12/2/24, 6:55 AM Untitled5.
ipynb - Colab
import random
import numpy as np
import matplotlib.pyplot as plt
#--- COST FUNCTION AND PENALTY ------------------------------------------------+
# Function to maximize
def func1(x):
return -(x[0]**5 - 81*x[0]**4 + 2330*x[0]**3 - 28750*x[0]**2 +
150000*x[0] + 0.5*x[1]**5 - 65*x[1]**4 + 2950*x[1]**3 -
53500*x[1]**2 + 305000*x[1])
# Penalty function for constraints
def constraint_penalty(x):
penalty = 0
# Constraint 1: x1 + 2x2 <= 110
if x[0] + 2*x[1] > 110:
penalty += (x[0] + 2*x[1] - 110)**2
# Constraint 2: 3x1 + x2 <= 120
if 3*x[0] + x[1] > 120:
penalty += (3*x[0] + x[1] - 120)**2
# Constraint 3: 0 <= x1 <= 36
if x[0] < 0:
penalty += (0 - x[0])**2
if x[0] > 36:
penalty += (x[0] - 36)**2
# Constraint 4: 0 <= x2 <= 50
if x[1] < 0:
penalty += (0 - x[1])**2
if x[1] > 50:
penalty += (x[1] - 50)**2
return penalty
# Combined cost function with penalty
def constrained_func(x):
base_cost = func1(x)
penalty = constraint_penalty(x)
return base_cost + penalty
#--- GENETIC ALGORITHM IMPLEMENTATION -----------------------------------------+
def genetic_algorithm(func, bounds, population_size, generations, mutation_rate, crossover_rate):
# Initialize population
population = [
[random.uniform(bounds[i][0], bounds[i][1]) for i in range(len(bounds))]
for _ in range(population_size)
]
# Track best solution
best_solution = None
best_score = float('-inf')
# Log data for visualization
gen_scores = []
best_x1 = []
best_x2 = []
for gen in range(generations):
# Evaluate fitness of the population
fitness = [func(ind) for ind in population]
# Find the best individual
gen_best_idx = np.argmax(fitness)
gen_best_ind = population[gen_best_idx]
gen_best_score = fitness[gen_best_idx]
if gen_best_score > best_score:
best_solution = gen_best_ind
best_score = gen_best_score
# Logging for plots
gen_scores.append(best_score)
best_x1.append(best_solution[0])
best_x2.append(best_solution[1])
# Selection (Roulette Wheel Selection)
total_fitness = sum(fitness)
probabilities = [f / total_fitness for f in fitness]
selected_population = random.choices(population, probabilities, k=population_size)
# Crossover
i []
https://colab.research.google.com/drive/1aUjS0Beoa1xzNzpubPXccOJrsYimFJmV#printMode=true 1/3
12/2/24, 6:55 AM Untitled5.ipynb - Colab
next_generation = []
for _ in range(population_size // 2):
if random.random() < crossover_rate:
parent1 = random.choice(selected_population)
parent2 = random.choice(selected_population)
crossover_point = random.randint(1, len(bounds) - 1)
child1 = parent1[:crossover_point] + parent2[crossover_point:]
child2 = parent2[:crossover_point] + parent1[crossover_point:]
next_generation.extend([child1, child2])
else:
next_generation.extend(random.sample(selected_population, 2))
# Mutation
for individual in next_generation:
if random.random() < mutation_rate:
index = random.randint(0, len(bounds) - 1)
individual[index] = random.uniform(bounds[index][0], bounds[index][1])
population = next_generation
# Plotting
plt.plot(range(generations), gen_scores, label='Objective Function Value')
plt.xlabel('Generation')
plt.ylabel('Objective Function Value')
plt.title('GA Optimization Progress')
plt.legend()
plt.grid()
plt.show()
print("Best Solution:", best_solution)
print("Best Score:", best_score)
# Plot positions vs iterations
plt.plot(range(generations), best_x1, label="x1")
plt.plot(range(generations), best_x2, label="x2")
plt.xlabel("Generation")
plt.ylabel("Position")
plt.legend()
plt.title("Variable Positions vs Generations")
plt.grid()
plt.show()
return best_solution, best_score
#--- RUN GA -------------------------------------------------------------------+
# Parameters
bounds = [(20, 21), (35,36)] # Variable bounds: [(x1_min, x1_max), (x2_min, x2_max)]
population_size = 50 # Number of individuals in the population
generations = 100 # Number of generations to run
mutation_rate = 0.1 # Probability of mutation
crossover_rate = 0.7 # Probability of crossover
# Initial run
best_sol, best_val = genetic_algorithm(
constrained_func, bounds, population_size, generations, mutation_rate, crossover_rate
)
# Print final result
print("Optimal Solution Found:", best_sol)
print("Optimal Value:", best_val)
https://colab.research.google.com/drive/1aUjS0Beoa1xzNzpubPXccOJrsYimFJmV#printMode=true 2/3
12/2/24, 6:55 AM Untitled5.ipynb - Colab
Best Solution: [20.059028968447304, 35.99723424389268]
Best Score: -717497.028987119
Optimal Solution Found: [20.059028968447304, 35.99723424389268]
Optimal Value: -717497.028987119
https://colab.research.google.com/drive/1aUjS0Beoa1xzNzpubPXccOJrsYimFJmV#printMode=true 3/3
