CBT_block.py

import numpy as np
import scipy as sc
import scipy.optimize as sc_opt
import matplotlib.pyplot as plt
import itertools as itr

#w = np.logspace(-5, 3, 1000)
#s = 1j*w

def G(s):
    # the matrix transfer function
    G = [[0.001*np.exp(-5*s)*(-34.54*(s+0.0572))/((s+1.72E-04)*(4.32*s+1)),
    0.001*np.exp(-5*s)*(1.913)/((s+1.72E-04)*(4.32*s+1))],
    [0.001*np.exp(-5*s)*(-32.22*s)/((s+1.72E-04)*(4.32*s+1)),
    0.001*np.exp(-5*s)*(-9.188*(s+6.95E-04))/((s+1.72E-04)*(4.32*s+1))]]
    return G

#freqresp = map (G, s)

#sigmas = np.array([Sigma for U, Sigma, V in map(np.linalg.svd, freqresp)])

#maximum input and output vectors
#Umax = np.array([U[:, 0] for U, sigma , V in map(np.linalg.svd, freqresp)])
#Vmax = np.array([V[:, 0] for U, sigma , V in map(np.linalg.svd, freqresp)])

#minimum input and output vectors
#Umin = np.array([U[:, -1] for U, sigma , V in map(np.linalg.svd, freqresp)])
#Vmin = np.array([V[:, -1] for U, sigma , V in map(np.linalg.svd, freqresp)])

def Sing(w_i):
    s = 1j*w_i
    [U, S, V] = np.linalg.svd(G(s))
    return np.max(S)-1

sc_opt.fsolve(Sing, 0.1)



#plt.loglog(w, sigmas)
#plt.show()