def hooke_jeeves(x0, epsilon, step, alpha):
it = 0
xB = x0
a = b = float('inf')
while c(x0) > epsilon or it < 1_000: #
# warunek stopu
if it == 0:
a = c(x0)
else:
a = b
b = c(x0)
if a == b:
break
it += 1
xB = x0
x0 = try_procedure(xB, step)
if c(x0) < c(xB):
while c(x0) >= c(xB):
xB_temp = xB
xB = x0
x0 = 2 * xB - xB_temp
x0 = try_procedure(x0, step)
else:
step = alpha * step
save(c(x0), 0)
save(xB, 0)
return xB
def hooke_jeeves(x0, epsilon, s, alpha, function):
it = 0
xB = x0
a = b = float('inf')
while s > epsilon or it < 1_000:
if it == 0:
a = c(x0, function)
else:
a = b
b = c(x0, function)
save(function, c(x0, function))
if a == b:
break
it += 1
xB = x0
x0 = try_procedure(xB, s, function)
if c(x0, function) < c(xB, function):
while c(x0, function) >= c(xB, function):
xB_temp = xB
xB = x0
x0 = 2 * xB - xB_temp
x0 = try_procedure(x0, s)
else:
s = alpha * s
return xB
def ivm_calculations(vector_a, Temp, _EPSILON_DOT, time, name):
Temp += 273.15
_KROK_CZASOWY = time / _ITERACJE
RO_0 = 10_000.0
historia_ro = [RO_0]
_Z = _EPSILON_DOT * exp(Q / (R * Temp))
mianownik = pow(_Z, vector_a[12])
_SWOBODNA_DROGA_DYSLOKACJI = vector_a[0] / mianownik
_B_L = _BURGERS * _SWOBODNA_DROGA_DYSLOKACJI
A1 = 1 / (_B_L)
A2 = vector_a[1] * pow(_EPSILON_DOT,
(-vector_a[8])) * exp(-vector_a[2] / (R * Temp))
A3 = vector_a[3] * (_TAU / D) * exp(-vector_a[4] / (R * Temp))
RO_CR = -vector_a[10] + vector_a[11] * pow(_Z, vector_a[9])
T_CR = -1
index_of_a3 = 0
for i in range(_ITERACJE):
if historia_ro[-1] >= RO_CR and historia_ro[-1] > 0:
temp = 1.0
if T_CR == -1:
T_CR = i
index_of_a3 = i - T_CR
last_step = A3*temp * \
pow(historia_ro[-1], vector_a[7])*historia_ro[index_of_a3]
else:
temp = 0.0
T_CR = -1
index_of_a3 = 0
last_step = 0.0
delta_ro = A1*_EPSILON_DOT - A2 * \
_EPSILON_DOT*historia_ro[-1] - last_step
RO_0 += delta_ro * _KROK_CZASOWY
historia_ro.append(RO_0)
save(RO_0, name)
return RO_0
def main():
search_space = s.Space(0.0, epsilon, 500)
particles_vector = [
p.Particle() for _ in range(search_space.num_particles)
]
search_space.particles = particles_vector
search_space.set_pbest()
search_space.set_gbest()
iteration = 0
n_iterations = 100
val_prev = val_prev_1 = 0.0
while iteration < n_iterations or search_space.gbest_value <= epsilon:
search_space.set_pbest()
search_space.set_gbest()
# JEŚLI PRZEZ TRZY ITERACJE NIE ZMIENI SIE WARTOŚĆ => BREAK
if iteration == 0:
val_prev = search_space.gbest_value
else:
val_prev_1 = val_prev
val_prev = search_space.gbest_value
#print("POŁOŻENIE: ", search_space.gbest_position, " WARTOŚĆ: ", search_space.gbest_value)
if abs(search_space.gbest_value - search_space.target
) <= search_space.epsilon or val_prev_1 == val_prev:
break
search_space.move_particles()
iteration += 1
save(search_space.gbest_value, 0)
print("Solution: ", search_space.gbest_position, " value ",
search_space.gbest_value)
file = open("model.txt", 'a')
file.write(str(search_space.gbest_position) + " \n")
file.close()
return search_space.gbest_position
_Z = _EPSILON_DOT * exp(Q / (R * Temp))
_BURGERS = 0.25 * pow(10, -9)
_SWOBODNA_DROGA_DYSLOKACJI = vector_a[0] / pow(_Z, vector_a[12])
_B2 = _BURGERS * _BURGERS
_TAU = (_KIRCHOFF * _B2 * 10.0**6) / 2
_B_L = _BURGERS * _SWOBODNA_DROGA_DYSLOKACJI
A1 = 1 / (_B_L)
A2 = vector_a[1] * pow(_EPSILON_DOT,
(-vector_a[8])) * exp(-vector_a[2] / (R * Temp))
A3 = vector_a[3] * (_TAU / D) * exp(-vector_a[4] / (R * Temp))
RO_CR = -vector_a[10] + vector_a[11] * pow(_Z, vector_a[9])
T_CR = -1
index_of_a3 = 0
save(RO_0, vector_a[6] + vector_a[5] * _KIRCHOFF * _BURGERS * sqrt(RO_0))
for i in range(_ITERACJE):
if historia_ro[-1] >= RO_CR:
temp = 1.0
if T_CR == -1:
T_CR = i
index_of_a3 = i - T_CR
else:
temp = 0.0
T_CR = -1
index_of_a3 = 0
print(A1, A2, A3)
delta_ro = A1 * _EPSILON_DOT - A2 * _EPSILON_DOT * historia_ro[
-1] - A3 * temp * pow(historia_ro[-1],
vector_a[7]) * historia_ro[index_of_a3]
# JEŚLI PRZEZ TRZY ITERACJE NIE ZMIENI SIE WARTOŚĆ => BREAK
if iteration == 0:
val_prev = search_space.gbest_value
else:
val_prev_1 = val_prev
val_prev = search_space.gbest_value
#print("POŁOŻENIE: ", search_space.gbest_position, " WARTOŚĆ: ", search_space.gbest_value)
if abs(search_space.gbest_value - search_space.target
) <= search_space.epsilon or val_prev_1 == val_prev:
break
search_space.move_particles()
iteration += 1
save(search_space.gbest_value, 0)
print("Solution: ", search_space.gbest_position, " value ",
search_space.gbest_value)
file = open("model.txt", 'a')
file.write(str(search_space.gbest_position) + " \n")
file.close()
return search_space.gbest_position
if __name__ == "__main__":
vector = main()
for i in range(9):
T, epsilon_dot = dict_T_epsilon_dot[i]
for j in range(21):
save(sigma(vector, 0.05 * (j + 1), epsilon_dot, T), i + 1)
from math import log, pow
from saveToFile import saveToFile as save
# WARTOŚCI A1, A2, A3, RO_CT I A8 PODMIENIAM DLA KAŻDEGO PRZYPADKU
A1 = 3.5
A2 = 5.0
A3 = 1
epsilon_dot = 1.0
ro_0 = 0.0
ro_cr = 0.4
a8 = 0.0
number_of_steps = 1000
historia_ro = [0.0]
t_cr = (1/A2) * log((ro_0-(A1/A2))/(ro_cr-(A1/A2)))
t_cr_step = int(t_cr*number_of_steps)
save(ro_0)
for i in range(number_of_steps):
if i < t_cr_step:
temp = 0.0
index = 0
else:
temp = 1.0
index = i-t_cr_step
delta = A1*epsilon_dot - A2*historia_ro[-1]*epsilon_dot - A3*pow(historia_ro[-1], a8)*temp*historia_ro[index]
ro_0 += delta*(1/number_of_steps)
historia_ro.append(ro_0)
save(ro_0)