W12_CompensatorDesign

English translation

!pip install control

Requirement already satisfied: control in /usr/local/lib/python3.10/dist-packages (0.10.0)
Requirement already satisfied: numpy>=1.23 in /usr/local/lib/python3.10/dist-packages (from control) (1.25.2)
Requirement already satisfied: scipy>=1.8 in /usr/local/lib/python3.10/dist-packages (from control) (1.11.4)
Requirement already satisfied: matplotlib>=3.6 in /usr/local/lib/python3.10/dist-packages (from control) (3.7.1)
Requirement already satisfied: contourpy>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (1.2.1)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (0.12.1)
Requirement already satisfied: fonttools>=4.22.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (4.51.0)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (1.4.5)
Requirement already satisfied: packaging>=20.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (24.0)
Requirement already satisfied: pillow>=6.2.0 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (9.4.0)
Requirement already satisfied: pyparsing>=2.3.1 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (3.1.2)
Requirement already satisfied: python-dateutil>=2.7 in /usr/local/lib/python3.10/dist-packages (from matplotlib>=3.6->control) (2.8.2)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-packages (from python-dateutil>=2.7->matplotlib>=3.6->control) (1.16.0)

import numpy as np
import control as ctl
import matplotlib.pyplot as plt

Phase-lead compensator using root locus

Phase-lead compensator is defined as, $C(s)=K\frac{(s-z_{0})}{(s-p_{0})}, \;\; |z_0|<|p_0|.$ Let $G(s)=\frac{40}{s(s+2)}$ and the requirements are

$K_v>20$
$\zeta=0.4$
$\omega_n=10$

s=ctl.tf('s')
G = 40/(s*(s+2))
G.name = 'G(s)'
zeta = 0.4
omega_n = 10
results = ctl.root_locus(G, xlim=(-5,0.5), ylim=(-1,15))

<Figure size 640x480 with 1 Axes>

Set $z_0=-\zeta \cdot \omega_n$

z_0 = -zeta*omega_n
z_0

-4.0

desired_root = z_0+1j*(omega_n**2-z_0**2)**0.5
desired_root

(-4+9.16515138991168j)

poles = ctl.poles(G)
angle_p_0 = np.pi-(sum(np.angle(desired_root-poles))-np.angle(desired_root-z_0))
angle_p_0

0.9444296476516962

p_0 = z_0 - (omega_n**2-z_0**2)**0.5 / np.tan(angle_p_0)
p_0

-10.631578947368421

C_lead = (s-z_0)/(s-p_0)
results = ctl.root_locus(C_lead*G, xlim=(-12,0), ylim=(-1, 10))
plot_root = plt.plot(desired_root.real, desired_root.imag, 'rs')

<Figure size 640x480 with 1 Axes>

rlist, klist = ctl.root_locus(C_lead*G, kvect=np.linspace(0, 10, 1000), plot=False)
distances = np.min(np.abs(rlist - desired_root),axis=1)
closest_idx = np.argmin(distances)
K = klist[closest_idx]
K

2.8928928928928928

C_lead = (s-z_0)/(s-p_0) * K
Kv = ctl.dcgain(ctl.minreal(G*C_lead*s))
Kv

1 states have been removed from the model

21.768302956421767

sys_cl = ctl.feedback(G*C_lead,1)
t,y = ctl.step_response(sys_cl, np.arange(0, 1.5, 0.01))
result = plt.plot(t,y)

<Figure size 640x480 with 1 Axes>

ctl.step_info(sys_cl)

{'RiseTime': 0.13956024882631715,
 'SettlingTime': 1.0641468973006682,
 'SettlingMin': 0.931400253036462,
 'SettlingMax': 1.3421386540504088,
 'Overshoot': 34.21386540504088,
 'Undershoot': 0,
 'Peak': 1.3421386540504088,
 'PeakTime': 0.3314555909625032,
 'SteadyStateValue': 1.0}

Phase-lead compensator using frequency response

Phase-lead compensator is defined as, $C(s)=K\frac{(s-z_{0})}{(s-p_{0})}, \;\; |z_0|<|p_0|.$ Let $G(s)=\frac{40}{s(s+2)}$ and the requirements are

$K_v>20$
PM $>40$

bode_G = ctl.bode(G)
gm, pm, Wcg, Wcp = ctl.margin(G)
pm

17.964235916371365

<Figure size 640x480 with 2 Axes>

pm_desired = 40
pm_required = pm_desired - pm + 8
pm_required

30.035764083628635

$sin(\phi_m) = \frac{\alpha-1}{\alpha+1}, \alpha = \frac{p_0}{z_0}$

from sympy.solvers import solve
from sympy import Symbol

a = Symbol('a')
sol = solve(np.sin(np.deg2rad(pm_required))-(a-1)/(a+1))
sol

[3.00432849083701]

alpha = float(sol[0])
C_gain_center = alpha**0.5
frequencies = np.linspace(1, 10, 1000)
mag, phase, omega = ctl.freqresp(G, frequencies)
closest_idx = np.argmin(np.abs(mag - 1/C_gain_center))
omega_m = omega[closest_idx]
omega_m

8.207207207207208

z_0 = -omega_m/C_gain_center
z_0

-4.735018607524815

p_0 = alpha*z_0
p_0

-14.225551307230202

C_lead_ = (s-z_0)/(s-p_0)
Kv = ctl.dcgain(ctl.minreal(G*C_lead_*s))
Kv

1 states have been removed from the model

6.657061656539413

C_lead = C_lead_ * 20/Kv
Kv = ctl.dcgain(ctl.minreal(G*C_lead*s))
Kv

1 states have been removed from the model

19.999999999999996

C_lead.name = 'G_c(s)'
G_comp = G*C_lead
G_comp.name = 'G_c(s)G(s)'
sys_list = [G, C_lead, G_comp]
result = ctl.bode(sys_list)

<Figure size 640x480 with 2 Axes>

gm_1, pm_1, Wcg_1, Wcp_1 = ctl.margin(G)
gm_2, pm_2, Wcg_2, Wcp_2 = ctl.margin(G_comp)
print('G, GM = %.2f dB (at %.2f rad/s) and PM = %.2f dB (at %.2f rad/s)' % (gm_1, Wcg_1, pm_1, Wcp_1))
print('G_comp, GM = %.2f dB (at %.2f rad/s) and PM = %.2f dB (at %.2f rad/s)' % (gm_2, Wcg_2, pm_2, Wcp_2))

G, GM = inf dB (at nan rad/s) and PM = 17.96 dB (at 6.17 rad/s)
G_comp, GM = inf dB (at nan rad/s) and PM = 43.73 dB (at 8.21 rad/s)

Phase-lag compensator using root locus

Phase-lag compensator is defined as, $C(s)=K\frac{(s-z_{0})}{(s-p_{0})}, \;\; |z_0|>|p_0|.$ Let $G(s)=\frac{40}{s(s+2)}$ and the requirements are

$K_v>20$
$\zeta=0.4$

Kv_desired = 20
zeta = 0.4
rlist, klist = ctl.root_locus(G, kvect=np.linspace(0, 10, 1000), xlim=(-2.5, 0.5), ylim=(-5, 5), plot=True)

<Figure size 640x480 with 1 Axes>

root_real = -1
theta = np.arccos(zeta)
root_angle = np.pi-theta
angles = np.min(np.abs(np.angle(rlist) - root_angle),axis=1)
closest_idx = np.argmin(angles)
K = klist[closest_idx]
K

0.16016016016016016

Kv = ctl.dcgain(ctl.minreal(K*G*s))
Kv

1 states have been removed from the model

3.203203203203203

alpha = Kv_desired / Kv
alpha

6.24375

z_0 = -0.1
p_0 = -0.1/alpha
z_0, p_0

(-0.1, -0.016016016016016016)

C_lag = K * (s-z_0) / (s-p_0)
C_lag

TransferFunction(array([0.16016016, 0.01601602]), array([1.        , 0.01601602]))

sys_cl = ctl.feedback(C_lag*G,1)
t,y = ctl.step_response(sys_cl, np.arange(0,12,0.1))
result = plt.plot(t,y)

<Figure size 640x480 with 1 Axes>

ctl.step_info(sys_cl)

{'RiseTime': 0.6027358403193015,
 'SettlingTime': 4.821886722554412,
 'SettlingMin': 0.9446154329113723,
 'SettlingMax': 1.3010305356637417,
 'Overshoot': 30.10305356637417,
 'Undershoot': 0,
 'Peak': 1.3010305356637417,
 'PeakTime': 1.4063836274117036,
 'SteadyStateValue': 1.0}

Phase-lag compensator using root locus

Phase-lag compensator is defined as, $C(s)=K\frac{(s-z_{0})}{(s-p_{0})}, \;\; |z_0|>|p_0|.$ Let $G(s)=\frac{40}{s(s+2)}$ and the requirements are

$K_v>20$
PM $>40$

bode_G = ctl.bode(G)
gm, pm, Wcg, Wcp = ctl.margin(G)
pm

17.964235916371365

<Figure size 640x480 with 2 Axes>

pm_desired = 50
phase_desired = np.deg2rad(-180+pm_desired)
frequencies = np.linspace(1, 10, 1000)
mag, phase, omega = ctl.freqresp(G, frequencies)
closest_idx = np.argmin(np.abs(phase - phase_desired))
omega_c = omega[closest_idx]
alpha = mag[closest_idx]
omega_c, alpha

(1.6756756756756757, 9.148792839892277)

z_0 = -omega_c/10
p_0 = z_0/alpha
K = 1/alpha
C_lag = K * (s-z_0) / (s-p_0)
C_lag

TransferFunction(array([0.10930404, 0.01831581]), array([1.        , 0.01831581]))

C_lag.name = 'G_c(s)'
G_comp = G*C_lag
G_comp.name = 'G_c(s)G(s)'
sys_list = [G, C_lag, G_comp]
result = ctl.bode(sys_list)

<Figure size 640x480 with 2 Axes>

gm_1, pm_1, Wcg_1, Wcp_1 = ctl.margin(G)
gm_2, pm_2, Wcg_2, Wcp_2 = ctl.margin(G_comp)
print('G, GM = %.2f dB (at %.2f rad/s) and PM = %.2f dB (at %.2f rad/s)' % (gm_1, Wcg_1, pm_1, Wcp_1))
print('G_comp, GM = %.2f dB (at %.2f rad/s) and PM = %.2f dB (at %.2f rad/s)' % (gm_2, Wcg_2, pm_2, Wcp_2))

G, GM = inf dB (at nan rad/s) and PM = 17.96 dB (at 6.17 rad/s)
G_comp, GM = inf dB (at nan rad/s) and PM = 44.88 dB (at 1.68 rad/s)