| Under the background of the " double carbon " strategic goal and the " whole county photovoltaic " policy,the development of China ’s distributed photovoltaic industry will show a leap-forward growth trend during the 14 th Five-Year Plan period.However,its largescale access has brought many problems to the planning and operation of the distribution network,including node voltage violation,power flow reverse transmission,etc.,which seriously affect the security of the distribution network.At the same time,the traditional passive management mode is difficult to adapt to the power grid operation under the influence of uncertain factors such as illumination and load.In order to further improve the hosting capacity of distribution network economically and reasonably,it is necessary to take into account certain economy and consider active management measures for comprehensive control.In order to make more efficient use of distributed photovoltaic on the basis of ensuring the safe and stable operation of distribution network,this paper studies the evaluation method of distributed photovoltaic hosting capacity in distribution network.The details are as follows:Firstly,the construction of source-load temporal scene is studied.Based on the principle of K-means scene clustering algorithm,the density center and DBI index are introduced to improve the algorithm flow.The improved clustering algorithm is used to cluster the annual time series of photovoltaic and load power,and its randomness and variation law are analyzed.According to the clustering results,the practicability of the improved K-means algorithm is verified.Taking the clustered scene and probability as the calculation scene of the distribution network hosting capacity research can greatly reduce the complexity of the operation while ensuring a certain accuracy.Secondly,the maximum hosting capacity of distribution network under different photovoltaic access conditions is studied.A mathematical model with the goal of maximizing photovoltaic installed capacity is established,and various constraints are considered.Different photovoltaic access modes are defined,and four photovoltaic distribution characteristic indexes of distribution,capacity dispersion,distance dispersion and comprehensive uniformity are constructed.The whale optimization algorithm is improved,and this method is applied to calculate the distributed photovoltaic hosting capacity of distribution network under different photovoltaic access states.Based on the IEEE33 example,the effectiveness of the proposed method is proved.At the same time,the guiding significance of the defined photovoltaic access state to the large-scale access of distributed photovoltaic is analyzed from both qualitative and quantitative perspectives.Finally,the evaluation method of distribution network hosting capacity considering economic cost is further studied.According to the principle of four active management measures,the mathematical model and constraint conditions are established.Considering the comprehensive economic cost of distributed photovoltaic access,a multi-objective optimization model of hosting capacity is constructed.The Pareto optimal solution set is obtained by the multi-objective optimization algorithm based on NSGA-II,and the optimal solution is selected by the comprehensive evaluation index of the optimal solution set.The simulation on the improved IEEE33 node system verifies the improvement of the hosting capacity by various active management measures and the change of the hosting capacity after taking into account the comprehensive economic cost,so as to further promote the economic and safe access of distributed photovoltaic. |
