Font Size: a A A

Research On Active Power Filter Based On Neural Network Control

Posted on:2012-04-05Degree:MasterType:Thesis
Country:ChinaCandidate:L S HanFull Text:PDF
GTID:2178330332991532Subject:Power electronics and electric drive
Abstract/Summary:PDF Full Text Request
With the using of power electronic equipments in power system,the harmonic pollution of power system is becoming more and more serious.Active power filter (APF) is regardedas an effective way to correct power harmonic,which is the focus in the field of power quality control. In the last few years, much attention had paid to researchAPF by applying modern control theory all over the world.A new way was opened up with the using of neural network,which has the strong ability of nonlinear mapping.The paper designs the APF controller by applying neural network and PI control theory based on the previous research.Firstly,the present research situation,characteristics and existent questions were analyzed,the merit and demerit of the methods applied to harmonic detection and current control were summarized and compared.In the end,the advanced neural network control theory was selected to detect harmonic in the APF design.Secondly,the APF theory is analyzed deeply,the math model of shunt APF is established.Besides,multi-adaptive linear neural network was adopted to detect harmonic,and the PI method was combined to control the direct voltage effectively,and the method of triangular wave comparator control was adopted to track and control the current.Thirdly,the shunt three-phase four-wire APF was designed,circuit model was established.Besides,the hardware and software also were designed detailedly.Inder to improve real-time of equipment,the AD synchronism sampling circuit was designed.Phase locked loop was designed to improve compensation effect.At the same time, a output filter was designed to eliminate the ultra harmonics effectively.Finally,simulation models of three-phase four-wire shunt avtive power filter system based on this method are built in environment of MATLAB software,in the meantime,the system experiments of harmonic detection and conpensation are made,the control system of which use DSP as a core CPU.The results of simulation and experiments show that APF can eliminate harmonic current effectively, in the meantime,the APF system have better compensation effect.
Keywords/Search Tags:Active power filter(APF), Harmonic currents, Neural network, DSP, Synchronism sampling
PDF Full Text Request
Related items