# !/usr/bin/env python
# Created by "Thieu" at 14:52, 17/03/2020 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from mealpy.optimizer import Optimizer
[docs]class BaseBES(Optimizer):
"""
The original version of: Bald Eagle Search (BES)
Links:
1. https://doi.org/10.1007/s10462-019-09732-5
Hyper-parameters should fine tuned in approximate range to get faster convergen toward the global optimum:
+ a_factor (int): default: 10, determining the corner between point search in the central point, in [5, 10]
+ R_factor (float): default: 1.5, determining the number of search cycles, in [0.5, 2]
+ alpha (float): default: 2, parameter for controlling the changes in position, in [1.5, 2]
+ c1 (float): default: 2, in [1, 2]
+ c2 (float): c1 and c2 increase the movement intensity of bald eagles towards the best and centre points
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy.swarm_based.BES import BaseBES
>>>
>>> def fitness_function(solution):
>>> return np.sum(solution**2)
>>>
>>> problem_dict1 = {
>>> "fit_func": fitness_function,
>>> "lb": [-10, -15, -4, -2, -8],
>>> "ub": [10, 15, 12, 8, 20],
>>> "minmax": "min",
>>> "verbose": True,
>>> }
>>>
>>> epoch = 1000
>>> pop_size = 50
>>> a_factor = 10
>>> R_factor = 1.5
>>> alpha = 2.0
>>> c1 = 2.0
>>> c2 = 2.0
>>> model = BaseBES(problem_dict1, epoch, pop_size, a_factor, R_factor, alpha, c1, c2)
>>> best_position, best_fitness = model.solve()
>>> print(f"Solution: {best_position}, Fitness: {best_fitness}")
References
~~~~~~~~~~
[1] Alsattar, H.A., Zaidan, A.A. and Zaidan, B.B., 2020. Novel meta-heuristic bald eagle
search optimisation algorithm. Artificial Intelligence Review, 53(3), pp.2237-2264.
"""
def __init__(self, problem, epoch=10000, pop_size=100, a_factor=10, R_factor=1.5, alpha=2.0, c1=2.0, c2=2.0, **kwargs):
"""
Args:
problem (dict): The problem dictionary
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
a_factor (int): default: 10, determining the corner between point search in the central point, in [5, 10]
R_factor (float): default: 1.5, determining the number of search cycles, in [0.5, 2]
alpha (float): default: 2, parameter for controlling the changes in position, in [1.5, 2]
c1 (float): default: 2, in [1, 2]
c2 (float): c1 and c2 increase the movement intensity of bald eagles towards the best and centre points
"""
super().__init__(problem, kwargs)
self.epoch = epoch
self.pop_size = pop_size
self.a_factor = a_factor
self.R_factor = R_factor
self.alpha = alpha
self.c1 = c1
self.c2 = c2
self.nfe_per_epoch = 3 * pop_size
self.sort_flag = False
def _create_x_y_x1_y1_(self):
""" Using numpy vector for faster computational time """
## Eq. 2
phi = self.a_factor * np.pi * np.random.uniform(0, 1, self.pop_size)
r = phi + self.R_factor * np.random.uniform(0, 1, self.pop_size)
xr, yr = r * np.sin(phi), r * np.cos(phi)
## Eq. 3
r1 = phi1 = self.a_factor * np.pi * np.random.uniform(0, 1, self.pop_size)
xr1, yr1 = r1 * np.sinh(phi1), r1 * np.cosh(phi1)
x_list = xr / max(xr)
y_list = yr / max(yr)
x1_list = xr1 / max(xr1)
y1_list = yr1 / max(yr1)
return x_list, y_list, x1_list, y1_list
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
## 0. Pre-definded
x_list, y_list, x1_list, y1_list = self._create_x_y_x1_y1_()
# Three parts: selecting the search space, searching within the selected search space and swooping.
## 1. Select space
pos_list = np.array([individual[self.ID_POS] for individual in self.pop])
pos_mean = np.mean(pos_list, axis=0)
pop_new = []
for idx in range(0, self.pop_size):
pos_new = self.g_best[self.ID_POS] + self.alpha * np.random.uniform() * (pos_mean - self.pop[idx][self.ID_POS])
pos_new = self.amend_position(pos_new)
pop_new.append([pos_new, None])
pop_new = self.update_fitness_population(pop_new)
pop_new = self.greedy_selection_population(self.pop, pop_new)
## 2. Search in space
pos_list = np.array([individual[self.ID_POS] for individual in pop_new])
pos_mean = np.mean(pos_list, axis=0)
pop_child = []
for idx in range(0, self.pop_size):
idx_rand = np.random.choice(list(set(range(0, self.pop_size)) - {idx}))
pos_new = pop_new[idx][self.ID_POS] + y_list[idx] * (pop_new[idx][self.ID_POS] - pop_new[idx_rand][self.ID_POS]) + \
x_list[idx] * (pop_new[idx][self.ID_POS] - pos_mean)
pos_new = self.amend_position(pos_new)
pop_child.append([pos_new, None])
pop_child = self.update_fitness_population(pop_child)
pop_child = self.greedy_selection_population(pop_new, pop_child)
## 3. Swoop
pos_list = np.array([individual[self.ID_POS] for individual in pop_child])
pos_mean = np.mean(pos_list, axis=0)
pop_new = []
for idx in range(0, self.pop_size):
pos_new = np.random.uniform() * self.g_best[self.ID_POS] + x1_list[idx] * (pop_child[idx][self.ID_POS] - self.c1 * pos_mean) \
+ y1_list[idx] * (pop_child[idx][self.ID_POS] - self.c2 * self.g_best[self.ID_POS])
pos_new = self.amend_position(pos_new)
pop_new.append([pos_new, None])
pop_new = self.update_fitness_population(pop_new)
self.pop = self.greedy_selection_population(pop_child, pop_new)