Prev: why first order filter have slope 6db/octave
Next: How to implement the filter give in Z transform using matlab
From: AmiHK on 28 Jan 2010 07:58
Hi All,

I have been recently placed on a new project related to Power Systems.I am
new in this field
and not getting a proper place to start with.So please give me your
advise.
The project is an Active Shunt Type Power Filter used to take out
harmonics from the power line.
We have an existing product selling in the market,but with an old
controller design, TI DSP(TMS320C40)<=====>Dual Ram<=====>TMS320F240(TI
MCU).
So we have thought of migrating on a new controller platform.

The existing source code we have is in assembly,and the documentation
written is not up to the mark(I mean the derivations..etc)
All that i can see in the docs are final equations with missing
explanation.Below is the section I am not able to understand,
mainly the equation (1),(2).....(5) their derivations.
---------------start here ------------------
Recursive real-time harmonics and reactive power detection algorithm
routine flow chart shown in Figure 4-20.
F240 controlled by the A / D converter samples the input data to the form
of signed integers stored in the designated memory unit. C32 first converts
integer to floating-point number, and then executes the recursive real-time
detection of harmonic and reactive power compensation algorithm
instructions, the calculation is as described below
According to symmetrical component method, in three-phase three-wire
system, the instantaneous load current ia, ib, ic can be decomposed into
positive sequence component group and the negative sequence components
Group. To use the index, 1 positive sequence, 2 negative sequence,k
harmonic frequency,I Indicated that peak current, ¢ first phase angle, N
sampling frequency cycle points,n Counts for the sampling points
n=0,1,2,…N-1. So, after sampling the load current can be expressed as


ix(n) = ∑k=1to ∞ [I1k sin(2∏nk/N + ￠1k -2l∏/3) + I2k
sin(2∏nk/N + ￠2k +2l∏/3)]

In which

l = 0,1,2 for x=a, x=b, x=c


Substituting in the following formula

ixypk(n) = {ixypk(n-1) – 2*[ ix(n-N)- ix(n)] * sin(2∏nk/N -
2l∏/3)}/N
ixyqk(n) = {ixyqk(n-1) – 2*[ ix(n-N)- ix(n)] * cos(2∏nk/N -
2l∏/3)}/N

-------------------------(1)

For the fundamental wave can be calculated
Matrix [A11(n) B11(n) , A21(n) B21(n)] = 1/3 matrix [iayp1(n)+ ibyp1(n)+
icyp1(n) iayq1(n)+
ibyq1(n)+ icyq1(n) , 2iayp1(n)- ibyp1(n)-
icyp1(n) 2iayq1(n)- ibyq1(n)- icyq1(n)]
------------------(2)
B11 A11,Represent the fundamental positive-sequence reactive current
components and fundamental positive-sequence active component of the
amplitude.
Suppose



ix11(n) = A11(n)* sin(2∏n/N + 2l∏/3) + B11(n)* cos(2∏n/N +
2l∏/3) --------(3)
ix21(n) = A21(n)* sin(2∏n/N + 2l∏/3) + B21(n)* cos(2∏n/N +
2l∏/3) --------(4)
ix1(n) =ixyp1(n)* sin(2∏n/N + 2l∏/3) + ixyq1(n)* cos(2∏n/N +
2l∏/3) --------(5)



If only the active power filter to compensate harmonic and fundamental
negative sequence, instead of reactive power compensation, then use
equation (3). Calculated to be fundamental positive sequence current
signals ix11, Then the load current signal minus the ix11 command signal
can be obtained icx*(n ) = ix(n) - ix11.If the demand compensation after
the network side of a symmetrical three-phase fundamental current, with
power factor 1,with equation (3) makes A11 = 0, Which can be fundamental
positive sequence of the active component ipx11, With the load current
signal minus ipx11,will get the command signal icx*(n ) = ix(n) - ipx11
If the active power filter to compensate harmonics and not just reactive
and negative sequence compensation,use equation (5), Calculated fundamental
current signal ix1. Then the load current signal minus the ix1,will get the
command signal.
For the harmonic,don't need to distinguish positive sequence component &
negative sequence component. If the active power filter is used to
compensate for nominated sub-harmonic,then by using equation (1),with
ixyqk(n), ixypk(n) be calculated as

ixk(n) =ixyqk(n)* cos(2∏kn/N ) + ixypk(n)* sin(2∏kn/N )
And is a sub harmonic current signal.
------------- end here -----------------------------------

Thanks in advance




 | 
Pages: 1
Prev: why first order filter have slope 6db/octave
Next: How to implement the filter give in Z transform using matlab