simulated_annealing.py

# -*- coding: utf-8 -*-
"""simulated_annealing.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/16rdBKrPto56m1hWIYIjKmTLM7XBxZmV5
"""

import time
import random
import math
import numpy as np
import matplotlib.pyplot as plt

# Customization section
i_temp = 273     # initial temperature
cool   = 10     # cooling coefficient
n_var  = 2       # number of variables
u_b    = [6,6]   # upper bounds
l_b    = [-6,-6] # lower bounds
c_time = 0.5   # computing time

def objective_function(X):
  x   = X[0]
  y   = X[1]
  val = 3*(1-x)**2*math.exp(-x**2 - (y+1)**2) - 10*(x/5 - x**3 - y**5)*math.exp(-x**2 - y**2) -1/3*math.exp(-(x+1)**2 - y**2)
  return val

# Simulated Annealing Algorithm
i_sol = np.zeros(n_var) # initial solution
for v in range(n_var):
  i_sol[v] = random.uniform(l_b[v],u_b[v])

c_sol  = i_sol # current solution
b_sol  = i_sol # best solution
n      = 1     # no of solutions accepted
best_f = objective_function(b_sol)
c_temp = i_temp # current temperature

start      = time.time()
n_attemp   = 100 # number of attempts in each level of temperature
rec_best_f = []  # record best fitness

for i in range(9999):
  for j in range(n_attemp):

    for k in range(n_var):
      c_sol[k] = b_sol[k] + 0.1 * (random.uniform(l_b[k],u_b[k]))
      c_sol[k] = max(min(c_sol[k], u_b[k]), l_b[k])
    
    c_fit = objective_function(c_sol)
    E     = abs(c_fit - best_f)
    if i==0 and j==0:
      EA =E

    if c_fit > best_f:
      p = math.exp(-E/(EA*c_temp))
      # make a decision to accept the worse solution or not
      if random.random()<p:
        accept=True
      else:
        accept=False
    else:
      accept=True
    if accept==True:
      b_sol  = c_sol                # update the best solution
      best_f = objective_function(b_sol)
      n      = n+1                  # count solutions accepted
      EA     = (EA * (n-1) + E) / n # update EA

  print('iteration: {}, best_solution: {}, best_fitness: {}'.format(i, b_sol, best_f))
  rec_best_f.append(best_f)

  # Cooling the temperature
  c_temp = c_temp * cool

  # Stop timer
  end = time.time()
  if end-start >= c_time:
    break

plt.plot(rec_best_f)