唐山地方志网站建设,杭州网站建设专业公司,wordpress采集商品,深圳做关键词优化平台#x1f368; 本文为#x1f517;365天深度学习训练营 中的学习记录博客#x1f356; 原作者#xff1a;K同学啊 目标
复用LSTM模型实现火灾温度预测
具体实现
#xff08;一#xff09;环境
语言环境#xff1a;Python 3.10 编 译 器: PyCharm 框 架: Pytorch 本文为365天深度学习训练营 中的学习记录博客 原作者K同学啊 目标
复用LSTM模型实现火灾温度预测
具体实现
一环境
语言环境Python 3.10 编 译 器: PyCharm 框 架: Pytorch
二具体步骤
1. 具体代码
import torch.nn.functional as F
import numpy as np
import pandas as pd
import torch
from torch import nn import matplotlib.pyplot as plt
import seaborn as sns device torch.device(cuda if torch.cuda.is_available() else cpu)
print(device) data pd.read_csv(data/woodpine2.csv)
print(data) plt.rcParams[savefig.dpi] 500
plt.rcParams[figure.dpi] 500 fig, ax plt.subplots(1, 3, constrained_layoutTrue, figsize(14, 3)) sns.lineplot(datadata[Tem1], axax[0])
sns.lineplot(datadata[CO 1], axax[1])
sns.lineplot(datadata[Soot 1], axax[2])
plt.show() from sklearn.preprocessing import MinMaxScaler
# 从原始数据中去除第一列通常为索引列保留所有行和其他列并创建数据副本
dataFrame data.iloc[:,1:].copy()
# 将数据归一化范围是0-1
sc MinMaxScaler(feature_range(0, 1))
for i in [CO 1, Soot 1, Tem1]: dataFrame[i] sc.fit_transform(dataFrame[i].values.reshape(-1, 1))
print(dataFrame)
print(dataFrame.shape) width_X 8
width_y 1 X []
y [] print(f原始数据长度: {len(dataFrame)})
print(f可创建的序列数量: {len(dataFrame) - width_X}) # 数据窗口划分主循环
# 功能将时间序列数据划分为输入样本X和对应标签y
# 参数说明
# data: 原始数据DataFrame
# dataFrame: 转换后的numpy数组
# width_X: 输入序列长度
# width_y: 输出序列长度
# 返回值
# X: 输入样本列表
# y: 标签列表
# 窗口滑动逻辑
# 1. 计算当前窗口结束索引
# 2. 检查窗口是否超出数据边界
# 3. 合法窗口则提取对应数据片段
# 4. 窗口起始位置逐步后移
in_start 0
for _, _ in data.iterrows(): in_end in_start width_X out_end in_end width_y if out_end len(dataFrame): X_ np.array(dataFrame.iloc[in_start:in_end, ]) y_ np.array(dataFrame.iloc[in_end:out_end, 0]) X.append(X_) y.append(y_) in_start 1 X np.array(X)
y np.array(y).reshape(-1, 1) print(X.shape)
print(y.shape) print(np.any(np.isnan(X)))
print(np.any(np.isnan(y))) max_samples min(5000, len(X))
X_train torch.tensor(np.array(X[:max_samples]), dtypetorch.float32).to(device)
y_train torch.tensor(np.array(y[:max_samples]), dtypetorch.float32).to(device)
X_test torch.tensor(np.array(X[max_samples:]), dtypetorch.float32).to(device)
y_test torch.tensor(np.array(y[max_samples:]), dtypetorch.float32).to(device)
print(训练集大小:, X_train.shape, y_train.shape)
print(测试集大小:, X_test.shape, y_test.shape) from torch.utils.data import TensorDataset, DataLoader
train_dl DataLoader(TensorDataset(X_train, y_train), batch_size64, shuffleFalse)
test_dl DataLoader(TensorDataset(X_test, y_test), batch_size64, shuffleFalse) class model_lstm(nn.Module): LSTM神经网络模型包含两个LSTM层和一个全连接层。 参数: input_size (int): 输入特征的维度默认为3CO、烟雾、温度 hidden_size (int): 隐藏层维度默认为320 num_layers (int): LSTM层数默认为1 batch_first (bool): 输入数据是否以batch为第一维度默认True 返回值: torch.Tensor: 形状为(batch_size, 1)的预测结果 def __init__(self): super(model_lstm, self).__init__() self.lstm0 nn.LSTM(input_size3, hidden_size320, num_layers1, batch_firstTrue) self.lstm1 nn.LSTM(input_size320, hidden_size320, num_layers1, batch_firstTrue) self.fc0 nn.Linear(320, 1) def forward(self, x): 前向传播函数。 参数: x (torch.Tensor): 输入张量形状为(batch_size, seq_length, input_size) 返回值: torch.Tensor: 输出张量形状为(batch_size, 1) out, hidden1 self.lstm0(x) out, _ self.lstm1(out, hidden1) out self.fc0(out) return out[:, -1, :] model model_lstm()
print(model) print(model(torch.rand(30, 8, 3)).shape) import copy
def train(train_dl, model, loss_fn, opt, lr_schedulerNone): 训练模型单个epoch。 参数: train_dl (DataLoader): 训练数据加载器 model (nn.Module): 神经网络模型 loss_fn (nn.Module): 损失函数 opt (Optimizer): 优化器 lr_scheduler (LRScheduler): 学习率调度器可选 返回值: float: 当前epoch的平均训练损失 size len(train_dl.dataset) num_batches len(train_dl) train_loss 0 for x, y in train_dl: x, y x.to(device), y.to(device) pred model(x) loss loss_fn(pred, y) opt.zero_grad() loss.backward() opt.step() train_loss loss.item() if lr_scheduler is not None: lr_scheduler.step() print(当前学习率:, lr_scheduler.get_last_lr()) train_loss / num_batches return train_loss def test(dataloader, model, loss_fn): 测试模型性能。 参数: dataloader (DataLoader): 测试数据加载器 model (nn.Module): 训练好的神经网络模型 loss_fn (nn.Module): 损失函数 返回值: float: 测试集的平均损失 size len(dataloader.dataset) num_batches len(dataloader) test_loss 0 with torch.no_grad(): for x, y in dataloader: x, y x.to(device), y.to(device) pred model(x) loss loss_fn(pred, y) test_loss loss.item() test_loss / num_batches return test_loss # 训练
model model_lstm().to(device)
loss_fn nn.MSELoss()
learn_rate 1e-1
opt torch.optim.SGD(model.parameters(), lrlearn_rate, weight_decay1e-4)
epochs 50
train_loss []
test_loss []
lr_scheduler torch.optim.lr_scheduler.CosineAnnealingLR(opt, T_maxepochs, last_epoch-1) for epoch in range(epochs): model.train() epoch_train_loss train(train_dl, model, loss_fn, opt, lr_scheduler) model.eval() epoch_test_loss test(test_dl, model, loss_fn) train_loss.append(epoch_train_loss) test_loss.append(epoch_test_loss) template Epoch:{:2d}, Train_loss:{:.3f}, Test_loss:{:.3f} print(template.format(epoch 1, epoch_train_loss, epoch_test_loss)) print(*20, Done, *20) from datetime import datetime
current_time datetime.now()
plt.figure(figsize(5, 3), dpi120)
plt.plot(train_loss, labelLSTM Training Loss)
plt.plot(test_loss, labelLSTM Testing Loss)
plt.title(Training and Valiadation Loss)
plt.xlabel(current_time)
plt.legend()
plt.show() # 调用模型进行预测
model.eval()
with torch.no_grad(): predicted_y_lstm sc.inverse_transform(model(X_test).cpu().detach().numpy().reshape(-1, 1))
y_test_1 sc.inverse_transform(y_test.cpu().reshape(-1, 1))
y_test_one [i[0] for i in y_test_1]
predicted_y_lstm_one [i[0] for i in predicted_y_lstm] plt.figure(figsize(10, 5), dpi120)
plt.plot(y_test_one[:2000], colorred, labelread_temp)
plt.plot(predicted_y_lstm_one[:2000], colorblue, labelpredict)
plt.title(Title)
plt.xlabel(X)
plt.ylabel(Y)
plt.legend()
plt.show()2. 结果
C:\Users\feng_\anaconda3\envs\Pytorch\python.exe E:\dev\AI\Pytorch\RNN\R4-LSTM火灾温度预测\LSTM-HUOZAI.py
cudaTime Tem1 CO 1 Soot 1
0 0.000 25.0 0.000000 0.000000
1 0.228 25.0 0.000000 0.000000
2 0.456 25.0 0.000000 0.000000
3 0.685 25.0 0.000000 0.000000
4 0.913 25.0 0.000000 0.000000
... ... ... ... ...
5943 366.000 295.0 0.000077 0.000496
5944 366.000 294.0 0.000077 0.000494
5945 367.000 292.0 0.000077 0.000491
5946 367.000 291.0 0.000076 0.000489
5947 367.000 290.0 0.000076 0.000487[5948 rows x 4 columns]Tem1 CO 1 Soot 1
0 0.000000 0.000000 0.000000
1 0.000000 0.000000 0.000000
2 0.000000 0.000000 0.000000
3 0.000000 0.000000 0.000000
4 0.000000 0.000000 0.000000
... ... ... ...
5943 0.957447 0.968672 0.968750
5944 0.953901 0.963659 0.964844
5945 0.946809 0.958647 0.958984
5946 0.943262 0.954887 0.955078
5947 0.939716 0.951128 0.951172[5948 rows x 3 columns]
(5948, 3)
原始数据长度: 5948
可创建的序列数量: 5940
(5939, 8, 3)
(5939, 1)
False
False
训练集大小: torch.Size([5000, 8, 3]) torch.Size([5000, 1])
测试集大小: torch.Size([939, 8, 3]) torch.Size([939, 1])
model_lstm((lstm0): LSTM(3, 320, batch_firstTrue)(lstm1): LSTM(320, 320, batch_firstTrue)(fc0): Linear(in_features320, out_features1, biasTrue)
)
torch.Size([30, 1])
当前学习率: [0.09990133642141358]
Epoch: 1, Train_loss:0.001, Test_loss:0.012
当前学习率: [0.0996057350657239]
Epoch: 2, Train_loss:0.014, Test_loss:0.012
当前学习率: [0.09911436253643444]
Epoch: 3, Train_loss:0.014, Test_loss:0.011
当前学习率: [0.09842915805643154]
Epoch: 4, Train_loss:0.013, Test_loss:0.011
当前学习率: [0.09755282581475767]
Epoch: 5, Train_loss:0.013, Test_loss:0.010
当前学习率: [0.09648882429441256]
Epoch: 6, Train_loss:0.012, Test_loss:0.010
当前学习率: [0.09524135262330098]
Epoch: 7, Train_loss:0.012, Test_loss:0.009
当前学习率: [0.09381533400219318]
Epoch: 8, Train_loss:0.011, Test_loss:0.009
当前学习率: [0.09221639627510075]
Epoch: 9, Train_loss:0.010, Test_loss:0.008
当前学习率: [0.09045084971874737]
Epoch:10, Train_loss:0.009, Test_loss:0.007
当前学习率: [0.08852566213878946]
Epoch:11, Train_loss:0.008, Test_loss:0.006
当前学习率: [0.08644843137107057]
Epoch:12, Train_loss:0.007, Test_loss:0.005
当前学习率: [0.08422735529643442]
Epoch:13, Train_loss:0.006, Test_loss:0.005
当前学习率: [0.08187119948743447]
Epoch:14, Train_loss:0.005, Test_loss:0.004
当前学习率: [0.07938926261462366]
Epoch:15, Train_loss:0.004, Test_loss:0.003
当前学习率: [0.07679133974894982]
Epoch:16, Train_loss:0.003, Test_loss:0.003
当前学习率: [0.07408768370508576]
Epoch:17, Train_loss:0.002, Test_loss:0.002
当前学习率: [0.07128896457825362]
Epoch:18, Train_loss:0.002, Test_loss:0.002
当前学习率: [0.06840622763423389]
Epoch:19, Train_loss:0.001, Test_loss:0.001
当前学习率: [0.06545084971874736]
Epoch:20, Train_loss:0.001, Test_loss:0.001
当前学习率: [0.06243449435824272]
Epoch:21, Train_loss:0.001, Test_loss:0.001
当前学习率: [0.05936906572928623]
Epoch:22, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.056266661678215216]
Epoch:23, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.053139525976465665]
Epoch:24, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.049999999999999996]
Epoch:25, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.046860474023534326]
Epoch:26, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.04373333832178478]
Epoch:27, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.040630934270713764]
Epoch:28, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.03756550564175726]
Epoch:29, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.03454915028125265]
Epoch:30, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.03159377236576612]
Epoch:31, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.028711035421746366]
Epoch:32, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.025912316294914226]
Epoch:33, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.023208660251050155]
Epoch:34, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.020610737385376346]
Epoch:35, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.01812880051256551]
Epoch:36, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.015772644703565562]
Epoch:37, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.013551568628929433]
Epoch:38, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.011474337861210542]
Epoch:39, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.009549150281252632]
Epoch:40, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.007783603724899257]
Epoch:41, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.0061846659978068205]
Epoch:42, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.004758647376699033]
Epoch:43, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.0035111757055874327]
Epoch:44, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.0024471741852423235]
Epoch:45, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.0015708419435684518]
Epoch:46, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.000885637463565564]
Epoch:47, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.00039426493427611173]
Epoch:48, Train_loss:0.000, Test_loss:0.001
当前学习率: [9.866357858642205e-05]
Epoch:49, Train_loss:0.000, Test_loss:0.001
当前学习率: [0.0]
Epoch:50, Train_loss:0.000, Test_loss:0.001Done 三总结
数据处理采用了滑动窗口技术将时间序列划分为长度为8的输入序列和长度为1的预测目标多变量输入 同时使用温度(Tem1)、CO浓度(CO 1)和烟雾浓度(Soot 1)三个特征双层LSTM结构: 两个320隐藏单元的LSTM层串联增强模型的时序建模能力单一全连接层直接输出预测结果
