The Exploratory Development Of Parameter Identification In DTC System Based On Colony Aptitude Arithmetic

The develop instance of DTC system is summarized systemically in this essay, and the basic theory of DTC system is introduced particularly. This essay brings forward a kind of aptitude control method based on NN for parameter identification in DTC system and proves its validity through experimentation.Comparing with vector control, the main advantage of DTC system is controlling the stator flux linkage and torque directly. This essay indicates that the key of DTC system is choosing stator voltage vector logically under the request of flux linkage and torque through analyzing the mathematical model of asynchronous machine.This essay ameliorates the methods of observing key parameter in DTC system through combining aptitude arithmetic and DTC theory, and validates the feasibility of combining these two theories with experimentation: trussing NN in DSP CMOS chip and observing asynchronous machine flux linkage combining rotate speed and torque.Basic NN contains the disadvantages that its convergent speed and precision are low, so this essay adopts PSO arithmetic to optimizing the operative formula of threshold value and weight in NN for increasing the convergent speed and precision of NN. The PSO arithmetic is a kind of iterative optimized arithmetic based on colony. The NN optimized by PSO can recognizes parameters of asynchronous machine effectually and quickly.This essay sets up a new method for observing the motor flux linkage according as NN which has excellent robustness. It possesses favorable dynamic response because of the parallel computing ability of modified NN. The experimental result proves the feasibility of this method, so this model can be the feedback quantity in speed governing system of asynchronous machine. At the same time this method can restrain the unstabilizing factor which aroused by stator resistance in low speed state.