Grid connected wind power system driven by PMSG with MPPT technique using neural network compensator

This paper presents an improved model of grid connected wind power system (GCWPS), driven by a Permanent Magnet Synchronous Generator (PMSG) involving an Artificial Intelligence control technique. The propose system uses two Insulated-Gate-Bipolar-Transistor (IGBT) based Voltage Source Converters(VSC), among them one is connected at generator side(Rectifier) and another one is connected at grid side (Inverter). In this paper, to collect a set of data, back propagation neural network trained model is employed to simulate the proposed technique and predict the Maximum Power Point Tracking (MPPT), from the knowledge of wind turbine characteristics. The merits of the proposed system is, generates the output optimal speed command for speed control loop of rotor flux oriented vector controlled scheme at grid side voltage inverter with the instantaneous active power as its input. The optimal speed commands, which track the maximum power points for the proposed system, are generated in accordance with the variation of active power output due to the change in the command speed generated by the controller. The concept is analyzed in a 2MW rating of direct driven variable speed PMSG and the random performance of the proposed system demonstrate its capability of MPPT under varying wind conditions. The proposed grid connected wind power system is modeled and simulated in Mat lab/semolina GUI environment.