GRID CONNECTED WIND POWER SYSTEM AND SENSORLESS MAXIMUM POWER POINT TRACKING CONTROL METHOD THEREOF USING NEURAL NETWORK

摘要

PURPOSE: A grid connected wind power generation system using a neural network and a sensorless maximum power point tracking control method thereof are provided to control a duty ratio of an input voltage of a boost convert in order to follow a maximum power point of a generator by optimizing the width of a fuzzy membership function by learning the neural network using an output power and an output voltage of the generator. CONSTITUTION: A grid connected wind power generation system comprises a blade(1) which generates a mechanical energy rotated by a wind speed; a generator(2) converting generated mechanical energy into electrical energy; a rectifier(3) rectifying the electrical energy; a boost convert(4) which boosts and outputs an output voltage of a current transformer; and an inverter(5) supplied to a grid by converting the output voltage of the boost convert into an alternating current. An output power fluctuation and an output voltage fluctuation of a generator varied according to the wind speed are inputted. A fuzzy membership function width of the output power fluctuation and a fuzzy membership function width of the voltage fluctuation are determined learning through the neural network. A fuzzy control for the output voltage fluctuation and the output power fluctuation of the generator is performed. A neuro fuzzy controller(6), which outputs an input voltage of a boost convert in which a duty ratio is controlled in order to reach a generator power output electricity to a maximum power point, is included. [Reference numerals] (3) Rectifier; (4) Boost convert; (5) Inverter; (6) Neuro fuzzy controller; (AA) Wind; (BB) Direct current terminal; (CC) Grid