Study On Low Voltage Fuzzy Evaluation Method For Distribution Network

With the increase of residential electricity demand,the power supply voltage quality becomes one of the main indicators of the power quality of the Low-Voltage(LV)distribution network.It is an important basis for the voltage monitoring department to find out the low-voltage users timely and accurately.In the 220V/380V low-voltage distribution network,due to the large number of nodes,the voltage monitoring facilities are outdated,the grid parameters are not fully collected,and the power flow calculation has high requirements on the parameter integrity and accuracy,which makes it difficult for the grid voltage monitoring department to control the distribution station all user voltage to achieve full coverage monitoring.Under this background,this thesis first calculates the voltage drop of low voltage distribution network based on the actual parameters of LV distribution network.By introducing "node load moment",the influence of user power on line voltage drop is approximately converted into node load Moment on the node voltage,thus establishing 220V low voltage distribution network user voltage full coverage estimation model.Combined with BP neural network to simulate the nature of unknown function based on data modeling,this thesis applies it to the calculation of full coverage of user voltage in LV distribution network.Based on the voltage measurement data of a few nodes,the voltage of most nodes that can not be measured is estimated,Obtaining the voltage estimates of all nodes in the low-voltage distribution area,the auxiliary power grid monitoring department grasps the voltage profile of the entire 220V/380V low-voltage distribution network with little data,and promptly finds the existence of low-voltage users.In order to verify the validity and practicability of the fuzzy comprehensive evaluation method of voltage coverage in this thesis,the actual operation data of the case area are collected and analyzed.The analysis results show that the voltage fuzzy evaluation method proposed in this thesis can meet the engineering practice requirements.In view of the flaws of BP neural network itself,by improving BP neural network,adding weighting factors in training process to dynamically adjust the learning rate and momentum factor of neural network to improve the characteristics of network training is not easy to converge and improve the learning efficiency.Due to the requirements of the neural network for the generalization ability of the sample,this thesis also compares the impact of the selection of different calling points on the estimation effect through actual data analysis,and draws the principle of selecting the voltage calling points that meet the actual needs of the algorithm.