Font Size: a A A

Research Of Grid-connected Control And Islanding Detection For Photovoltaic System

Posted on:2011-02-11Degree:MasterType:Thesis
Country:ChinaCandidate:E L ZhangFull Text:PDF
GTID:2192330338983579Subject:Electrical theory and new technology
Abstract/Summary:PDF Full Text Request
With the increasing of the world's power demand and worsening of the environment, the Photovoltaic (PV) application become a hot spot.According to the actual situation of power supply, Grid-Connected PV System will be the new trend among PV applications.This paper focuses on the research of grid-connected control and islanding detection for PV System. The main contents are as follows:First, in the Grid-Connected PV System, the filter and the control strategy directly affect on the quality of current. Based on the analysis of the mathematical model of filter, the paper designs the LCL filter, and the simulation results verify its feasibility. This paper mainly analyses grid-connected control of three-phase PV system, and designs the synchronous PI control method base on the dq transformation. The simulation results with MATLAB/Simulink prove that the method of separate control of the active and reactive current can satisfy grid-connected requirements.Secondly, through the analysis of traditional neural network algorithm, PI control method based on the neural network algorithm is proposed, then a new PV grid-connected control method is formatted. In addition, in order to improve the speed of this complex neural network control, this paper also proposes a new embedded neural network PI control method. Simulation results show that the two methods can achieve grid-connected control, while the latter has better performance.Thirdly, in the PV system, the islanding detection method consists of two types, which include active and passive detection methods. Based on the analysis of existing islanding detection methods, according to the character of the active detection methods, the power-matching islanding detection method based on slight asymmetric frequency is brought forward. Theoretical analysis and simulation results show that this method can detect power-matching islanding, and affect little on the power quality, and provide the basis for intentional islanding operation and uninterrupted power supply technology.
Keywords/Search Tags:Grid-connected PV System, LCL Filter, PI Control, Neural Network, Embedded, Islanding Detection, Slight Asymmetric Frequency
PDF Full Text Request
Related items