期刊名称：International Journal of Reviews in Computing
印刷版ISSN：2076-3328
电子版ISSN：2076-3336
出版年度：2011
卷号：6
出版社：Little Lion Scientific Research and Developement
摘要：This paper presents Artificial Neural Network (ANN) based Direct Torque control scheme for Matrix Converter fed Induction Motor. Matrix Converter drives have received more attention due to its bidirectional current flow, sinusoidal input current, sinusoidal output voltage and adjustable displacement angle. Direct Torque Control scheme provides direct control of motor torque and flux and hence provides fast torque response and flux control. The DTC technique has the drawback of torque and flux ripple. The drawbacks of DTC technique has been addressed some extent by Fuzzy Logic. The accuracy may be increased by using fuzzy logic, but at the same time it can be used only a static environment. In case of dynamic decision-making, the artificial neural network is used. In this paper, the optimisation of the neural network using fuzzy inputs to estimate the reference voltage vectors using various combinations of flux position, stator flux and torque values is done. Then the ANN controller replaces space vector modulator which generates the appropriate switching states for Matrix Converter. The Lavenberg-Marquardt back propagation technique has been used to train the neural network. The simulation model of the proposed scheme is presented. The speed and torque response of the model is presented for various switching frequencies.
