High Performance Control of Multilevel Converter for Induction Generator with Variable Speed Wind Turbine

Abstract

Nowadays, wind turbine capacities are increasing rapidly. New capacities of 5 MW wind turbines now exist and yet are still increasing continuously, especially for offshore wind turbines. Generation voltage has been increased to comply with the new power demand. Multilevel converter is used to decrease switching voltage to an acceptable level. In this thesis, maximum power point tracking for variable speed wind turbine with two techniques are used. The first technique uses optimum maximum power curve fitting Pmax(Wmax) to generate an optimal reference speed. The second maximum power point tracking for variable speed wind turbine is based on adaptive neural fuzzy inference system algorithm. It predicts the optimal reference speed of squirrel cage induction generator (input of direct torque controller for the machine side multilevel converter) with wind speed variations The diode clamped three, four and five level converters with direct torque controller for the machine side converter with proportional integral controller are used. The high level converter enhances the performance of the induction generator torque, current, flux, speed, active and reactive power during start up and wind speed variations. Moreover, the total harmonic distortion is decreased compared with the classical two levels converter without the need of using complex controllers. The diode clamped three, four and five level converters with direct power controller for the grid side converter with proportional integral controller is used. This topology is used to improve the performance of the delivered power flowing from the dc-link to the electric utility-grid. It decreases ripples of the DC-link voltage and total harmonic distortion and improves the power factor of the injected power to the utility-grid. The proposed system r is done using