160 KiB
160 KiB
In [9]:
from math import sqrt from numpy import concatenate from matplotlib import pyplot import pandas as pd import numpy as np from sklearn.preprocessing import MinMaxScaler from sklearn.preprocessing import LabelEncoder from sklearn.metrics import mean_squared_error from tensorflow.keras import Sequential from tensorflow.keras.layers import Dense from tensorflow.keras.layers import LSTM from tensorflow.keras.layers import Dropout from sklearn.model_selection import train_test_split import matplotlib.pyplot as plt
In [10]:
# 加载数据 path1 = r"D:\project\小论文1-基于ICEEMDAN分解的时序高维变化的短期光伏功率预测模型\CEEMAN-PosConv1dbiLSTM-LSTM\模型代码流程\单多步可视化\icial模型\t+1\xin9999低频_forecast(T+1).csv"#数据所在路径 #我的数据是excel表，若是csv文件用pandas的read_csv()函数替换即可。 f_low= pd.DataFrame(pd.read_csv(path1))
In [11]:
# 加载数据 path2 = r"D:\project\小论文1-基于ICEEMDAN分解的时序高维变化的短期光伏功率预测模型\CEEMAN-PosConv1dbiLSTM-LSTM\模型代码流程\单多步可视化\icial模型\t+1\xin99939高频re_forecast(t+1).csv"#数据所在路径 #我的数据是excel表，若是csv文件用pandas的read_csv()函数替换即可。 f_high= pd.DataFrame(pd.read_csv(path2))
In [12]:
path3= r"D:\project\小论文1-基于ICEEMDAN分解的时序高维变化的短期光伏功率预测模型\CEEMAN-PosConv1dbiLSTM-LSTM\模型代码流程\单多步可视化\icial模型\t+1\xin9999低频_test(T+1).csv"#数据所在路径 #我的数据是excel表，若是csv文件用pandas的read_csv()函数替换即可。 true_low= pd.DataFrame(pd.read_csv(path3))
In [13]:
path4= r"D:\project\小论文1-基于ICEEMDAN分解的时序高维变化的短期光伏功率预测模型\CEEMAN-PosConv1dbiLSTM-LSTM\模型代码流程\单多步可视化\icial模型\t+1\xin99939高频re_test(t+1).csv"#数据所在路径 #我的数据是excel表，若是csv文件用pandas的read_csv()函数替换即可。 true_high= pd.DataFrame(pd.read_csv(path4))
In [14]:
path5= r"D:\project\小论文1-基于ICEEMDAN分解的时序高维变化的短期光伏功率预测模型\CEEMAN-PosConv1dbiLSTM-LSTM\模型代码流程\单多步可视化\icial模型\t+1\test.csv"#数据所在路径 #我的数据是excel表，若是csv文件用pandas的read_csv()函数替换即可。 true_2= pd.DataFrame(pd.read_csv(path5))
In [50]:
true_2
Out[50]:
| column_name | |
|---|---|
| 0 | 4.679599 |
| 1 | 4.675801 |
| 2 | 4.636000 |
| 3 | 4.572200 |
| 4 | 4.525266 |
| ... | ... |
| 1557 | 0.000000 |
| 1558 | 0.000000 |
| 1559 | 0.000000 |
| 1560 | 0.000000 |
| 1561 | 0.000000 |
1562 rows × 1 columns
In [15]:
f_low
Out[15]:
| column_name | |
|---|---|
| 0 | 1.681645 |
| 1 | 1.683139 |
| 2 | 1.684575 |
| 3 | 1.685911 |
| 4 | 1.687193 |
| ... | ... |
| 1557 | 1.545752 |
| 1558 | 1.545890 |
| 1559 | 1.546023 |
| 1560 | 1.546150 |
| 1561 | 1.546288 |
1562 rows × 1 columns
In [16]:
pre_data=f_low+f_high pre_data true=true_low+true_high true # df1 = pd.DataFrame(pre_data, columns=['column_name']) # 指定文件路径和文件名，保存DataFrame到CSV文件中 # df1.to_csv('(t+3)经过ICEEMDAN分解预测的预测集.csv', index=False)
Out[16]:
| column_name | |
|---|---|
| 0 | 4.679599e+00 |
| 1 | 4.675801e+00 |
| 2 | 4.636000e+00 |
| 3 | 4.572200e+00 |
| 4 | 4.525266e+00 |
| ... | ... |
| 1557 | 2.220446e-16 |
| 1558 | -4.440892e-16 |
| 1559 | 0.000000e+00 |
| 1560 | 0.000000e+00 |
| 1561 | -4.440892e-16 |
1562 rows × 1 columns
In [18]:
import numpy as np def update_pre_based_on_true(true, pre): # 确保 true 和 pre 是 NumPy 数组 true = np.array(true) pre = np.array(pre) # 使用布尔索引将 pre 中对应位置的值设为0 pre[true == 0] = 0 return pre updated_pre = update_pre_based_on_true(true_2, pre_data) print(updated_pre) df1 = pd.DataFrame(updated_pre, columns=['column_name']) # 指定文件路径和文件名，保存DataFrame到CSV文件中 df1.to_csv('(t+1)经过ICEEMDAN分解预测的预测集.csv', index=False)
[[4.6702959] [4.6353927] [4.6446364] ... [0. ] [0. ] [0. ]]
In [33]:
plt.figure(figsize=(16,8)) plt.plot(true_2, label='true') plt.plot(updated_pre, label='pre') plt.legend() plt.show()
找论文 引用 峰值达不到 偏低 情况 通过分解之后 提高了 加在分析 结论 两个数据集篇幅一样 附录 图 未来数据引入 文章对比 3个亮点
In [24]:
from sklearn.metrics import mean_squared_error, mean_absolute_error # 评价指标 # 使用sklearn调用衡量线性回归的MSE 、 RMSE、 MAE、r2 from math import sqrt from sklearn.metrics import mean_absolute_error from sklearn.metrics import mean_squared_error from sklearn.metrics import r2_score print('mean_squared_error:', mean_squared_error(updated_pre, true)) # mse) print("mean_absolute_error:", mean_absolute_error(updated_pre, true)) # mae print("rmse:", sqrt(mean_squared_error(pre_data, true))) print("r2 score:", r2_score(updated_pre, true))#
mean_squared_error: 0.0018723911403168718 mean_absolute_error: 0.015867955050384196 rmse: 0.04488873526226243 r2 score: 0.9994400158090125
In [45]:
true3=true_2[150:400] pre_data3=updated_pre[150:400]
In [48]:
# 假设true_2和updated_pre是你的NumPy数组 true3 = true_2[150:400] pre_data3 = updated_pre[150:400]
In [51]:
pre_data3
Out[51]:
array([[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0.0163369 ],
[ 0.0251948 ],
[ 0.041127 ],
[ 0.0720103 ],
[ 0.1019873 ],
[ 0.1241176 ],
[ 0.1625308 ],
[ 0.2392393 ],
[ 0.3379075 ],
[ 0.3366777 ],
[ 0.3438205 ],
[ 0.453801 ],
[ 0.733962 ],
[ 1.10062312],
[ 1.1571449 ],
[ 1.10106503],
[ 1.1546293 ],
[ 0.94037127],
[ 1.06687378],
[ 1.01553673],
[ 1.68440343],
[ 1.83468444],
[ 1.47566818],
[ 1.78731345],
[ 1.71659963],
[ 2.22625894],
[ 1.64490715],
[ 2.2160093 ],
[ 2.7114284 ],
[ 2.5523071 ],
[ 2.5810688 ],
[ 2.9307843 ],
[ 2.941154 ],
[ 3.1618417 ],
[ 3.3204204 ],
[ 3.358411 ],
[ 3.4240855 ],
[ 3.5163293 ],
[ 3.6084709 ],
[ 3.6867704 ],
[ 3.752888 ],
[ 3.8386279 ],
[ 3.9089437 ],
[ 3.9630318 ],
[ 4.0262716 ],
[ 4.0721401 ],
[ 4.1171014 ],
[ 4.1841362 ],
[ 4.238177 ],
[ 4.2544167 ],
[ 4.3330357 ],
[ 4.3175416 ],
[ 4.4193332 ],
[ 4.4418901 ],
[ 4.3219046 ],
[ 4.4783611 ],
[ 4.4945666 ],
[ 4.5501185 ],
[ 4.5723112 ],
[ 4.6125982 ],
[ 4.6380826 ],
[ 4.6540792 ],
[ 4.6313653 ],
[ 4.6655451 ],
[ 4.6778368 ],
[ 4.7202235 ],
[ 4.730606 ],
[ 4.7420423 ],
[ 4.6920747 ],
[ 4.7461967 ],
[ 4.720302 ],
[ 4.7038498 ],
[ 4.6931526 ],
[ 4.6983388 ],
[ 4.7022051 ],
[ 4.7118301 ],
[ 4.7128725 ],
[ 4.6984447 ],
[ 4.7051 ],
[ 4.6634191 ],
[ 4.6624637 ],
[ 4.660742 ],
[ 4.6593221 ],
[ 4.6186729 ],
[ 4.548237 ],
[ 4.5489672 ],
[ 4.5194306 ],
[ 4.47022 ],
[ 4.490508 ],
[ 4.4213457 ],
[ 4.4191448 ],
[ 4.3844216 ],
[ 4.3697215 ],
[ 4.2856085 ],
[ 4.2563456 ],
[ 4.1913535 ],
[ 4.1107695 ],
[ 4.0634076 ],
[ 4.0212931 ],
[ 3.9695915 ],
[ 3.8896301 ],
[ 3.8627817 ],
[ 3.7772458 ],
[ 3.6876558 ],
[ 3.6219018 ],
[ 3.5707506 ],
[ 3.4685108 ],
[ 3.3918019 ],
[ 3.3126507 ],
[ 3.2466422 ],
[ 3.1425753 ],
[ 3.0552442 ],
[ 2.9830179 ],
[ 2.8893795 ],
[ 2.7793767 ],
[ 2.6845963 ],
[ 2.6033527 ],
[ 2.4811151 ],
[ 2.396697 ],
[ 2.28874156],
[ 2.1929247 ],
[ 2.097524 ],
[ 1.97145515],
[ 1.8743583 ],
[ 1.7611796 ],
[ 1.6390783 ],
[ 1.50743994],
[ 1.39524042],
[ 1.25902684],
[ 1.1469124 ],
[ 1.02292133],
[ 0.91121595],
[ 0.79405735],
[ 0.68875556],
[ 0.56993494],
[ 0.4554422 ],
[ 0.3385904 ],
[ 0.2156578 ],
[ 0.0804994 ],
[ 0.0229209 ],
[ 0.0284618 ],
[ 0.0292833 ],
[ 0.0188559 ],
[ 0.0067573 ],
[-0.0049999 ],
[-0.0103893 ],
[-0.0064329 ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ],
[ 0. ]])
In [56]:
# 假设df是你的DataFrame true31 = true3['column_name'].to_numpy() true31
Out[56]:
array([0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 7.06666700e-03, 2.01333300e-02,
3.87333260e-02, 5.92666570e-02, 7.54666780e-02, 1.03399977e-01,
1.54066697e-01, 2.47066677e-01, 2.42999941e-01, 2.61666685e-01,
3.75400007e-01, 6.25133395e-01, 1.01493323e+00, 1.13400018e+00,
9.97799873e-01, 9.71266568e-01, 8.37799966e-01, 9.07066643e-01,
9.70266700e-01, 1.55040002e+00, 1.84593356e+00, 1.38419986e+00,
1.50386667e+00, 1.67359996e+00, 2.15926647e+00, 1.63333345e+00,
2.23159981e+00, 2.48473311e+00, 2.52906680e+00, 2.43453336e+00,
3.03593326e+00, 2.84746695e+00, 3.17600060e+00, 3.19239974e+00,
3.26446676e+00, 3.35959959e+00, 3.45073366e+00, 3.53786659e+00,
3.60366631e+00, 3.66599941e+00, 3.75313353e+00, 3.82739925e+00,
3.88766718e+00, 3.95066643e+00, 4.00006676e+00, 4.04713345e+00,
4.10600042e+00, 4.18293333e+00, 4.21853399e+00, 4.25460005e+00,
4.23586655e+00, 4.33613300e+00, 4.38013363e+00, 4.21306658e+00,
4.37960052e+00, 4.39233303e+00, 4.51066637e+00, 4.54153299e+00,
4.62066650e+00, 4.63080025e+00, 4.67740011e+00, 4.64033365e+00,
4.65366697e+00, 4.68333292e+00, 4.72733355e+00, 4.75113344e+00,
4.75366688e+00, 4.70033312e+00, 4.73833370e+00, 4.76153374e+00,
4.72873306e+00, 4.71820068e+00, 4.72226620e+00, 4.73426676e+00,
4.74899960e+00, 4.73766708e+00, 4.69826698e+00, 4.72093391e+00,
4.67413330e+00, 4.65986633e+00, 4.65800047e+00, 4.63566637e+00,
4.60559940e+00, 4.53426600e+00, 4.50513315e+00, 4.49013329e+00,
4.41386700e+00, 4.45499992e+00, 4.37726641e+00, 4.41346645e+00,
4.34786653e+00, 4.34599972e+00, 4.29546642e+00, 4.27073288e+00,
4.22719955e+00, 4.16773415e+00, 4.10826588e+00, 4.07379961e+00,
4.01966667e+00, 3.95240021e+00, 3.91913366e+00, 3.82226634e+00,
3.74139977e+00, 3.67379951e+00, 3.63913369e+00, 3.54620004e+00,
3.46993303e+00, 3.41440010e+00, 3.34759975e+00, 3.23346663e+00,
3.16666675e+00, 3.09766650e+00, 3.01139998e+00, 2.92106676e+00,
2.82239986e+00, 2.73320007e+00, 2.61460018e+00, 2.51999998e+00,
2.42013359e+00, 2.31419992e+00, 2.20986652e+00, 2.09686661e+00,
1.99059999e+00, 1.88639998e+00, 1.77793360e+00, 1.64793324e+00,
1.54139996e+00, 1.42546666e+00, 1.30626667e+00, 1.19073319e+00,
1.07273352e+00, 9.58733380e-01, 8.42133343e-01, 7.15199947e-01,
5.81333339e-01, 4.45266664e-01, 3.14866692e-01, 2.03333318e-01,
1.25466660e-01, 8.83999990e-02, 6.87333350e-02, 4.97333260e-02,
3.25333250e-02, 1.56666660e-02, 2.40000000e-03, 1.33333000e-04,
6.67000000e-05, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00,
0.00000000e+00, 0.00000000e+00])
In [60]:
plt.figure(figsize=(6,6)) plt.plot(true31, label='true') plt.plot(pre_data3, label='pre') plt.legend() plt.show()
In [61]:
df1 = pd.DataFrame(updated_pre[150:400], columns=['column_name']) # 指定文件路径和文件名，保存DataFrame到CSV文件中 df1.to_csv('1天的经过ICEEMDAN分解预测的预测集1.csv', index=False)
In [62]:
df2 = pd.DataFrame(true[150:400], columns=['column_name']) # 指定文件路径和文件名，保存DataFrame到CSV文件中 df2.to_csv('1天的经过ICEEMDAN分解预测的真实集.csv', index=False)
In [36]:
# 使用MinMaxScaler进行归一化 from sklearn.preprocessing import MinMaxScaler scaler = MinMaxScaler(feature_range=(0, 1)) pre = scaler.fit_transform(updated_pre) print(pre.shape)
(10415, 1)
In [37]:
from sklearn.preprocessing import MinMaxScaler scaler = MinMaxScaler(feature_range=(0, 1)) true_data = scaler.fit_transform(true) print(true_data.shape)
(10415, 1)
In [38]:
from sklearn.metrics import mean_squared_error, mean_absolute_error # 评价指标 # 使用sklearn调用衡量线性回归的MSE 、 RMSE、 MAE、r2 from math import sqrt from sklearn.metrics import mean_absolute_error from sklearn.metrics import mean_squared_error from sklearn.metrics import r2_score print('mean_squared_error:', mean_squared_error(updated_pre, true)) # mse) print("mean_absolute_error:", mean_absolute_error(updated_pre, true)) # mae print("rmse:", sqrt(mean_squared_error(pre_data, true))) print("r2 score:", r2_score(pre[900:2100], true_data[900:2100]))#
mean_squared_error: 0.04675641413227651 mean_absolute_error: 0.0798491015148862 rmse: 0.21689357303163628 r2 score: 0.9912435196234671
In [ ]: