# encoding=utf8
import numpy as np
from niapy.algorithms.algorithm import Algorithm
__all__ = ['GreyWolfOptimizer']
[docs]class GreyWolfOptimizer(Algorithm):
r"""Implementation of Grey wolf optimizer.
Algorithm:
Grey wolf optimizer
Date:
2018
Author:
Iztok Fister Jr. and Klemen Berkovič
License:
MIT
Reference paper:
* Mirjalili, Seyedali, Seyed Mohammad Mirjalili, and Andrew Lewis. "Grey wolf optimizer." Advances in engineering software 69 (2014): 46-61.
* Grey Wolf Optimizer (GWO) source code version 1.0 (MATLAB) from MathWorks
Attributes:
Name (List[str]): List of strings representing algorithm names.
See Also:
* :class:`niapy.algorithms.Algorithm`
"""
Name = ['GreyWolfOptimizer', 'GWO']
[docs] @staticmethod
def info():
r"""Get algorithm information.
Returns:
str: Algorithm information.
See Also:
* :func:`niapy.algorithms.Algorithm.info`
"""
return r"""Mirjalili, Seyedali, Seyed Mohammad Mirjalili, and Andrew Lewis. "Grey wolf optimizer." Advances in engineering software 69 (2014): 46-61."""
[docs] def init_population(self, task):
r"""Initialize population.
Args:
task (Task): Optimization task.
Returns:
Tuple[numpy.ndarray, numpy.ndarray, Dict[str, Any]]:
1. Initialized population.
2. Initialized populations fitness/function values.
3. Additional arguments:
* alpha (numpy.ndarray): Alpha of the pack (Best solution)
* alpha_fitness (float): Best fitness.
* beta (numpy.ndarray): Beta of the pack (Second best solution)
* beta_fitness (float): Second best fitness.
* delta (numpy.ndarray): Delta of the pack (Third best solution)
* delta_fitness (float): Third best fitness.
See Also:
* :func:`niapy.algorithms.Algorithm.init_population`
"""
pop, fpop, d = super().init_population(task)
si = np.argsort(fpop)
alpha = np.copy(pop[si[0]])
alpha_fitness = fpop[si[0]]
beta = np.copy(pop[si[1]])
beta_fitness = fpop[si[1]]
delta = np.copy(pop[si[2]])
delta_fitness = fpop[si[2]]
d.update({
'alpha': alpha,
'alpha_fitness': alpha_fitness,
'beta': beta,
'beta_fitness': beta_fitness,
'delta': delta,
'delta_fitness': delta_fitness
})
return pop, fpop, d
[docs] def run_iteration(self, task, population, population_fitness, best_x, best_fitness, **params):
r"""Core function of GreyWolfOptimizer algorithm.
Args:
task (Task): Optimization task.
population (numpy.ndarray): Current population.
population_fitness (numpy.ndarray): Current populations function/fitness values.
best_x (numpy.ndarray):
best_fitness (float):
**params (Dict[str, Any]): Additional arguments.
Returns:
Tuple[numpy.ndarray, numpy.ndarray, numpy.ndarray, float, Dict[str, Any]]:
1. New population
2. New population fitness/function values
3. Additional arguments:
* alpha (numpy.ndarray): Alpha of the pack (Best solution)
* alpha_fitness (float): Best fitness.
* beta (numpy.ndarray): Beta of the pack (Second best solution)
* beta_fitness (float): Second best fitness.
* delta (numpy.ndarray): Delta of the pack (Third best solution)
* delta_fitness (float): Third best fitness.
"""
alpha = params.pop('alpha')
alpha_fitness = params.pop('alpha_fitness')
beta = params.pop('beta')
beta_fitness = params.pop('beta_fitness')
delta = params.pop('delta')
delta_fitness = params.pop('delta_fitness')
a = 2 - task.evals * (2 / task.max_evals)
for i, w in enumerate(population):
a1, c1 = 2 * a * self.random(task.dimension) - a, 2 * self.random(task.dimension)
x1 = alpha - a1 * np.fabs(c1 * alpha - w)
a2, c2 = 2 * a * self.random(task.dimension) - a, 2 * self.random(task.dimension)
x2 = beta - a2 * np.fabs(c2 * beta - w)
a3, c3 = 2 * a * self.random(task.dimension) - a, 2 * self.random(task.dimension)
x3 = delta - a3 * np.fabs(c3 * delta - w)
population[i] = task.repair((x1 + x2 + x3) / 3, rng=self.rng)
population_fitness[i] = task.eval(population[i])
for i, f in enumerate(population_fitness):
if f < alpha_fitness:
alpha, alpha_fitness = population[i].copy(), f
elif alpha_fitness < f < beta_fitness:
beta, beta_fitness = population[i].copy(), f
elif beta_fitness < f < delta_fitness:
delta, delta_fitness = population[i].copy(), f
best_x, best_fitness = self.get_best(alpha, alpha_fitness, best_x, best_fitness)
return population, population_fitness, best_x, best_fitness, {'alpha': alpha, 'alpha_fitness': alpha_fitness,
'beta': beta, 'beta_fitness': beta_fitness,
'delta': delta, 'delta_fitness': delta_fitness}
