Research On Short-term Wind Speed Forecasting Based On PSO-Elman Optimized By Association Rule

Energy is an important guarantee for the growth and development of human society.It can promote social progress and improve the living standard of mankind.But the deterioration of environment has become increasingly serious,causing the attention of countries around the world.As an environmentally friendly renewable energy source wind energy is one of the effective ways to alleviate the current energy shortage and environmental degradation.Wind power has been the community’s universal attention and vigorously develop.But the randomness and volatility of the wind can lead to the instability of the wind power,which affects the quality of the power and the reliability of the power system,and has serious impact on the grid dispatch,operation and control.Accurate short wind speed prediction can provide the basis for real-time power grid scheduling and control,and can enhance the safety,reliability and controllability of wind power system.Elman neural network,as a typical dynamic local recursive network,has the ability to map the dynamic characteristics by storing the internal state,and can simulate the dynamic change of the system better.But the algorithm has the initial value of the sensitive,easy to fall into the local extremes of the shortcomings.In this paper,an optimized PSO-Elman algorithm is proposed.The algorithm is based on historical experience,and encourages the particles in the particle swarm algorithm to improve the search ability of the algorithm according to the best position of the learning experience.And the optimized particle swarm optimization algorithm is added to the Elman algorithm to optimize the threshold and weight of each layer.The optimal solution is found by particle swarm optimization to improve the overall dynamic performance of the algorithm.At the same time,combined with the association rule analysis to consider meteorological factors.Analyzing the correlation of wind speed and related to the weather factor temperature,base station pressure,sea level pressure.The Apriori algorithm was used to mine the association rules of wind speed and other meteorological factors,and the wind speed forecasting value was modified and compensated by the association rules.Finally,a Short-term wind speed forecasting based on PSO-Elman optimized by association rule is proposed and used in the example to analyze short-term wind speed.The experimental results show that the predicted model has a significant improvement compared with the single prediction model.At the same time,it is proved that the combination of the association rules considering the meteorological factors can reduce the wind speed prediction error,can be used in wind speed short-term wind speed prediction.