# !/usr/bin/env python
# Created by "Thieu" at 07:03, 18/03/2020 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from copy import deepcopy
from mealpy.optimizer import Optimizer
[docs]class BaseHGSO(Optimizer):
"""
The original version of: Henry Gas Solubility Optimization (HGSO)
Links:
1. https://www.sciencedirect.com/science/article/abs/pii/S0167739X19306557
Hyper-parameters should fine tuned in approximate range to get faster convergen toward the global optimum:
+ n_clusters (int): [2, 10], number of clusters, default = 2
Examples
~~~~~~~~
>>> import numpy as np
>>> from mealpy.physics_based.HGSO import BaseHGSO
>>>
>>> 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
>>> n_clusters = 3
>>> model = BaseHGSO(problem_dict1, epoch, pop_size, n_clusters)
>>> best_position, best_fitness = model.solve()
>>> print(f"Solution: {best_position}, Fitness: {best_fitness}")
References
~~~~~~~~~~
[1] Hashim, F.A., Houssein, E.H., Mabrouk, M.S., Al-Atabany, W. and Mirjalili, S., 2019. Henry gas solubility
optimization: A novel physics-based algorithm. Future Generation Computer Systems, 101, pp.646-667.
"""
def __init__(self, problem, epoch=10000, pop_size=100, n_clusters=2, **kwargs):
"""
Args:
problem (dict): The problem dictionary
epoch (int): maximum number of iterations, default = 10000
pop_size (int): number of population size, default = 100
n_clusters (int): number of clusters, default = 2
"""
super().__init__(problem, kwargs)
self.nfe_per_epoch = pop_size
self.sort_flag = False
self.epoch = epoch
self.pop_size = pop_size
self.n_clusters = n_clusters
self.n_elements = int(self.pop_size / self.n_clusters)
self.T0 = 298.15
self.K = 1.0
self.beta = 1.0
self.alpha = 1
self.epxilon = 0.05
self.l1 = 5E-2
self.l2 = 100.0
self.l3 = 1E-2
self.H_j = self.l1 * np.random.uniform()
self.P_ij = self.l2 * np.random.uniform()
self.C_j = self.l3 * np.random.uniform()
self.pop_group, self.p_best = None, None
def _create_group(self, pop):
pop_group = []
for idx in range(0, self.n_clusters):
pop_group.append(pop[idx * self.n_elements:(idx + 1) * self.n_elements])
return pop_group
def _flatten_group(self, group):
pop = []
for idx in range(0, self.n_clusters):
pop += group[idx]
return pop
[docs] def initialization(self):
self.pop = self.create_population(self.pop_size)
_, self.g_best = self.get_global_best_solution(self.pop)
self.pop_group = self._create_group(self.pop)
self.p_best = self._get_best_solution_in_team(self.pop_group) # multiple element
def _get_best_solution_in_team(self, group=None):
list_best = []
for i in range(len(group)):
_, best_agent = self.get_global_best_solution(group[i])
list_best.append(best_agent)
return list_best
[docs] def evolve(self, epoch):
"""
The main operations (equations) of algorithm. Inherit from Optimizer class
Args:
epoch (int): The current iteration
"""
nfe_epoch = 0
## Loop based on the number of cluster in swarm (number of gases type)
for i in range(self.n_clusters):
### Loop based on the number of individual in each gases type
pop_new = []
for j in range(self.n_elements):
F = -1.0 if np.random.uniform() < 0.5 else 1.0
##### Based on Eq. 8, 9, 10
self.H_j = self.H_j * np.exp(-self.C_j * (1.0 / np.exp(-epoch / self.epoch) - 1.0 / self.T0))
S_ij = self.K * self.H_j * self.P_ij
gama = self.beta * np.exp(- ((self.p_best[i][self.ID_TAR][self.ID_FIT] + self.epxilon) /
(self.pop_group[i][j][self.ID_TAR][self.ID_FIT] + self.epxilon)))
X_ij = self.pop_group[i][j][self.ID_POS] + F * np.random.uniform() * gama * \
(self.p_best[i][self.ID_POS] - self.pop_group[i][j][self.ID_POS]) + \
F * np.random.uniform() * self.alpha * (S_ij * self.g_best[self.ID_POS] - self.pop_group[i][j][self.ID_POS])
pos_new = self.amend_position(X_ij)
pop_new.append([pos_new, None])
nfe_epoch += 1
self.pop_group[i] = self.update_fitness_population(pop_new)
self.pop = self._flatten_group(self.pop_group)
## Update Henry's coefficient using Eq.8
self.H_j = self.H_j * np.exp(-self.C_j * (1.0 / np.exp(-epoch / self.epoch) - 1.0 / self.T0))
## Update the solubility of each gas using Eq.9
S_ij = self.K * self.H_j * self.P_ij
## Rank and select the number of worst agents using Eq. 11
N_w = int(self.pop_size * (np.random.uniform(0, 0.1) + 0.1))
## Update the position of the worst agents using Eq. 12
sorted_id_pos = np.argsort([x[self.ID_TAR][self.ID_FIT] for x in self.pop])
pop_new = []
pop_idx = []
for item in range(N_w):
id = sorted_id_pos[item]
X_new = np.random.uniform(self.problem.lb, self.problem.ub)
pos_new = self.amend_position(X_new)
pop_new.append([pos_new, None])
pop_idx.append(id)
nfe_epoch += 1
pop_new = self.update_fitness_population(pop_new)
for idx, id_selected in enumerate(pop_idx):
self.pop[id_selected] = deepcopy(pop_new[idx])
self.pop_group = self._create_group(self.pop)
self.p_best = self._get_best_solution_in_team(self.pop_group)
self.nfe_per_epoch = nfe_epoch