Font Size: a A A

Research On Dynamic Modeling Method Of Switching Converter Converter Based On System Identification

Posted on:2013-06-14Degree:MasterType:Thesis
Country:ChinaCandidate:J YangFull Text:PDF
GTID:2232330395980553Subject:Circuits and Systems
Abstract/Summary:PDF Full Text Request
The design of switching supply system depends on the mathematical model for every linkof the system, while the dynamic modeling for the switching converter is the key to design aswitching converter. But, it is difficult to obtain the accurate small signal modeling by existingmethods especially for the switch resonant converter with complex circuit structure and manyoperating modes, such as FPRDC and LLC, which are used in MPM power supply. So, it isnecessary to explore an accurate and effective small signal modeling method for switchconverters.The thesis is primarily focuses on dynamic modeling method for the switching converter,especially for the complex switching converters. Considering that the small signal modeling isdifficult to be built for these converters, a new modeling method is proposed based on the systemidentification. The method treats switching converter as a black-box, only with the input andoutput data of switch power converter comes from MATLAB/Simulink or Pspice simulation byusing the least-squares or Recursive-least-squares parameter estimation algorithm of systemidentification to fit the step response curve for modeling. It can avoid the analysis of the complexworking model. So it solves the difficulty in building the small signal model of switchingconverter. Moreover, it is suitable for PWM converters as well as PFM converters because thismethod only depends on the input and output data.Firstly, the paper analyses various parts of switching supply system and their mathematicalmodels. Focusing on the PWM switching converter’s characteristics, a new modeling methodbased on Least-squares parameter estimation is proposed from the perspective of fitting theoutput-voltage curve. A detail algorithm derivation of this method is given in this paper.Secondly, the small signal modeling of Buck, Boost, Buck-Boost and Cuk converters is built bymaking use of this new method. Compared with these models built by state-space averagingmethod, this new method is found much more precise than state-space averaging method. Also,the small signal modeling of APCDC converter is built by this method. Thirdly, the paperanalyses the PFM switching converter’s mathematical models. On the basis of Least-squaresparameter estimation method, an online small signal method called Recursive-least-squaresparameter estimation is deduced. Then, the small signal modeling of LLC and FPRDCconverters is built by making use of this method. The successful applications show that the newmethod based on system identification applies to not only PWM switch converters but also PFMswitch converters. Finally, a Graphical User Interfaces of system identification software isprogrammed, which can save time labor and be reusable for power supply designers.
Keywords/Search Tags:small signal modeling, switching converter, system identification, parameterestimation, curve fitting, GUI
PDF Full Text Request
Related items