Short-term Wind Power Prediction Method Based On Wind Speed Variation Characteristics

The use of renewable energy,especially wind energy,is highly regarded by governments and the private sector as it is considered to be one of the best and most competitive alternative energy sources in the current energy transition that many countries in the world are currently adopting.Wind power plays an important role in reducing greenhouse gas emissions and reducing global warming.Another contribution of wind power is that it allows countries to diversify their energy mix,which is especially important for countries with a large share of hydropower.The uncertainty of wind energy has become the most important problem restricting the development of wind power generation,and the wind power prediction model can solve this problem well and provide relevant departments with better information to support the decisionmaking process.Based on this,this paper has carried out corresponding research on the topic of short-term wind power prediction.The main work is as follows:A short-term wind power prediction model based on similar curve cluster and SVM model is proposed.Firstly,the similar curve clusters are extracted from the historical wind speed sequence.The similarity degree is used as the similarity criterion.The similarity between a large number of historical wind speed sequences and the test set wind speed sequence is judged,and then the similarly good wind speed curve clusters and curves are found.The power corresponding to each wind speed point in the cluster is taken as the final training sample,and then the SVM model is used to predict the wind power.The comparison test with a wind field in Shanghai shows that the method reduces the redundant information of the training data,saves the training time of the model,and can significantly improve the accuracy and efficiency of short-term wind power prediction.It has practical significance.From the perspective of error analysis,the proposed similarity-SVM model is used as the predictive model to analyze the overall characteristics of short-term wind power prediction error,and the shortcomings of the model in short-term wind power prediction are analyzed.Then,the relationship between the change of wind speed and the absolute error is further analyzed.At the same time,the speed change is defined as the wind speed change rate sequence.The Pearson correlation coefficient is used to analyze the correlation between the wind speed and the absolute error sequence.Provide a theoretical basis in the predictive model.Finally,it is verified that the prediction error distribution of the model accords with the Gaussian distribution,which provides the theoretical basis for the Gaussian regression process as the prediction model.A short-term wind power prediction model based on wind speed change rate and Gaussian process regression is proposed.Taking into account the influence of the magnitude of the wind speed and the change of the wind speed on the prediction error,the wind speed value and the wind speed change rate are simultaneously used as inputs to the prediction model.At the same time,considering that the error of short-term wind power prediction accords with the Gaussian distribution,Gaussian regression model is used as the prediction model in order to utilize the information of the distribution characteristics of the error as much as possible.The experimental results show that the proposed model has better stability and better tracking performance of actual power,which enhances the generalization performance of the model for wind power data prediction.