-
Notifications
You must be signed in to change notification settings - Fork 1
/
PSO.py
92 lines (86 loc) · 4.6 KB
/
PSO.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 os
import numpy as np
#import matplotlib.pyplot as plt
class PSO(object):
def __init__(self, population_size, max_steps):
self.w = 0.6 # 惯性权重
self.c1 = self.c2 = 2
self.population_size = population_size # 粒子群数量
self.dim = 4 # 搜索空间的维度
#self.dim = 2 # 搜索空间的维度
self.max_steps = max_steps # 迭代次数
self.x_bound = [8, 12] # 解空间范围 TODO 更改这部分，表示nz,nz,ny范围
self.n_bound = [1, 4] #TODO 更改这部分，表示进程数范围
self.core = 24 #机器核心数量 TODO 更改这部分
self.x1 = np.random.uniform(self.x_bound[0], self.x_bound[1],
(self.population_size, self.dim - 1)) # 初始化粒子群位置
self.x2 = np.random.randint(1,4,size=(self.population_size, 1))
self.x = np.hstack((self.x1, self.x2)) #拼接
self.v = np.random.rand(self.population_size, self.dim) # 初始化粒子群速度
fitness = self.calculate_fitness(self.x, self.core)
self.p = self.x # 个体的最佳位置
self.pg = self.x[np.argmax(fitness)] # 全局最佳位置
self.individual_best_fitness = fitness # 个体的最优适应度
self.global_best_fitness = np.max(fitness) # 全局最佳适应度
best_pg = np.rint(self.pg)
print("全局最优gflop:%.5f ,平均gflop:%.5f"%(self.global_best_fitness, np.mean(fitness)))
print("全局最优nx值:"+str(int(best_pg[0])) + \
" ,全局最优ny值:" + str(int(best_pg[1])) + \
" ,全局最优nz值:" + str(int(best_pg[2])) + \
" ,全局最优进程数:" + str(int(best_pg[3])) + \
" ,全局最优线程数:" + str(int(self.core / int(best_pg[3]))))
print('\n')
def calculate_fitness(self, x, core):
x = np.rint(x)
y = np.zeros(len(x))
for i in range(len(x)):
print("=========目前的参数==========")
os.environ['nx'] = str(int(x[i][0]))
os.environ['ny'] = str(int(x[i][1]))
os.environ['nz'] = str(int(x[i][2]))
os.environ['n'] = str(int(x[i][3]))
os.environ['OMP_NUM_THREADS'] = str(int(core / int(x[i][3])))
print("nx值:"+str(int(x[i][0])) + \
" ,ny值:" + str(int(x[i][1])) + \
" ,nz值:" + str(int(x[i][2])) + \
" ,进程数:" + str(int(x[i][3])) + \
" ,线程数:" + str(int(core / int(x[i][3]))))
os.system('mpirun -n $n ./xhpcg --nx $nx --ny $ny --nz $nz >> a.txt')
gflop = os.popen('bash result.sh').read().strip('\n')
y[i] = float(gflop)
print("第%d个个体计算所得值:%f"%(i,y[i]))
return y
def evolve(self):
#fig = plt.figure()
for step in range(self.max_steps):
print("第%d轮寻优"%step)
r1 = np.random.rand(self.population_size, self.dim)
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-self.x)+self.c2*r2*(self.pg-self.x)
self.x = self.v + self.x
#plt.clf()
#plt.scatter(self.x[:, 0], self.x[:, 1], s=30, color='k')
#plt.xlim(self.x_bound[0], self.x_bound[1])
#plt.ylim(self.x_bound[0], self.x_bound[1])
#plt.pause(0.01)
fitness = self.calculate_fitness(self.x, self.core)
# 需要更新的个体
update_id = np.greater(self.individual_best_fitness, fitness)
self.p[update_id] = self.x[update_id]
self.individual_best_fitness[update_id] = fitness[update_id]
# 新一代出现了更大的fitness，所以更新全局最优fitness和位置
if np.max(fitness) > self.global_best_fitness:
self.pg = self.x[np.argmax(fitness)]
self.global_best_fitness = np.max(fitness)
best_pg = np.rint(self.pg)
print('全局最优gflop: %.5f, 平均最优gflop: %.5f' % (self.global_best_fitness, np.mean(fitness)))
print("全局最优nx值:"+str(int(best_pg[0])) + \
" ,全局最优ny值:" + str(int(best_pg[1])) + \
" ,全局最优nz值:" + str(int(best_pg[2])) + \
" ,全局最优进程数:" + str(int(best_pg[3])) + \
" ,全局最优线程数:" + str(int(self.core / int(best_pg[3]))))
print('\n')
pso = PSO(2, 10)
pso.evolve()
#plt.show()