Influencing And Controlling The Grid-connected Point Voltage Of Photovoltaic Generation System With Power Output Forecasting

In recent years,photovoltaic power generation has become a new growth point of renewable energy generation after wind power generation.As the fluctuation of the output of the Photovoltaic generation system(PVGS),it has a serious impact on the safety and reliability of the power grid,power quality and economic operation.Therefore,the accurate prediction of PV output can grasp the output of PV grid-connected power system in advance and lay a foundation for the rational distribution of active and reactive power of PVGS to improve the voltage quality of grid-connected grid.The access of photovoltaic grid-connected systems will have an impact on the voltage stability of the grid.To reduce this adverse effect,it is critical to study a reasonable control strategy,it will help improve the voltage quality of grid-connected grid.The main contents of this paper are the output prediction of PVGS and the control strategy to improve the voltage quality of grid-connected grid.This paper briefly introduces the general situation of PV grid-connected system and the principle of PQ control strategy.It designs and simulates the simulation model of PV grid-connected system on Matlab/Simulink platform.The PLL & Measurements module,PQ control module and Current control module in the PQ control strategy are introduced in detail.The simulation results show that PVGS with PQ control strategy can adjust the grid-connected power and adjust the grid-connected voltage,which verifies the accuracy of the model.Aiming at the problems that the low accuracy of the forecasting model for photovoltaic generation when the historical meteorological data are not enough or fluctuation of the weather is severe,a segmented forecasting model is proposed for the case that the sample is insufficient and weathers are rich.In this paper,the segmentation neural network model and similarity period screening method are combined.Using the actual data to train and verify the model,the results show that the comprehensive forecasting method can reflect the fluctuation of PV output and reduce the requirement of meteorological data when compared to the neural network model for whole points and the similarity period screening method by using the actual data to train and verify the model.In allusion to voltage fluctuation of point of common coupling caused by distribution network connection of PVGS.In this paper,the influencing factors are quantitatively calculated and analyzed by using the common conductor model,typical impedance value and 20 kVA PVGS accessing to the distribution network system.The requirements of active power output and the influence of voltage on the connection point are studied emphatically by four reactive power setting schemes,such as constant power factor,full utilization of residual reactive power,maximum voltage support and suppression of voltage fluctuation.Then,the reactive power setting scheme of PVGS under the different active power is presented,and the reactive power setting scheme with the maximum support of the point vol tage is proposed,and the reactive power setting scheme that allows the existence of certain voltage deviations and suppresses voltage fluctuations is proposed.Finally,through the typical weather conditions,the validity and feasibility of the proposed s cheme are verified in the normal grid condition and voltage sag condition.