Careless or ignorance?
Friday, March 30, 2007 by HongJunX
About several days ago,I settle down the basic direct current control simulation problems of three-phase PWM rectifier.And these days, I am trying using SVM modulation method.But first,I am focused on triangle modulation method.This time,the current control loop is based on d-q rotating coordinates ,with the angular speed same as that of the sinusoidal voltage source .Because if we want the unity input power factor,we set Iq reference value to zero.And if we get Id well controlled,and we get output voltage well controlled.We call Id as the active current component and Iq the reactive current component.In other means, we control these two current components,we are actually get the active and reactive power controlled.Maybe,this way is commonly used in SVG and UPFC,which is my final target.
The three-phase rectifier control model under d-q reference frame is far simple and easy to understand.There exists three control loop.In order to control output voltage,we build a outer voltage loop.The inner current loop are Id/Iq control loop.Outputs of these two current loop are Dd and Dq,which are switch duty-circle under d-q reference frame .In order to control the conduction of the IGBTs,we need to inverse transform the duty-circle under d-q reference frame to abc coordinates.In steady state,the abc duty-circle changes sinusoidally.Be careful that if the sequential signal which we used as carrier, repeats from 0 to 1,then make sure the abc duty-circle waves are in this range.If not,just use a proper gain and do some shift to make this condition satisfied.
After building up the simulation model,problems follows.The big problem is that I can not get Iq controlled.I changed the PI parameters again and again,and I failed again and again.Out of patient,I turn my help to my roommate also teammate,sir Xialixun.He helped me checking if there exists any errors.Without detecting any problems,he is also confused,so he copied his simulation model to me to let get a detail comparison.After check and comparison ,I finally found
the problem.It is the problem of the three-phase voltage sources.The three-balance voltage source I set the phase angles are 0 degree,-120 degree,120 degree,But his are 90 degree,-30 degree,210 degree.After about server minutes' thinking,I think the problem come from the coordinates.As plot showed left,the three-dimensional static abc coordinates represent the ABC phase's spacial position.Alpha-beta is two dimensional plane.But d-q is the rotating coordinate which here we assume at a speed same as voltage source.The abc-dq transformer I used in my model based on the initial angle starts from alpha axis.But actually when Va=Vm*sin(0),Vb=Vm*sin(-2*pi/3),Vc=Vm*sin(2*pi/3),the real start point is from beta axis!There exist 90 degree difference and no wonder Iq can not be controlled.I had to confess that this is not because of my careless,but my ignorance.I didn't think about this question before. After I fixing this problem,the model works.Although the control performance is not that good because of the controller parameters.Any way,I have made a little progress forward.Next step,I need to review my control theory textbook.I really need to occupy the skill of controller design method.
The three-phase rectifier control model under d-q reference frame is far simple and easy to understand.There exists three control loop.In order to control output voltage,we build a outer voltage loop.The inner current loop are Id/Iq control loop.Outputs of these two current loop are Dd and Dq,which are switch duty-circle under d-q reference frame .In order to control the conduction of the IGBTs,we need to inverse transform the duty-circle under d-q reference frame to abc coordinates.In steady state,the abc duty-circle changes sinusoidally.Be careful that if the sequential signal which we used as carrier, repeats from 0 to 1,then make sure the abc duty-circle waves are in this range.If not,just use a proper gain and do some shift to make this condition satisfied.
After building up the simulation model,problems follows.The big problem is that I can not get Iq controlled.I changed the PI parameters again and again,and I failed again and again.Out of patient,I turn my help to my roommate also teammate,sir Xialixun.He helped me checking if there exists any errors.Without detecting any problems,he is also confused,so he copied his simulation model to me to let get a detail comparison.After check and comparison ,I finally found
the problem.It is the problem of the three-phase voltage sources.The three-balance voltage source I set the phase angles are 0 degree,-120 degree,120 degree,But his are 90 degree,-30 degree,210 degree.After about server minutes' thinking,I think the problem come from the coordinates.As plot showed left,the three-dimensional static abc coordinates represent the ABC phase's spacial position.Alpha-beta is two dimensional plane.But d-q is the rotating coordinate which here we assume at a speed same as voltage source.The abc-dq transformer I used in my model based on the initial angle starts from alpha axis.But actually when Va=Vm*sin(0),Vb=Vm*sin(-2*pi/3),Vc=Vm*sin(2*pi/3),the real start point is from beta axis!There exist 90 degree difference and no wonder Iq can not be controlled.I had to confess that this is not because of my careless,but my ignorance.I didn't think about this question before. After I fixing this problem,the model works.Although the control performance is not that good because of the controller parameters.Any way,I have made a little progress forward.Next step,I need to review my control theory textbook.I really need to occupy the skill of controller design method.
|
|
0 Comments:
Post a Comment