-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathoptimization.py
193 lines (172 loc) · 9.06 KB
/
optimization.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
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
#!/usr/bin/python
# -*- coding:utf8 -*-
import numpy as np
import math
import os
import json
from collections import defaultdict
import pandas as pd
from util import get_Dataset, visualize_prediction, array_operate_with_nan, t2np, SimpleLogger,visualize_prediction_power,visualize_prediction_power
import torch
import argparse
import pickle
from matplotlib import pyplot as plt
parser = argparse.ArgumentParser(description='Models for Continuous Stirred Tank dataset')
parser.add_argument("--low", type=float, default=12, help='The temperature activating the work of cooling')
parser.add_argument("--high", type=float, default=20, help='The temperature stopping the cooling system')
parser.add_argument("--model", type=str, default='best_dev_model.pt')
parser.add_argument("--bptt", type=int, default=800, help="bptt")
parser.add_argument("--data", type=str, default='test.csv')
parser.add_argument("--save_dir", type=str, default='None')
parser.add_argument("--datasets", type=list, default=['train_P-1.7k','train_P-3.8k','train_P-6.3k'], help="datasets")
parser.add_argument("--datasets2", type=list, default=['heat load-1.7k','heat load-3.8k','heat load-6.3k'], help="datasets")
if __name__ == '__main__':
paras = parser.parse_args()
paras.save_dir = os.path.join('results', paras.save_dir)
save_dir_img = os.path.join(paras.save_dir,'optimization')
if not os.path.exists(save_dir_img):
os.mkdir(save_dir_img)
log_file_all = os.path.join(save_dir_img, 'all.txt' )
logging_all = SimpleLogger(log_file_all) # log to file
filename = "./results/optimization_test/data.pkl"
file = open(filename, "rb")
data = pickle.load(file)
X_mean = data['X_mean']
X_std = data['X_std']
Y_mean = data['Y_mean']
Y_std = data['Y_std']
sum_powers_all = []
cops_all = []
pue_all = []
for idx,everdata in enumerate(paras.datasets):
sum_powers = []
cops = []
pues = []
Min = 12
Max = 18.5
step = 0.5
Xtest, Ytest, ttest, stest = [df.to_numpy(dtype=np.float32) for df in
get_Dataset('./data/' + everdata + '.csv')]
dttest = np.append(ttest[1:] - ttest[:-1], ttest[1] - ttest[0]).reshape(-1, 1)
# Y_mean, Y_std = array_operate_with_nan(Ytest, np.mean), array_operate_with_nan(Ytest, np.std) # 归一化
# X_mean, X_std = array_operate_with_nan(Xtest, np.mean), array_operate_with_nan(Xtest, np.std)
Ytest = (Ytest - Y_mean) / Y_std
Xtest = (Xtest - X_mean) / X_std
Xtest_tn = torch.tensor(Xtest, dtype=torch.float).unsqueeze(0)
Ytest_tn = torch.tensor(Ytest, dtype=torch.float).unsqueeze(0)
dttest_tn = torch.tensor(dttest, dtype=torch.float).unsqueeze(0)
stest_tn = torch.tensor(stest, dtype=torch.int).unsqueeze(0) # (1, Ntrain, 1)
if torch.cuda.is_available():
Xtest_tn = Xtest_tn.cuda()
Ytest_tn = Ytest_tn.cuda()
dttest_tn = dttest_tn.cuda()
stest_tn = stest_tn.cuda()
save_dir = os.path.join(save_dir_img,everdata)
if not os.path.exists(save_dir):
os.mkdir(save_dir)
log_file = os.path.join(save_dir, '%s-%.2f-log.txt' % (everdata,paras.low))
logging = SimpleLogger(log_file) # log to file
logging('dataset: %s, Min: %.2f, Max: %.3f, ' % (everdata,Min, Max))
for low in np.arange(Min, Max, step):
device = torch.device('cuda:0') if torch.cuda.is_available() else torch.device('cpu')
model = torch.load(os.path.join(paras.save_dir, paras.model),
map_location=device) # load模型
model.aj_odes_forward.transforms = defaultdict(list)
model.aj_odes_forward.add_transform(1, 2, [[0, 'geq', paras.high]])
model.aj_odes_forward.add_transform(4, 1, [[0, 'leq',low]])
Ytest_pred, test_state_pred = model.encoding_plus_predict( # 模型预测
Xtest_tn, dttest_tn, Ytest_tn[:, :paras.bptt], stest_tn[:, :paras.bptt], paras.bptt, None)
test_aj_state_pred_array = model.select_aj_states(
test_state_pred[0]).int().detach().cpu().numpy() # 把状态也拿出来
visualize_prediction_power( # 可视化画图专用
Ytest[paras.bptt:] * Y_std + Y_mean, t2np(Ytest_pred) * Y_std + Y_mean, test_aj_state_pred_array,
Xtest[paras.bptt:, 0] * X_std[0] + X_mean[0],low
,save_dir,
seg_length=1000, dir_name='vis-%.1f' % (low))
predicted_power = (t2np(Ytest_pred) * Y_std + Y_mean)[..., -1].reshape(-1) # 算power
predicted_power[predicted_power < 0] = 0
sum_power = (
predicted_power[:7200] *
dttest_tn[:, paras.bptt:paras.bptt+7200].reshape(-1).detach().cpu().numpy()/(3.6*100000)
).sum()
logging('Low: %.2f, Predicted Power: %.3f' % (low, float(sum_power)))
sum_powers.append(sum_power)
#cop
pcooling_power = (t2np(Ytest_pred) * Y_std + Y_mean)[..., 1].reshape(-1)
pcooling_power[pcooling_power <0 ] = 0
sum_pcooling_power = (
pcooling_power[:7200] *
dttest_tn[:, paras.bptt:paras.bptt+7200].reshape(-1).detach().cpu().numpy()/(3.6*100000)
).sum()
logging('Low: %.2f, sum_pcooling_powerr: %.3f' % (low, float(sum_pcooling_power)))
cop = round((sum_pcooling_power/sum_power), 2) #百分比
logging('Low: %.2f, cop: %.3f' % (low, float(cop)))
cops.append(cop)
#pue
pserver = (Xtest*X_std+X_mean)[...,0].reshape(-1)
pserver_power = (pserver[paras.bptt:paras.bptt+7200] *
dttest_tn[:, paras.bptt:paras.bptt+7200].reshape(-1).detach().cpu().numpy()/(3.6*100000)
).sum()
pue = round((1+(sum_power/pserver_power)), 2) #百分比
logging('Low: %.2f, pue: %.3f' % (low, float(pue)))
pues.append(pue)
min_sum_power = min(sum_powers)
ratio_power_decline = ((sum_powers[0]-min_sum_power)/sum_powers[0])*100
logging_all('%s ratio_power_decline: %.2f' % (everdata,ratio_power_decline))
cops_all.append(cops)
pue_all.append(pues)
sum_powers_all.append(sum_powers)
plt.title('all' + " - " + 'Power')
plt.figure(figsize=(8, 5))
color = ['blue', 'red', 'green', 'purple']
line = ['-', '--', ':']
for i, sum_power in enumerate(sum_powers_all):
# plt.title('Compare the power consumption at different lower temperature limits', fontsize=20)
plt.plot(np.arange(Min, Max, step), sum_power, label=paras.datasets2[i], linestyle=line[i], color=color[i])
min_idx = sum_power.index(min(sum_power))
plt.plot(Min + (min_idx * step), sum_power[min_idx], marker='x', color=color[i], markersize='20')
plt.legend(fontsize=18)
plt.text(13.98, 4.35, ' Optimal\nlower threshold', fontsize=16, color='black')
plt.arrow(15.75,4.4,-1.1, -0.55,shape='full',head_width=0.1,head_length=0.2,length_includes_head=True,ec ='black')
plt.arrow(15.75, 4.4, -0.28, 0.8, shape='full', head_width=0.1, head_length=0.2, length_includes_head=True, ec='black')
plt.arrow(15.75,4.4,0.26, 1.3, shape='full',head_width=0.1,head_length=0.2,length_includes_head=True,ec ='black')
plt.xticks(fontsize=19)
plt.yticks(fontsize=19)
plt.xlabel('$Ti_{\min}(℃)$', fontsize=20)
plt.ylabel('Energy consumption(kwh)', fontsize=20)
plt.tight_layout()
plt.savefig(os.path.join(save_dir_img, 'power%.1f-%.1f-%.1f.png' % (Min, Max, step)))
plt.close()
# cop
plt.figure(figsize=(8, 5))
color = ['blue', 'red', 'green', 'purple']
line = ['-', '--', ':']
for i, cops in enumerate(cops_all):
# plt.title('Compare the power consumption at different lower temperature limits', fontsize=20)
plt.plot(np.arange(Min, Max, step), cops, label=paras.datasets2[i], linestyle=line[i], color=color[i])
#plt.legend(fontsize=18)
plt.legend(fontsize=18)
plt.xticks(fontsize=19)
plt.yticks(fontsize=19)
plt.xlabel('$Ti_{\min}(℃)$', fontsize=20)
plt.ylabel('COP', fontsize=20)
plt.tight_layout()
plt.savefig(os.path.join(save_dir_img, 'cop%.1f-%.1f-%.1f.png' % (Min, Max, step)))
plt.close()
# pue
plt.figure(figsize=(8, 5))
color = ['blue', 'red', 'green', 'purple']
line = ['-', '--', ':']
for i, pues in enumerate(pue_all):
# plt.title('Compare the power consumption at different lower temperature limits', fontsize=20)
plt.plot(np.arange(Min, Max, step), pues, label=paras.datasets2[i], linestyle=line[i], color=color[i])
min_idx = pues.index(min(pues))
plt.plot(Min + (min_idx * step), pues[min_idx], marker='x', color=color[i], markersize='20')
plt.legend(fontsize=18)
plt.xticks(fontsize=19)
plt.yticks(fontsize=19)
plt.xlabel('$Ti_{\min}(℃)$', fontsize=20)
plt.ylabel('PUE', fontsize=20)
plt.tight_layout()
plt.savefig(os.path.join(save_dir_img, 'pue%.1f-%.1f-%.1f.png' % (Min, Max, step)))
plt.close()