Probabilistic Power Flow Calculation And Optimization Planning For Active Power Distribution

With more and more distributed power access,distribution network was transformed from the passive structure into an active structure,which had a far-reaching impact on distribution network structure.On the other hand,due to wind speed and light intensity,the power output of wind turbines and photovoltaic power generation equipment presents a large random fluctuation characteristic,which increases the difficulty of research.Therefore,the following research work was carried out:(1)In allusion to the uncertainty and correlation of distributed generation output in the active distribution network,a new method of probabilistic load flow analysis was proposed based on improved sequence operation method.This method utilized the nonparametric kernel density estimation method and established a probabilistic model of distributed generation daily output considering the timing characteristics.Then,based on the present multidimensional sequence operation theory,an improved dependent probabilistic sequence method was proposed and combined with the traditional linear power flow calculation to achieve the probability distribution information of system power flow.Finally,the simulation was carried out on the improved 33-bus distribution system.Results indicated that the proposed method can achieve the probabilistic load flow calculation more accurately and has higher computational efficiency.(2)Considering a self-consumption trade pattern and a multi-interest equilibrium,this paper proposed a novel planning method.Firstly,a bi-level optimization model is designed,which is based on distribution network operators and DG owners.The sitting of new line and DG was taken as the decision variables and the comprehensive cost of distribution network operators was taken as the optimization target in the upper level.The lower level took PV capacity and storage capacity as decision variables and the comprehensive cost of DG owners.Secondly,particle swarm optimization and game theory are adopted to solve the optimization problem,in which binary particle swarm optimization was adopted in the upper level while non-cooperative Nash game method was utilized in the lower level.Finally,simulation tests were carried out on a modified 33-bus distribution system.Test results verified the reasonable configuration and access of the optical storage system can effectively reduce the comprehensive cost of the users,and also can effectively reduce the network loss of the distribution system.