessor technology, power electronics devices and motor speed regulation control theory, the induction motor speed regulation control system is moving in the direction of high-performance and intelligent, of which the speed regulation control system of induction motor speed sensor-less DTC (Direct Torque Control, referred to as DTC) has grown into the most promising high-performance speed regulation control system.DTC is a high-performance control method of induction motor speed regulation after FOC, with its novel control thought, simple system structure, direct torque control means, the speed observer improving the system redundancy effectively, robustness, dynamic performance and good properties, receives widespread attention and has become a hot topic of induction motor speed regulation control research areas. There are still some key technologies to be resolved in induction motor speed sensor-less DTC system, such as identification of the motor stator flux and torque ripple increase at low speed, the motor parameters affected by environmental change. Therefore, this dissertation has finished the experimental and theoretic research on the control strategy of torque and flux of speed sensor-less DTC for induction motor and the identification method of stator flux, stator resistance, motor speed etc.But the open-loop identification does not have the adaptive ability to eliminate errors, by introducing a mean compensation algorithm into stator flux parameter closed-loop identifier, this dissertation put forward a new closed-loop stator flux estimator based on PI correction. Simulation and experimental results show that this closed-loop stator flux identification method not only preserves the advantages of the mean integrator compensation algorithm solving the problem of the accumulated error of integrator, but also the closed-loop stator flux estimator has a certain degree of correction ability to uncertain disturbance, improving the accuracy of the stator flux parameter identification.Identification of stator resistance parameter in the DTC, since the identification of the main parameters, such as the stator flux, speed and torque and so on, relies on the motor stator resistance of the control object model, When stator resistance of the DTC system influenced by environmental temperature changes, it will make the control performance of induction motor DTC falling, especially at low speed. In order to solve this problem effectively, this dissertation uses fuzzy logic theory to study the relationship between the change of stator resistance and the error of stator current amplitude, and proposed a fuzzy logic stator resistance parameter identifier based on the error of stator current amplitude. Simulation results show that this new parameter identification of stator resistance not only has the advantages of simple algorithm, good real-time control, robustness etc, but also improve the identification accuracy of the motor stator flux and speed etc, and improve the control performance of the speed and torque of DTC system both dynamic and static states.Speed sensor-less DTC speed regulation system of induction motor has become the hot research topic of high-performance speed regulation system. However, speed observer has the problems of complex algorithm, low-speed accuracy seriously affected by the change of stator resistance. In recent years, in the study of realizing the speed sensor-less DTC system running of high performance induction motor, people use MRAS theory widely to solve motor control problems with characteristics of nonlinear time-varied. In the DTC research of speed parameter observed, this dissertation analyses a kind of speed observation method, and presents a DTC interactive MRAS speed based on the MRAS theory. This interactive MRAS method of speed identification using motor voltage model as reference model and motor current mode as an adjustable model constructing MRAS speed observer exchanges adjustable model reference model and the method in real-time exchange, and in line adjusts the speed and the stator resistance parameters. Simulation and experimental results show that interactive new method of MRAS speed observer using the stability theory obtain a stable, high-speed identification and improves low-speed induction motor DTC system dynamic and static performance,and the elimination of the stator resistance changes.Conventional induction motor DTC system uses flux and torque hysterics control method, only six basic voltage space vectors in360°space are used in the motor flux and torque direct control, so many problems exists such as inherent torque, flux, current ripple and switching frequency is not fixed and so on. On account of traditional control method existing problems of not fixed switching frequency and so on, this dissertation studies on a twelve voltage space vector subdivision DTC-SVM control method, the control scheme increases stator voltage space vector from six to twelve, and synthesis of any size and direction of the control voltage vector in SVM, so it has more sophisticated voltage space vector and reduce the induction motor torque ripple at low speed effectively. To improve flux and torque control characteristics of DTC system, fuzzy PI self-tuning control is used in flux and torque control of DTC system. This fuzzy PI self-tuning controller uses flux and torque error and error change rate as input, PI controller parameters is in real-time online adjusted by fuzzy logic rules, so good flux and torque control performance are obtained. Simulation and experiment show this DTC-SVM control method based on flux and torque fuzzy PI self-tuning control not only improve the system robustness and adaptability of different control objects, but also ensure the DTC system has good flux and torque tracking control dynamic quality and cause torque pulse and electromagnetic noise suppressed better.Based on establishing the traditional induction motor DTC system simulation model and the improved induction motor speed sensor-less DTC system simulation model in Matlab7.8/Simulink simulation programming environment, this dissertation simulates the two models and simulation results show that the proposed DTC method of parameter identification and method of flux and torque control are correct. Speed sensor-less DTC induction motor system test platform is constructed by using dsPIC30F6010A produced by Microchip Corporation in USA and PM50CLA120produced by Mitsubishi Corporation in Japan and many experiments is done in this test platform. In software design, DSP control software successfully is developed by using mixed C and assembly language programming method, and achieves high performance control of induction motor. A large number of experimental results show that the proposed method of parameter identification of DTC induction motor speed sensor-less and fuzzy PI self-tuning control method based on flux and torque are effective and correct.