lowPassReturn low pass filter roots as needed for block for given cut-off frequency |
This information is part of the Modelica Standard Library maintained by the Modelica Association.
The goal is to implement the filter in the following form:
// real pole: der(x) = r*x - r*u y = x // complex conjugate poles: der(x1) = a*x1 - b*x2 + ku*u; der(x2) = b*x1 + a*x2; y = x2; ku = (a^2 + b^2)/b
This representation has the following transfer function:
// real pole: s*y = r*y - r*u or (s-r)*y = -r*u or y = -r/(s-r)*u comparing coefficients with y = cr/(s + cr)*u -> r = -cr // r is the real eigenvalue // complex conjugate poles s*x2 = a*x2 + b*x1 s*x1 = -b*x2 + a*x1 + ku*u or (s-a)*x2 = b*x1 -> x2 = b/(s-a)*x1 (s + b^2/(s-a) - a)*x1 = ku*u -> (s(s-a) + b^2 - a*(s-a))*x1 = ku*(s-a)*u -> (s^2 - 2*a*s + a^2 + b^2)*x1 = ku*(s-a)*u or x1 = ku*(s-a)/(s^2 - 2*a*s + a^2 + b^2)*u x2 = b/(s-a)*ku*(s-a)/(s^2 - 2*a*s + a^2 + b^2)*u = b*ku/(s^2 - 2*a*s + a^2 + b^2)*u y = x2 comparing coefficients with y = c0/(s^2 + c1*s + c0)*u -> a = -c1/2 b = sqrt(c0 - a^2) ku = c0/b = (a^2 + b^2)/b comparing with eigenvalue representation: (s - (a+jb))*(s - (a-jb)) = s^2 -2*a*s + a^2 + b^2 shows that: a: real part of eigenvalue b: imaginary part of eigenvalue time -> infinity: y(s=0) = x2(s=0) = 1 x1(s=0) = -ku*a/(a^2 + b^2)*u = -(a/b)*u
cr_in |
Type: Real[:] Description: Coefficients of real poles of base filter |
---|---|
c0_in |
Type: Real[:] Description: Coefficients of s^0 term of base filter if conjugate complex pole |
c1_in |
Type: Real[size(c0_in, 1)] Description: Coefficients of s^1 term of base filter if conjugate complex pole |
f_cut |
Type: Frequency (Hz) Description: Cut-off frequency |
r |
Type: Real[size(cr_in, 1)] Description: Real eigenvalues |
---|---|
a |
Type: Real[size(c0_in, 1)] Description: Real parts of complex conjugate eigenvalues |
b |
Type: Real[size(c0_in, 1)] Description: Imaginary parts of complex conjugate eigenvalues |
ku |
Type: Real[size(c0_in, 1)] Description: Input gain |