hobinkwak / comparison-of-PSO-and-GA

Particle Swarm Optimization vs. Genetic Algorithm Test

comparison of PSO and GA

• Particle Swarm Optimization vs. Genetic Algorithm Test

Test

• Rosenbrock function (=Banana function)
  • minimum solution : (1, 1)

def f(X):
    a, b = 1, 100
    return ((a - X[0]) ** 2) + b*(X[1]-X[0]**2)**2

• Rastrigin function
  • minimum solution : (0, 0)

def f(X):
    A = 10
    return A*len(X) + sum([(x**2 - A * np.cos(2 * np.pi * x)) for x in X])

PSO

• Particle Swarm Optimization Code in pso.py
  • Swarm size : size (default: 1000, init parameter)
  • Number of iterations : n_iter (default: 1000, init parameter)
  • Inertia Weight : w (default: 0.5, init parameter)
  • Cognitive Weight : c1 (default: 0.25, init parameter)
  • Social Weight : c2 (default: 0.25, init parameter)

Requirements

• geneticalgorithm==1.0.2
• matplotlib==3.5.1
• numpy==1.20.0