Wind Power Prediction Strategy Based On Ant Colony Algorithm And Particle Swarm Optimization

With the development of the wind power,the studies on the wind power thchnique are deeponed all over the word.Due to random of wind energy resources, the fluctuations and intermittent in wind power make difficult for large-scale wind power accessing to power system, especially, when the wind power changes, wind power is likely to bring harmonic pollution, voltage fluctuation and flicker problem to grid connection, and bring a lot of problems for power generation and scheduling plan. It is not conducive to spinning reserve of power system, lowering operating costs and improvement of wind power penetration.So the wind power prediction is necessary.In this paper , the studies on wind power prediction in existence is comprehensively summarize and their defects are analyzed.On the basis of the existing neural network methods, the improvements are made,multi-layer forward neural network based on intelligent optimization is established.According to the characteristic of the hidden layer nodes,the structural gene is conceived and is computed by the improced ant colony algorithm.In the light of the trait of the network connection weights and thresholds,the parameter gene is designed and is calculated by the particle swarm optimization.It is shown that compared with the traditional prediction strategy,the improved neural network improves the forecast accuracy,which is more adapt to wind power prediction.