Optimal Configuration Of User-Side Multi-Scenario Energy Storage For Adapting To The Dual-Carbon Target

Under the background of the "carbon peaking and carbon neutrality" dual carbon goals,the installed capacity of photovoltaic and wind energy continues to increase,and in order to smooth the fluctuation of wind and solar output,energy storage technology has also become the focus of development in the field of new energy.Among them,user-side energy storage can store new energy,reduce energy loss in the process of transmission and distribution,and improve the efficiency of new energy utilization.When the scale of new energy access is large,user-side energy storage can smooth the fluctuation of the power grid and improve the stability of the distribution network.Optimizing the configuration of energy storage capacity on the user side can guide project investment and operation,but the load characteristics differ greatly.Therefore,this paper studies the optimization of multi-scenario energy storage.Taking the user side as the starting point,a cluster generation method integrating PCA and improving k-means is proposed.For the massive load data,the Principal Component Analysis(PCA)is used to reduce the dimension,and the data of the smallest dimension is obtained on the basis of the maximum possible coverage of the original information.The reduced data is substituted into the improved k-means clustering,the maximum and minimum distance method is used to find the initial cluster center,and the Va index is introduced after clustering to test the validity of the clustering results,determine the optimal number of clusters and clustering results,and generate a typical daily load curve on the user side.Finally,simulation analysis shows that the proposed method has the characteristics of high accuracy and strong antiinterference compared with traditional k-means clustering,and generates typical daily curves for wind power and photovoltaic,which provides a data basis for the subsequent energy storage configuration.In the solution of multi-objective configuration model considering economy,technology and environmental protection,the NSGAⅡ is used to solve the optimal solution set of Pareto multi-objective model,and the optimal solution is selected by linear weighting.The weight values of each target in the model are obtained by the maximum and minimum method,and the final optimal solution with weights is the initial scheme of user-side energy storage configuration under the double carbon goal in this paper.In the case simulation,the results of the 27 scenarios obtained by the clustering algorithm are still guaranteed to be within 95% of the confidence interval,which meets the user-side energy storage capacity requirements in most cases,and can put forward effective guidance suggestions and construction planning for the user-side energy storage.Considering that the linear weight evaluation method ignores the mutual influence and interdependence between various targets,this paper proposes to improve the multi-objective optimal configuration model by using SHAP value.Based on the Pareto optimal solution set and optimal solution input into the neural network,a "black box" model is constructed,and SHAP values are introduced to analyze the potential influence between indicators.Any combination of indicators in the configuration model is made to determine whether the conditions for the use of SHAP values are met,the marginal benefit of each combination is calculated when the requirements are met,the marginal contribution of each index in the combination is analyzed,and the SHAP value of each index is obtained.The obtained SHAP value is used as the modified weight of each index,combined with the initial weight weighting,and the new weight value is substituted into the multi-objective optimization model,and the Pareto optimal solution is re-solved to obtain the final energy storage optimization configuration scheme.Combined with simulation analysis,the improved configuration scheme can be obtained with stronger comprehensive benefits and more in line with the actual situation of users.Some of the research results in this paper have been applied to the energy storage configuration technology project of a provincial power grid economic research institute.