Research On Power Quality Inprovement Method For Distributed Photovoltaic Output Prediction And Optimized Configuration

In today's society,fossil energy is increasingly in short supply,and new energy development technologies such as photovoltaics and wind power are becoming more mature.Among them,distributed power supply has begun to access a large number of grids to provide consumption in production and life due to its flexible installation and low cost.Distributed photovoltaic power generation is particularly important.However,due to the intermittent and uncertainties of photovoltaic power generation,when a large number of distributed photovoltaics are connected to the grid,it will have a certain impact on the original power grid current,which will cause some power quality problems.The impact of distributed photovoltaic grid-connected power on the power quality of 10 kV distribution network is analyzed in this paper.The power quality of distribution network with distributed photovoltaic system is evaluated.The two improvement measures of power quality are from the power supply side and distribution network.Side raised.The spatio-temporal distribution of distributed photovoltaic power and node load power is analyzed.Difference in three power quality of voltage deviation,harmonic and voltage fluctuation of distribution network were studied.Based on the comprehensive evaluation index of power quality,a 10 kV distribution network optimization model was established.The power quality problem caused by the random fluctuation of distributed photovoltaic power generation and its generation mechanism are studied,and the installation location and network structure of the distributed PV grid-connected system in the medium and low voltage distribution network are further optimized.At the same time,the actual grid structure and data of the 10 kV distribution network of the Rangxia Line in Songyuan City,Jilin Province were used for theoretical verification and prove the correctness of the model.Several important factors affecting photovoltaic power generation are analyzed in this paper.Based on the meteorological information provided by the weather forecast,a short-term power prediction method for distributed photovoltaic power generation on the power supply side is proposed,and a short-term prediction model of distributed photovoltaic power generation is established.Historical output data and meteorological data for photovoltaic power plants are used as input variables.The particle swarm optimization least squares support vector machine algorithm(PSO-LSSVM)is used to solve the model and output the predicted power of photovoltaic power generation.At the same time,compared with the traditional BP neural network algorithm,the results show that the prediction results of PSO-LSSVM are more accurate under the premise of small sample size.The simulation results show the correctness and applicability of the model.