The Application Research Of Intelligent Algorithms In Direct Torque Control System

The paper entirely summarizes the actuality and development of Direct Torque Control (DTC) system of induction motors,and detailedly introduces the basic theory and principles. The application of intelligent algorithms in the DTC is presented and MATLAB/Simulink experiment approve the validity.In order to realize the combination of the intelligent algorithms and DTC theory,On the basic of analysis characteristic of wavelet network, genetic algorithms(GA) and BP algorithm.Mixed intelligent algorithms is put forward according to the different peculiarities of intelligent algorithms.ultimately,schemes of intelligent DTC is presented.For realizeing high performance asynchronous motor control system,speed information must be know exactly.However,the use of speed sensor not only reduce system reliability,but also add system hardware cost.So,according to mathematic model of asynchronous motor,wavelet network motor speed identification model is constructed and mixed intelligent algorithms is used to train wavelet network,mixed algorithms actualize the combination of GA and BP algorithms,improve convergence rate of network obviously.Seen from simulation results that DTC system with wavelet network speed identification hold favourable dynamic, static performance,which indicate that wavelet network speed identification may replace speed sensor absolutely in DTC.Consequently,speed sensorless drive control system is carried out.At the same time,for enhancing control performance, shorten control algorithms,wavelet network switch state control is structured in this paper,owing to side-by-side operation ability of network,delay time of control is reduce,thereby,real-time control ability is improved in DTC system.