-
Notifications
You must be signed in to change notification settings - Fork 5
/
粒子群算法.py
92 lines (83 loc) · 4.63 KB
/
粒子群算法.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
import numpy as np
import matplotlib.pyplot as plt
"""
粒子群算法实现步骤：
(1)初始化：
初始化参数：学习因子c1,c2:一般取2; 惯性权重w:一般取0.9-1.2; 种群数量N：50-1000;迭代次数：100-4000;空间维数：自变量的个数;位置限制：自变量的取值范围;速度限制：可以控制在0-1之间。
初始化粒子群：包括粒子群的初始位置和速度
根据初始化得到的粒子群计算p_best,g_best
(2)更新粒子位置和粒子的速度
(3)计算每个粒子的适应度值
(4)根据适应度值更新p_best(个体最佳历史位置),g_best(全局最佳历史位置)
(5)判断是否满足结束条件，满足输出最优解，不满足转到第(2)
"""
class PSO(object):
def __init__(self, population_size, max_steps):
self.x = None
self.w = 0.6 # 惯性权重
self.c1 = self.c2 = 2 # 学习因子
self.population_size = population_size # 粒子群数量
self.dim = 4 # 搜索空间的维度
self.max_steps = max_steps # 迭代次数
self.bound = [] # 自变量的约束范围
self.bound.append([1, 1, 1, 1])
self.x = np.zeros((self.population_size, self.dim))
self.bound.append([30, 30, 30, 30]) # 解空间范围
for i in range(self.population_size): # 对于每一只蚂蚁来说都会有一个潜在解
for j in range(self.dim): # 取出每一个变量,产生一个在[var_num_min,var_num_max]范围内的随机数
self.x[i][j] = np.random.uniform(self.bound[0][j], self.bound[1][j])
self.v = np.random.rand(self.population_size, self.dim) # 初始化粒子群速度
# print("v", self.v) # population行4列，
# fitness = self.calculate_fitness(self.x)
self.Fitness = []
for i in range(self.population_size):
fitness = self.calculate_fitness(self.x[i])
self.Fitness.append(fitness)
self.p_best = self.x # 个体的最优位置
print("个体最优位置：", self.p_best)
self.g_best = self.x[np.argmax(self.Fitness)] # 全局最优位置
print("全局最优位置：", self.g_best)
self.individual_best_fitness = self.Fitness # 个体的最优适应度
print("个体最优适应度：", self.individual_best_fitness)
self.global_best_fitness = np.max(self.Fitness) # 全局最佳适应度
print("全局最优适应度：", self.global_best_fitness)
def calculate_fitness(self, x):
x1 = x[0]
x2 = x[1]
x3 = x[2]
x4 = x[3]
y = x1 ** 2 + x2 ** 2 + x3 ** 3 + x4 ** 4
# print("fitness:", y)
return y
def evolve(self): # evolve进化的意思
fig = plt.figure()
for step in range(self.max_steps):
r1 = np.random.rand(self.population_size, self.dim) # 返回一个或者一组0-1之间的随机数或者随机数组
r2 = np.random.rand(self.population_size, self.dim)
# 更新速度和权重
#
self.v = self.w * self.v + self.c1 * r1 * (self.p_best - self.x) + self.c2 * r2 * (self.g_best - self.x)
self.x = self.v + self.x
for i in range(self.population_size):
for j in range(self.dim):
if self.v[i][j] < self.bound[0][j]:
self.v[i][j] = self.bound[0][j]
if self.x[i][j] > self.bound[1][j]:
self.x[i][j] = self.bound[1][j]
for i in range(self.population_size): # 对于每一个粒子.
fitness = self.calculate_fitness(self.x[i]) # 新的粒子群的适应度，100行1列
# 需要更新的个体
if fitness > self.individual_best_fitness[i]:
self.p_best[i] = self.x[i]
self.individual_best_fitness[i] = fitness # ppt中没有说要更新个体历史适应度，但我觉得应该是要更新的，这样收敛的速度会加快
print("用当前位置更新粒子个体的历史最优位置p_best")
print("更改之后的粒子最优位置：", self.p_best[i])
else:
print("当前个体适应度值 小于 这个个体的历史最优位置的适应度值，不用更新")
if fitness > self.global_best_fitness:
self.g_best = self.x[i]
self.global_best_fitness = fitness
print("个体最优适应度：", self.individual_best_fitness)
print('best fitness : %.5f,mean fitness : %.5f' % (self.global_best_fitness, np.mean(fitness)))
pso = PSO(10, 100)
pso.evolve()