Font Size: a A A

Study On Optical Power Prediction And Inverter Of Photovoltaic Power Station

Posted on:2020-05-13Degree:MasterType:Thesis
Country:ChinaCandidate:Y F LiFull Text:PDF
GTID:2392330572980645Subject:Control engineering
Abstract/Summary:PDF Full Text Request
With the development of photovoltaic power industry at home and abroad,the technology of optical power prediction and inverter has been developed rapidly?However,the accuracy of the current optical power prediction system is not ideal,and the conversion efficiency of inverter still has room for improvement?In this paper,two corresponding methods are proposed;(1)Apply BP neural network to optical power prediction;(2)change the fixed-step length conductance increment method of the maximum power point of the photovoltaic cell to variable step length conductance increment method,and further improve the increment method of the variable step length.This article mainly completed the following work:(1)using the MATLAB design to build a model of BP neural network,The network was trained with meteorological data from a photovoltaic power station,Finally,meteorological data of the other day of the photovoltaic power station is entered into the neural network,The predicted output power is compared with the photovoltaic power station's output power by using own optical power prediction system.(2)the simulation model and test model of photovoltaic cells were built using MATLAB design,And compared with the actual model of a photovoltaic cell.The model of the battery is the same as the actual battery,and the MPPT model is built,The ability to track the maximum power points of photovoltaic cells is compared with the fixed step length Conductance increment method and variable step length Conductance increment method.Then,the power of the maximum power point of the photovoltaic cell is tracked by the increment method of the variable step length and the improved step-length conductance increment method.The simulation results show that:The BP neural network model is better than the photovoltaic power station's own optical power prediction system in Prediction accuracy.It is proved that the variable step length conductance increment method is higher than the fixed step length conductance increment method for the maximum power point tracking capability of the battery.The improved variable step length conductance method has better tracking effect on the maximum power point of the battery than the variable step length conductance method,Which can produce more electricity.The method discussed in this paper is only to improve the original system from one aspect,and the improvement plan is not considered from all aspects,Only comprehensive consideration can meet the Higher and higher requirements of the accuracy and efficiency of inverter in the industry.
Keywords/Search Tags:photovoltaic cell, MPPT, conductance incremental method, BP neural network
PDF Full Text Request
Related items