| As one of the cleanest renewable energy sources,wind power has been widely used all over the world.However,the unpredictability and volatility of wind power bring a series of technical challenges to the stable operation of power system.In order to solve the above problems,this paper studies from two aspects: one is to improve the accuracy of wind power prediction and establish the uncertainty set of wind power to quantify the impact of its uncertainty.The other is to study the multi-energy complementary system and establish a robust optimal scheduling model to improve the wind power consumption level under uncertainty.The main research work of this study is as follows:(1)A wind power prediction method based on data preprocessing,long-term and short-term memory neural network and power curve modeling is proposed to improve the accuracy of wind power prediction.Aiming at the problems of incomplete extraction of relevant features and covering up important variables caused by a single algorithm for predicting wind speed,a short-term wind speed prediction method combining cluster analysis discrete wavelet transform mutual information wind speed data preprocessing and short-term memory neural network is proposed.Firstly,the daily clustering of wind speed is carried out according to the meteorological data,and then the components of different frequencies are decomposed by discrete wavelet transform to improve the regularity of wind speed data;Considering the different correlation between input variables and different frequency components,the mutual information method is used to screen important variables as the input of prediction.After data preprocessing,the long-term and short-term memory neural network is used to predict wind speed.Finally,the differential evolution algorithm is used to optimize the four-parameter logic function to realize the power curve modeling of wind power.Simulation results show that the proposed method has higher prediction accuracy than the mainstream prediction methods.(2)In order to promote wind power consumption and optimize wind power dispatching,advanced adiabatic compressed air energy storage technology is introduced,and a robust optimal dispatching model of wind fire energy storage system based on advanced adiabatic compressed air energy storage is established.Considering the uncertainty of wind power prediction,the uncertainty of wind power in the model is reflected in the form of uncertainty set.At the same time,combined with the peak shaving and valley filling of energy storage,the dispatching of power system is improved.The model comprehensively considers the energy cost,environmental cost and reserve capacity cost,and promotes the consumption of wind power through energy cost and environmental cost.(3)Aiming at the problem that the robustness and economy of conventional robust optimal scheduling are difficult to balance and the solution is too complex,a robust optimal scheduling method of wind and fire storage based on limit scenario and wind power prediction uncertainty is proposed.Firstly,the uncertainty set of wind power is established by using the limit scenario method.Then,the training data of wind power prediction is used to fit the probability density function of prediction error,and the compensation cost outside the dispatching plan is obtained.Afterwards,the comprehensive cost is calculated based on the operation cost and compensation cost.Through the leverage of uncertain set-in operation and compensation costs,the optimal comprehensive cost is obtained.The proposed method is verified on the IEEE39 node test system.The results show that the proposed method not only takes into account the operation and compensation costs,but also considers the uncertainty of wind power output,and realizes the balance of system robustness and economy.At the same time,the proposed method also solves the problem that the establishment of uncertain set of traditional robust optimal scheduling is too simple or too dependent on probability distribution. |
PDF Full Text Request