def test_step(self):
"""Test function ``step``."""
figure(); plot_shape = (1, 3)
#Test SISO system
A, B, C, D = self.make_SISO_mats()
sys = ss(A, B, C, D)
#print(sys)
#print("gain:", dcgain(sys))
subplot2grid(plot_shape, (0, 0))
t, y = step(sys)
plot(t, y)
subplot2grid(plot_shape, (0, 1))
T = linspace(0, 2, 100)
X0 = array([1, 1])
t, y = step(sys, T, X0)
plot(t, y)
#Test MIMO system
A, B, C, D = self.make_MIMO_mats()
sys = ss(A, B, C, D)
subplot2grid(plot_shape, (0, 2))
t, y = step(sys)
plot(t, y)
def testStep(self, siso):
"""Test step()"""
t = np.linspace(0, 1, 10)
# Test transfer function
yout, tout = step(siso.tf1, T=t)
youttrue = np.array([0, 0.0057, 0.0213, 0.0446, 0.0739,
0.1075, 0.1443, 0.1832, 0.2235, 0.2642])
np.testing.assert_array_almost_equal(yout, youttrue, decimal=4)
np.testing.assert_array_almost_equal(tout, t)
# Test SISO system with direct feedthrough
sys = siso.ss1
youttrue = np.array([9., 17.6457, 24.7072, 30.4855, 35.2234, 39.1165,
42.3227, 44.9694, 47.1599, 48.9776])
yout, tout = step(sys, T=t)
np.testing.assert_array_almost_equal(yout, youttrue, decimal=4)
np.testing.assert_array_almost_equal(tout, t)
# Play with arguments
yout, tout = step(sys, T=t, X0=0)
np.testing.assert_array_almost_equal(yout, youttrue, decimal=4)
np.testing.assert_array_almost_equal(tout, t)
X0 = np.array([0, 0])
yout, tout = step(sys, T=t, X0=X0)
np.testing.assert_array_almost_equal(yout, youttrue, decimal=4)
np.testing.assert_array_almost_equal(tout, t)
yout, tout, xout = step(sys, T=t, X0=0, return_x=True)
np.testing.assert_array_almost_equal(yout, youttrue, decimal=4)
np.testing.assert_array_almost_equal(tout, t)
def test_step(self):
"""Test function ``step``."""
figure()
plot_shape = (1, 3)
#Test SISO system
A, B, C, D = self.make_SISO_mats()
sys = ss(A, B, C, D)
#print(sys)
#print("gain:", dcgain(sys))
subplot2grid(plot_shape, (0, 0))
t, y = step(sys)
plot(t, y)
subplot2grid(plot_shape, (0, 1))
T = linspace(0, 2, 100)
X0 = array([1, 1])
t, y = step(sys, T, X0)
plot(t, y)
#Test MIMO system
A, B, C, D = self.make_MIMO_mats()
sys = ss(A, B, C, D)
subplot2grid(plot_shape, (0, 2))
t, y = step(sys)
plot(t, y)
def test_step(self, SISO_mats, MIMO_mats, mplcleanup):
"""Test function ``step``."""
figure()
plot_shape = (1, 3)
#Test SISO system
A, B, C, D = SISO_mats
sys = ss(A, B, C, D)
#print(sys)
#print("gain:", dcgain(sys))
subplot2grid(plot_shape, (0, 0))
y, t = step(sys)
plot(t, y)
subplot2grid(plot_shape, (0, 1))
T = linspace(0, 2, 100)
X0 = array([1, 1])
y, t = step(sys, T, X0)
plot(t, y)
# Test output of state vector
y, t, x = step(sys, return_x=True)
#Test MIMO system
A, B, C, D = MIMO_mats
sys = ss(A, B, C, D)
subplot2grid(plot_shape, (0, 2))
y, t = step(sys)
plot(t, y[:, 0, 0])
def process_data(num11, den11, num21, den21):
w11 = ctrl.tf(num11, den11)
w21 = ctrl.tf(num21, den21)
print('результат w11={} w21={}'.format(w11, w21))
TimeLine = []
for i in range(1, 3000):
TimeLine.append(i / 1000)
plt.figure(0, figsize=[7, 6])
[y11, x11] = ctrl.step(w11, TimeLine)
[y21, x21] = ctrl.step(w21, TimeLine)
plt.plot(x11, y11, "r", label='Исходная')
plt.plot(x21, y21, "b", label='Увеличенная k и уменшенная Т')
plt.title('Переходная функция звена')
plt.ylabel('Амплитуда')
plt.xlabel('Время(с)')
plt.grid(True)
plt.show()
[y11, x11] = ctrl.impulse(w11, TimeLine)
[y21, x21] = ctrl.impulse(w21, TimeLine)
plt.plot(x11, y11, "r", label='Исходная')
plt.plot(x21, y21, "b", label='Увеличенная k и уменшенная Т')
plt.title('Импульсная функция звена')
plt.ylabel('Амплитуда')
plt.xlabel('Время(с)')
plt.grid(True)
plt.show()
ctrl.mag, ctrl.phase, ctrl.omega = ctrl.bode(w11, w21, dB=False)
plt.plot()
plt.show()
return w11, w21
def testStep_mimo(self, mimo):
"""Test step for MIMO system"""
sys = mimo.ss1
t = np.linspace(0, 1, 10)
youttrue = np.array([9., 17.6457, 24.7072, 30.4855, 35.2234, 39.1165,
42.3227, 44.9694, 47.1599, 48.9776])
y_00, _t = step(sys, T=t, input=0, output=0)
y_11, _t = step(sys, T=t, input=1, output=1)
np.testing.assert_array_almost_equal(y_00, youttrue, decimal=4)
np.testing.assert_array_almost_equal(y_11, youttrue, decimal=4)
def BAGUVIX(GG1, name_gg1, GG2, name_gg2,
t): # функция для построения графиков характеристик
topic = {
'G1': 'безынерционного звена',
'G2': 'апериодического звена',
'G3': 'интегрирующего звена',
'G4': 'реального диф. звена',
'G5': 'идеального диф. звена',
} # словарь для графика
if name_gg1 in topic: # определяем какой именно строим, для графика
k1 = topic[name_gg1]
plt.figure(1) # Вывод графиков в отдельном окне
y1, t1 = con.step(GG1, t)
y2, t2 = con.step(GG2, t)
lines = [y1, y2]
# plt.subplot(1, 1, 1) # 1цифра - количество строк в графике, 2 -тьиьтиколичество графиков в строке, 3 -номер графика
lines[0], lines[1] = plt.plot(t, y1, "r", t, y2, "b")
plt.legend(lines, ['h(t) для 1', 'h(t) для 2'],
loc='best',
ncol=2,
fontsize=10)
plt.title('Переходная характеристика' + '\n для ' + k1, fontsize=10)
plt.ylabel('h')
plt.xlabel('t, c')
plt.grid()
plt.figure(2)
y2, t2 = con.impulse(GG2, t)
y1, t1 = con.impulse(GG1, t)
lines[0], lines[1] = plt.plot(t, y1, "r", t, y2, "b")
plt.legend(lines, ['w(t) для 1', 'w(t) для 2'],
loc='best',
ncol=2,
fontsize=10)
plt.title('Импульсная характеристика' + '\n для ' + k1, fontsize=10)
plt.ylabel('w')
plt.xlabel('t, c')
plt.grid()
plt.figure(3)
mag1, phase1, omega1 = con.bode(GG1, dB=False)
# plt.plot()
plt.title('Частотные характеристики' + "\n для " + k1,
fontsize=10,
y=2.2)
mag1, phase1, omega1 = con.bode(GG2, dB=False)
plt.plot()
plt.title('Частотные характеристики' + "\n для " + k1,
fontsize=10,
y=2.2)
plt.show()
def stepResponse(Ts):
num = Poly(Ts.as_numer_denom()[0],s).all_coeffs()
den = Poly(Ts.as_numer_denom()[1],s).all_coeffs()
tf = matlab.tf(map(float,num),map(float,den))
y,t = matlab.step(tf)
plt.plot(t,y)
plt.title("Step Response")
plt.grid()
plt.xlabel("time (s)")
plt.ylabel("y(t)")
info = "OS:%f%s"%(round((y.max()/y[-1]-1)*100,2),'%')
try:
i10 = next(i for i in range(0,len(y)-1) if y[i]>=y[-1]*.10)
Tr = round(t[next(i for i in range(i10,len(y)-1) if y[i]>=y[-1]*.90)]-t[i10],2)
except StopIteration:
Tr = "unknown"
try:
Ts = round(t[next(len(y)-i for i in range(2,len(y)-1) if abs(y[-i]/y[-1])>1.02)]-t[0],2)
except StopIteration:
Ts = "unknown"
info += "\nTr: %s"%(Tr)
info +="\nTs: %s"%(Ts)
print info
plt.legend([info],loc=4)
plt.show()
def second_var():
"""
Второй способ релизации метода Зиглера-Никольса
:return: время западнывания tet, постоянную времени T и коэффициент передачи k
"""
# получает передаточную функцию объекта управления
w = SchemeBody().get_scheme_solving()
t = np.linspace(0, stop=100, num=2000)
y1, t1 = step(w, t)
y1 = list(y1)
max_dif = 0
# номер элемента в списке, имеющего максимальную разницу с предыдущим
nmd = 0
for num, yy in enumerate(y1[1:]):
if (yy - y1[num]) > max_dif:
max_dif, nmd = (yy - y1[num]), (num + 1)
if y1[num - 1] < y1[num] > y1[
num + 1]: # дошли до первого максимума, отбой
break
# находим время запаздывания и постоянную времени через уравенение прямой
tet = -y1[nmd] / (y1[nmd] - y1[nmd - 1]) * (t[nmd] -
t[nmd - 1]) + t[nmd]
T = (y1[-1] - y1[nmd]) / (y1[nmd] -
y1[nmd - 1]) * (t[nmd] - t[nmd - 1]) + t[nmd]
return tet, T, y1[num]
def plot_Res(t, y, g):
#%matplotlib qt
Y, T = mt.step(g, t)
plt.plot(T, Y)
plt.plot(t, y)
plt.grid()
plt.show()
def create_Plot(var,name,num, den):
W=ml.tf(num,den)
if var!=5:
#Переходная функция
plt.figure().canvas.set_window_title(name)
y,x=ml.step(W,timeVector)
plt.plot(x,y,"b")
plt.title('Переходная функция')
plt.ylabel('Амплитуда, о.е.')
plt.xlabel('Время, с.')
plt.grid(True)
#TimeLine=[]
#for i in range(0;3000):
# TimeLine = [i/1000]
plt.show()
#Импульсная функция
plt.figure().canvas.set_window_title(name)
y,x=ml.impulse(W, timeVector)
plt.plot(x,y,"r")
plt.title('Импульсная функция')
plt.ylabel('Амплитуда, о.е.')
plt.xlabel('Время, с.')
plt.grid(True)
plt.show()
#Диаграмма Боде
plt.figure().canvas.set_window_title(name)
mag, phase, omega = ml.bode(W, dB=False)
plt.plot()
plt.xlabel('Частота, Гц')
plt.show()
return
def pidplot(num, den, Kp, Ki, Kd, desired_settle=1.):
'''
Plot system step response when open loop system is subjected to feedback PID compensation.
Also plots 2% settling lines, and a vertical line at a desired settling time.
y, t =pidplot(num,den,Kp,Ki,Kd,desired_settle=1.)
Parameters:
:param num: [array-like], coefficients of numerator of open loop transfer function
:param den: [array-like], coefficients of denominator of open loop transfer function
:param Kp: [float] Proportional gain
:param Ki: [float] Integral gain
:param Kd: [float] Derivative gain
:param desired_settle: [float] Desired settling time, for tuning PID controllers, default = 1.0
Returns:
:return: y: [array-like] time step response
:return: t: [array-like] time vector
'''
numc = [Kd, Kp, Ki]
denc = [0, 1, 0]
numcg = convolve(numc, num)
dencg = convolve(denc, den)
Gfb = feedback(tf(numcg, dencg), 1)
y, t = step(Gfb)
yss = dcgain(Gfb)
plot(t, y, 'r')
plot(t, 1.02 * yss * ones(len(t)), 'k--')
plot(t, 0.98 * yss * ones(len(t)), 'k--')
plot(desired_settle * ones(15), linspace(0, yss + 0.25, 15), 'b-.')
xlim(0)
xlabel('Time [s]')
ylabel('Magnitude')
grid()
show()
return y, t
def test_dcgain_2(self):
"""Test function dcgain with different systems"""
#Create different forms of a SISO system
A, B, C, D = self.make_SISO_mats()
num, den = scipy.signal.ss2tf(A, B, C, D)
# numerator is only a constant here; pick it out to avoid numpy warning
Z, P, k = scipy.signal.tf2zpk(num[0][-1], den)
sys_ss = ss(A, B, C, D)
#Compute the gain with ``dcgain``
gain_abcd = dcgain(A, B, C, D)
gain_zpk = dcgain(Z, P, k)
gain_numden = dcgain(np.squeeze(num), den)
gain_sys_ss = dcgain(sys_ss)
# print('gain_abcd:', gain_abcd, 'gain_zpk:', gain_zpk)
# print('gain_numden:', gain_numden, 'gain_sys_ss:', gain_sys_ss)
#Compute the gain with a long simulation
t = linspace(0, 1000, 1000)
y, _t = step(sys_ss, t)
gain_sim = y[-1]
# print('gain_sim:', gain_sim)
#All gain values must be approximately equal to the known gain
assert_array_almost_equal(
[gain_abcd, gain_zpk, gain_numden, gain_sys_ss, gain_sim],
[0.026948, 0.026948, 0.026948, 0.026948, 0.026948],
decimal=6)
def step_sys(sys, final_time, setpoint):
# open-loop system transfer function
try:
num, den = model(sys)
except:
# for error detection
print("Err: system in not defined")
return
Gs = control.tf(num, den)
#print(Gs)
# closed-loop unity-feedback transfer function
Ts = control.feedback(Gs, 1)
# simulation time parameters
initial_time = 0
nsteps = 40 * final_time # number of time steps
t = np.linspace(initial_time, final_time, round(nsteps))
output, t = matlab.step(Ts, t)
output = setpoint*output
# covert numpy arrays to lists
t = list(t)
output = list(output)
# round lists to 6 decimal digits
ndigits = 6
t = [round(num, ndigits) for num in t]
output = [round(num, ndigits) for num in output]
return t, output
def assert_systems_behave_equal(self, sys1, sys2):
'''
Test if the behavior of two LTI systems is equal. Raises ``AssertionError``
if the systems are not equal.
Works only for SISO systems.
Currently computes dcgain, and computes step response.
'''
#gain of both systems must be the same
assert_array_almost_equal(dcgain(sys1), dcgain(sys2))
#Results of ``step`` simulation must be the same too
y1, t1 = step(sys1)
y2, t2 = step(sys2, t1)
assert_array_almost_equal(y1, y2)
def analysis(self, w, block, which):
nume = [int(n) for n in block.nume_coef]
if block.deno_coef == []:
deno = [1]
else:
deno = [int(d) for d in block.deno_coef]
nume.reverse()
deno.reverse()
print(nume)
print(deno)
system = matlab.tf(nume, deno)
if which == 'bode':
matlab.bode(system)
plt.show()
elif which == 'rlocus':
matlab.rlocus(system)
plt.show()
elif which == 'nyquist':
matlab.nyquist(sys)
plt.show()
elif which == 'impulse':
t = np.linspace(0, 3, 1000)
yout, T = matlab.impulse(system, t)
plt.plot(T, yout)
plt.axhline(0, color="b", linestyle="--")
plt.xlim(0, 3)
plt.show()
elif which == 'step':
t = np.linspace(0, 3, 1000)
yout, T = matlab.step(system, t)
plt.plot(T, yout)
plt.axhline(1, color="b", linestyle="--")
plt.xlim(0, 3)
plt.show()
def assert_systems_behave_equal(self, sys1, sys2):
'''
Test if the behavior of two Lti systems is equal. Raises ``AssertionError``
if the systems are not equal.
Works only for SISO systems.
Currently computes dcgain, and computes step response.
'''
#gain of both systems must be the same
assert_array_almost_equal(dcgain(sys1), dcgain(sys2))
#Results of ``step`` simulation must be the same too
t, y1 = step(sys1)
_t, y2 = step(sys2, t)
assert_array_almost_equal(y1, y2)
def test_dcgain_2(self):
"""Test function dcgain with different systems"""
#Create different forms of a SISO system
A, B, C, D = self.make_SISO_mats()
Z, P, k = scipy.signal.ss2zpk(A, B, C, D)
num, den = scipy.signal.ss2tf(A, B, C, D)
sys_ss = ss(A, B, C, D)
#Compute the gain with ``dcgain``
gain_abcd = dcgain(A, B, C, D)
gain_zpk = dcgain(Z, P, k)
gain_numden = dcgain(np.squeeze(num), den)
gain_sys_ss = dcgain(sys_ss)
print('gain_abcd:', gain_abcd, 'gain_zpk:', gain_zpk)
print('gain_numden:', gain_numden, 'gain_sys_ss:', gain_sys_ss)
#Compute the gain with a long simulation
t = linspace(0, 1000, 1000)
y, _t = step(sys_ss, t)
gain_sim = y[-1]
print('gain_sim:', gain_sim)
#All gain values must be approximately equal to the known gain
assert_array_almost_equal([gain_abcd[0,0], gain_zpk[0,0],
gain_numden[0,0], gain_sys_ss[0,0], gain_sim],
[0.026948, 0.026948, 0.026948, 0.026948,
0.026948],
decimal=6)
#Test with MIMO system
A, B, C, D = self.make_MIMO_mats()
gain_mimo = dcgain(A, B, C, D)
print('gain_mimo: \n', gain_mimo)
assert_array_almost_equal(gain_mimo, [[0.026948, 0 ],
[0, 0.026948]], decimal=6)
def test_dcgain_2():
"""Test function dcgain with different systems"""
#Create different forms of a SISO system
A, B, C, D = make_SISO_mats()
Z, P, k = scipy.signal.ss2zpk(A, B, C, D)
num, den = scipy.signal.ss2tf(A, B, C, D)
sys_ss = ss(A, B, C, D)
#Compute the gain with ``dcgain``
gain_abcd = dcgain(A, B, C, D)
gain_zpk = dcgain(Z, P, k)
gain_numden = dcgain(np.squeeze(num), den)
gain_sys_ss = dcgain(sys_ss)
print 'gain_abcd:', gain_abcd, 'gain_zpk:', gain_zpk
print 'gain_numden:', gain_numden, 'gain_sys_ss:', gain_sys_ss
#Compute the gain with a long simulation
t = linspace(0, 1000, 1000)
_t, y = step(sys_ss, t)
gain_sim = y[-1]
print 'gain_sim:', gain_sim
#All gain values must be approximately equal to the known gain
assert_array_almost_equal([
gain_abcd[0, 0], gain_zpk[0, 0], gain_numden[0, 0], gain_sys_ss[0, 0],
gain_sim
], [0.026948, 0.026948, 0.026948, 0.026948, 0.026948],
decimal=6)
#Test with MIMO system
A, B, C, D = make_MIMO_mats()
gain_mimo = dcgain(A, B, C, D)
print 'gain_mimo: \n', gain_mimo
assert_array_almost_equal(gain_mimo, [[0.026948, 0], [0, 0.026948]],
decimal=6)
def test_dcgain_2(self):
"""Test function dcgain with different systems"""
#Create different forms of a SISO system
A, B, C, D = self.make_SISO_mats()
num, den = scipy.signal.ss2tf(A, B, C, D)
# numerator is only a constant here; pick it out to avoid numpy warning
Z, P, k = scipy.signal.tf2zpk(num[0][-1], den)
sys_ss = ss(A, B, C, D)
#Compute the gain with ``dcgain``
gain_abcd = dcgain(A, B, C, D)
gain_zpk = dcgain(Z, P, k)
gain_numden = dcgain(np.squeeze(num), den)
gain_sys_ss = dcgain(sys_ss)
# print('gain_abcd:', gain_abcd, 'gain_zpk:', gain_zpk)
# print('gain_numden:', gain_numden, 'gain_sys_ss:', gain_sys_ss)
#Compute the gain with a long simulation
t = linspace(0, 1000, 1000)
y, _t = step(sys_ss, t)
gain_sim = y[-1]
# print('gain_sim:', gain_sim)
#All gain values must be approximately equal to the known gain
assert_array_almost_equal([gain_abcd[0,0], gain_zpk[0,0],
gain_numden[0,0], gain_sys_ss[0,0], gain_sim],
[0.026948, 0.026948, 0.026948, 0.026948,
0.026948],
decimal=6)
def testDcgain(self, siso):
"""Test dcgain() for SISO system"""
# Create different forms of a SISO system using scipy.signal
A, B, C, D = siso.ss1.A, siso.ss1.B, siso.ss1.C, siso.ss1.D
Z, P, k = sp.signal.ss2zpk(A, B, C, D)
num, den = sp.signal.ss2tf(A, B, C, D)
sys_ss = siso.ss1
# Compute the gain with ``dcgain``
gain_abcd = dcgain(A, B, C, D)
gain_zpk = dcgain(Z, P, k)
gain_numden = dcgain(np.squeeze(num), den)
gain_sys_ss = dcgain(sys_ss)
# print('\ngain_abcd:', gain_abcd, 'gain_zpk:', gain_zpk)
# print('gain_numden:', gain_numden, 'gain_sys_ss:', gain_sys_ss)
# Compute the gain with a long simulation
t = linspace(0, 1000, 1000)
y, _t = step(sys_ss, t)
gain_sim = y[-1]
# print('gain_sim:', gain_sim)
# All gain values must be approximately equal to the known gain
np.testing.assert_array_almost_equal(
[gain_abcd, gain_zpk, gain_numden, gain_sys_ss, gain_sim],
[59, 59, 59, 59, 59])
def plot_trans_func(w, toFindT=False, toPlotTrans=False):
"""
функция находит и возвращает период переходной характеристики
"""
t = np.linspace(0, stop=100, num=2000)
y1, t1 = step(w, t)
t = list(t)
if toPlotTrans:
plt.plot(t, y1, "r")
plt.title('Step Response')
plt.ylabel('Amplitude h(t)')
plt.xlabel('Time(sec)')
plt.grid(True)
plt.show()
if toFindT:
two_maxes = []
for num in range(len(y1[1:-1])):
if y1[num - 1] < y1[num] > y1[num + 1]:
two_maxes.append(num)
if len(two_maxes) == 2: break
return t[two_maxes[1]] - t[two_maxes[0]]
def grafic_stepw(f, name):
y1, x1 = cm.step(f, TimeLine)
plt.plot(x1, y1, "b")
plt.title('Переходная функция {}'.format(name))
plt.ylabel('Амплитудное значение')
plt.xlabel('Время(c)')
plt.grid(True, linewidth=1, alpha=0.5)
plt.show()
def degrau(G):
y1,t1 = co.step(co.feedback(G,1))
plt.figure()
plt.plot(t1,y1)
plt.title("Step")
plt.xlabel("Tempo[s]")
plt.ylabel("Amplitude")
plt.grid()
print( co.stepinfo(co.feedback(G,1)))
def residuals(p, y, t):
[k,alpha] = p
print(k, type(k))
alpha = p[1]
print(alpha, type(alpha))
g = tf(k,[1,alpha,0])
Y,T = mt.step(g,t)
err=y-Y
return err
def pereh(W, t):
plt.figure(1) # Вывод графиков в отдельном окне
y1, t = con.step(W, t)
lines = [y1]
lines[0] = plt.plot(t, y1, "r")
plt.legend(lines[0], ['h(t) для 1'], loc='best', fontsize=10)
plt.title('Переходная характеристика', fontsize=10)
plt.ylabel('h')
plt.xlabel('t, c')
plt.grid()
plt.show()
def get_trans_func(self):
print(self.w)
y1, t1 = step(self.w, self.t)
plt.plot(self.t, y1, "r")
plt.title('Step Response')
plt.ylabel('Amplitude h(t)')
plt.xlabel('Time(sec)')
plt.grid(True)
plt.show()
return
def step_response(self):
if self._verbose:
print("[PID Design] Calculating PID gains")
pid_tf, closedLoop_tf = self.pid_design()
end_time = 5 # [s]
npts = int(old_div(end_time, self._dt)) + 1
T = np.linspace(0, end_time, npts)
yout, T = cnt.step(closedLoop_tf, T)
u, _, _ = cnt.lsim(pid_tf, 1 - yout, T)
return T, u, yout
def action_pid(sys, final_time, setpoint, Kp, Ki, Kd):
# open-loop system transfer function
try:
num, den = model(sys)
except:
# for error detection
print("Err: system in not defined")
return
Gs = control.tf(num, den)
s = matlab.tf('s')
Ds = Kp + Ki/s + Kd*s
# closed-loop unity-feedback transfer function
Ts = control.feedback(Ds*Gs, 1)
# simulation time parameters
initial_time = 0
nsteps = 40 * int(final_time) # number of time steps
t = np.linspace(initial_time, final_time, round(nsteps))
output, t = matlab.step(Ts, t)
output = setpoint*output
# calculate list of error
setpoint_arr = setpoint * np.ones(nsteps)
err = setpoint_arr - output
action = []
sum = 0
for i in range(len(err)):
if i == 0:
action.append(Kp*err[i] + Kd*(err[i]-0)/t[1])
else:
sum += t[1]*(err[i]+err[i-1])/2
action.append(Kp*err[i] + Kd*(err[i]-err[i - 1])/t[1] + Ki * sum)
# round lists to 6 decimal digits
ndigits = 6
t = [round(num, ndigits) for num in t]
action = [round(num, ndigits) for num in action]
# calculate maximum control action
max_action = max(action)
min_action = min(action)
return t, action, min_action, max_action
def action_zpk(sys, final_time, setpoint, z, p, k):
# open-loop system transfer function
try:
num, den = model(sys)
except:
# for error detection
print("Err: system in not defined")
return
Gs = control.tf(num, den)
z = np.array([z])
p = np.array([p])
num, den = matlab.zpk2tf(z, p, k)
Ds = matlab.tf(num, den)
# closed-loop unity-feedback transfer function
Ts = control.feedback(Ds*Gs, 1)
# simulation time parameters
initial_time = 0
nsteps = 40 * int(final_time) # number of time steps
t = np.linspace(initial_time, final_time, round(nsteps))
output, t = matlab.step(Ts, t)
output = setpoint*output
# calculate list of error
setpoint_arr = setpoint * np.ones(nsteps)
err = setpoint_arr - output
# calculate control action
action = matlab.lsim(Ds, err, t)
# covert numpy arrays to lists
t = list(t)
action = list(action[0])
# round lists to 6 decimal digits
ndigits = 6
t = [round(num, ndigits) for num in t]
action = [round(num, ndigits) for num in action]
# calculate maximum control action
max_action = max(action)
min_action = min(action)
return t, action, min_action, max_action
def intergalnaya_Otsenka(W):
a = 0
b = 100
n = 1000
h = (b - a) / n
t = np.linspace(a, b, num=n)
y, x = con.step(W, t) # х-время ПП
func = 0
x0 = x[0]
y0 = y[0]
# нахождение площади методом трапеций
for i in range(0, len(x), 1):
xi = x[i]
y1 = y[i]
func += abs(1 * (xi - x0) - 0.5 * (y1 + y0) * (xi - x0))
x0 = xi
y0 = y1
# print(func)
print("Интегральная оценка за ", b, " c = ", func)
def control(s, Y, D):
end = 10
G = 1 / (3 * s * (s + 1))
F = (3 * s * (s + 1))
KP, KD, KI = 16, 7, 4
H_pd = (KP + KD * s)
H_pid = (KP + KD * s + KI / s)
PD = cm.feedback((KP + KD * s) * G, 1) # PD controller
PID = cm.feedback((KP + KD * s + KI / s) * G, 1) # PID controller
PD_FF = cm.tf([1], [1])
PID_FF = cm.tf([1], [1]) # PID controller
D_PD = (G) / (1 + (KP + KD * s) * (G))
D_PID = (G) / (1 + (KP + KD * s + KI / s) * (G))
# Now, according to the Example, we calculate the response of
# the closed loop system to a step input of size 10:
out_PD, t = cm.step(Y * PD, np.linspace(0, end, 200))
out_PID, t = cm.step(Y * PID, np.linspace(0, end, 200))
out_PD_FF, t = cm.step(Y * PD_FF, np.linspace(0, end, 200))
out_PID_FF, t = cm.step(Y * PID_FF, np.linspace(0, end, 200))
out_PD_D, t = cm.step(-D * D_PD, np.linspace(0, end, 200))
out_PID_D, t = cm.step(-D * D_PID, np.linspace(0, end, 200))
theta_PD = out_PD + out_PD_D
theta_PID = out_PID + out_PID_D
theta_PD_FF = out_PD_FF + out_PD_D
theta_PID_FF = out_PID_FF + out_PID_D
y_out, t = cm.step(Y, np.linspace(0, end, 200))
plt.plot(t, theta_PD, lw=2, label="PD")
plt.plot(t, theta_PID, lw=2, label="PID")
if D != 0:
plt.plot(t, theta_PD_FF, lw=2, label="PD_FF")
plt.plot(t, theta_PID_FF, lw=2, label="PID_FF")
plt.plot(t, y_out, lw=1, label="Reference")
plt.xlabel('Time')
plt.ylabel('Position')
plt.legend()
def step_zpk(sys, final_time, setpoint, z, p, k):
# open-loop system transfer function
try:
num, den = model(sys)
except:
# for error detection
print("Err: system in not defined")
return
Gs = control.tf(num, den)
# define compensator transfer function
# convert zero and pole to numpy arrays
z = np.array([z])
p = np.array([p])
num, den = matlab.zpk2tf(z, p, k)
Ds = matlab.tf(num, den)
# Compensated open-loop transfer function
DsGs = Ds*Gs
# closed-loop unity-feedback transfer function
Ts = control.feedback(DsGs, 1)
# simulation time parameters
initial_time = 0
nsteps = 40 * final_time # number of time steps
t = np.linspace(initial_time, final_time, round(nsteps))
output, t = matlab.step(Ts, t)
output = setpoint*output
# covert numpy arrays to lists
t = list(t)
output = list(output)
# round lists to 6 decimal digits
ndigits = 6
t = [round(num, ndigits) for num in t]
output = [round(num, ndigits) for num in output]
return t, output
def do_direct_method(w):
"""
definition for analyzing regulator quality by step responce.
parameters in research:
- regulation time
- hesistation
- overshoot
- degree of attenuation
:return: keys for handle changing regulator quality
"""
def get_degree(ideal, actual):
"""
функция оценивает полученное значение по пятибальной шкале
:param ideal: идеальное значение
:param actual: реальное значение
:return: оценка
"""
mas = [0, 1, 1.2, 1.4, 1.6, 1.8, 2]
for i in range(len(mas) - 1):
if mas[i] * ideal <= actual < mas[i + 1] * ideal:
return len(mas) - 2 - i
elif actual > mas[-1] * ideal:
return 0
t = np.linspace(0, stop=100, num=2000)
counter = regulation_time = t_vr_reg = integral_mean = 0
y1, t1 = step(w, t)
y1 = list(y1)
max_y = max(y1)
last_y = y1[-1]
"""перерегулирование и его оценка"""
overshoot = (max_y - last_y) / last_y
key_per = get_degree(27, overshoot)
"""величина и время достижения первого максимума и их оценка"""
key_vel_max = get_degree(1.1, max(y1))
key_vr_max = get_degree(1, t[y1.index(max(y1))])
two_maxes = []
for num in range(len(y1[1:-1])):
if y1[num - 1] < y1[num] > y1[num + 1]:
two_maxes.append(y1[num])
if len(two_maxes) == 10: break
"""степень затухания и ее оценка"""
if len(two_maxes) <= 1:
key_deg = 5.0
else:
degree_of_attenuation = (1 - two_maxes[1] / two_maxes[0]) * 100
if degree_of_attenuation <= 0:
key_deg = -1
else:
key_deg = get_degree(1 / 6.6, 1 / degree_of_attenuation)
for i in range(len(y1)):
if 0.95 * last_y < y1[i] < 1.05 * last_y:
"""
counter - счетчик входящих в диапазон установившегося значения точек
устраняет вероятность ошибки попадания условия в нулевой промежуток колебательной
функции
"""
counter += 1
if counter == 20: # функция внутри диапазона, удовл. достаточности уст. режима.
regulation_time = t[i]
# номер нахождения в массиве t значения времени регулирования
t_vr_reg = i
break
else: # функция еще не в установившемся значении
counter = 0
"""оценка времени регулирования"""
key_reg = get_degree(15, regulation_time)
"""оценка показателя колебательности"""
if len(two_maxes) <= 1:
key_koleb = 5.0
else:
koleb = two_maxes[1] / two_maxes[0] * 100
key_koleb = get_degree(1.19, koleb)
"""интеграл и его оценка"""
for i in range(0, t_vr_reg):
integral_mean = integral_mean + abs(y1[t_vr_reg] - y1[i]) * t[1]
key_int = get_degree(0.3, integral_mean)
# проверка системы на устойчивость
poles, zeros = pzmap(w, Plot=False)
if not is_sustainable(poles):
return [-100]
return [
key_koleb, key_reg, key_per, key_deg, key_vel_max, key_vr_max, key_int
]
#print_c2d_matlablike(syst_fake_dis) # Python: # zero order hold by default # 4.96670866519e-05 z 4.93370716897e-05 # ----------------------- # z^2 1.0 z^2 -1.97990166083 z 0.980198673307 # # Sample time: 0.01 seconds # Discrete-time transfer function. # MATLAB: # [output,t]=step(syst_fake_dis) ### OLD scipy.signal.step way! ###output, t = step(syst_fake_dis[0][0], syst_fake_dis[1])# ,T=200) # MATLAB calculates the T size automatically based on black magic different than python one, see [MATLAB]/toolbox/shared/controllib/engine/@DynamicSystem/step.m: output, t = step(syst_fake_dis) # MATLAB: # plot(output) plot(output[0]) show() # out_len = len(output) out_len = len(output[0]) print "out_len is:" print out_len print "output is:" print output[0] # input=1:650; # input(:)=1; input_ = np.ones(out_len) # [num,den]=stmcb(output,input,0,2)
8., 0., 0., 0.; \
0., 4., 0., 0.; \
0., 0., 1., 0.')
B = np.matrix('2.; 0.; 0.; 0.')
C = np.matrix('0.5, 0.6875, 0.7031, 0.5')
D = np.matrix('0.')
# The full system
fsys = StateSpace(A,B,C,D)
# The reduced system, truncating the order by 1
ord = 3
rsys = msimp.balred(fsys,ord, method = 'truncate')
# Comparison of the step responses of the full and reduced systems
plt.figure(1)
y, t = mt.step(fsys)
yr, tr = mt.step(rsys)
plt.plot(t.T, y.T)
plt.plot(tr.T, yr.T)
# Repeat balanced reduction, now with 100-dimensional random state space
sysrand = mt.rss(100, 1, 1)
rsysrand = msimp.balred(sysrand,10,method ='truncate')
# Comparison of the impulse responses of the full and reduced random systems
plt.figure(2)
yrand, trand = mt.impulse(sysrand)
yrandr, trandr = mt.impulse(rsysrand)
plt.plot(trand.T, yrand.T, trandr.T, yrandr.T)
def plot_sys(sys): ys, ts = step(sys) plot(ts, ys)
