Abstract:

Wind power is intermittent and random.Aiming at improving the power prediction accuracy of the power generation side and using the reactive power compensation capability of the doubly-fed wind turbine to improve the traditional reactive power optimization model,this paper first uses the particle swarm algorithm to optimize the initial weights and thresholds of the BP neural network based on the weather forecast and the historical output data of the wind farm,and uses the improved BP neural network algorithm to predict the output curve of the wind turbine.To minimize the active power loss and the voltage deviation of each branch as the objective function,considering the combination of the reactive power compensation capability of the doubly-fed wind turbine and the traditional reactive power optimization measures,the particle swarm algorithm is used to solve the reactive power optimization model.Finally,the improved IEEE 30 node system is used for simulation verification.The results show that the research has practical significance for improving the voltage safety level of the distribution network and reducing the power loss.

Key words: doubly fed induction generator, voltage security, improved BP neural network algorithm, reactive power optimization