PyTorch搭建ANN实现时间序列预测（风速预测）

数据集

数据集为Barcelona某段时间内的气象数据，其中包括温度、湿度以及风速等。本文将简单搭建来对风速进行预测。

特征构造

对于风速的预测，除了考虑历史风速数据外，还应该充分考虑其余气象因素的影响。因此，我们根据前24个时刻的风速+下一时刻的其余气象数据来预测下一时刻的风速。

数据处理

1.数据预处理

数据预处理阶段，主要将某些列上的文本数据转为数值型数据，同时对原始数据进行归一化处理。文本数据如下所示：

经过转换后，上述各个类别分别被赋予不同的数值，比如"sky is clear"为0，"few clouds"为1。

def load_data():
    global Max, Min
    df = pd.read_csv('Barcelona/Barcelona.csv')
    df.drop_duplicates(subset=[df.columns[0]], inplace=True)
    # weather_main
    listType = df['weather_main'].unique()
    df.fillna(method='ffill', inplace=True)
    dic = dict.fromkeys(listType)
    for i in range(len(listType)):
        dic[listType[i]] = i
    df['weather_main'] = df['weather_main'].map(dic)
    # weather_description
    listType = df['weather_description'].unique()
    dic = dict.fromkeys(listType)
    for i in range(len(listType)):
        dic[listType[i]] = i
    df['weather_description'] = df['weather_description'].map(dic)
    # weather_icon
    listType = df['weather_icon'].unique()
    dic = dict.fromkeys(listType)
    for i in range(len(listType)):
        dic[listType[i]] = i
    df['weather_icon'] = df['weather_icon'].map(dic)
    # print(df)

    columns = df.columns
    Max = np.max(df['wind_speed'])  # 归一化
    Min = np.min(df['wind_speed'])

    for i in range(2, 17):
        column = columns[i]
        if column == 'wind_speed':
            continue
        df[column] = df[column].astype('float64')
        if len(df[df[column] == 0]) == len(df):  # 全0
            continue
        mx = np.max(df[column])
        mn = np.min(df[column])
        df[column] = (df[column] - mn) / (mx - mn)
    # print(df.isna().sum())
    return df

2.数据集构造

利用当前时刻的气象数据和前24个小时的风速数据来预测当前时刻的风速：

def nn_seq():
    """
    :param flag:
    :param data: 待处理的数据
    :return: X和Y两个数据集,X=[当前时刻的year，month, hour, day, lowtemp, hightemp, 前一天当前时刻的负荷以及前23小时负荷]
                              Y=[当前时刻负荷]
    """
    print('处理数据:')
    data = load_data()
    speed = data['wind_speed']
    speed = speed.tolist()
    speed = torch.FloatTensor(speed).view(-1)
    data = data.values.tolist()
    seq = []
    for i in range(len(data) - 30):
        train_seq = []
        train_label = []
        for j in range(i, i + 24):
            train_seq.append(speed[j])
        # 添加温度、湿度、气压等信息
        for c in range(2, 7):
            train_seq.append(data[i + 24][c])
        for c in range(8, 17):
            train_seq.append(data[i + 24][c])
        train_label.append(speed[i + 24])
        train_seq = torch.FloatTensor(train_seq).view(-1)
        train_label = torch.FloatTensor(train_label).view(-1)
        seq.append((train_seq, train_label))
    # print(seq[:5])

    Dtr = seq[0:int(len(seq) * 0.5)]
    Den = seq[int(len(seq) * 0.50):int(len(seq) * 0.75)]
    Dte = seq[int(len(seq) * 0.75):len(seq)]

    return Dtr, Den, Dte

任意输出其中一条数据：

(tensor([1.0000e+00, 1.0000e+00, 2.0000e+00, 1.0000e+00, 1.0000e+00, 1.0000e+00,
        1.0000e+00, 1.0000e+00, 0.0000e+00, 1.0000e+00, 5.0000e+00, 0.0000e+00,
        2.0000e+00, 0.0000e+00, 0.0000e+00, 5.0000e+00, 0.0000e+00, 2.0000e+00,
        2.0000e+00, 5.0000e+00, 6.0000e+00, 5.0000e+00, 5.0000e+00, 5.0000e+00,
        5.3102e-01, 5.5466e-01, 4.6885e-01, 1.0066e-03, 5.8000e-01, 6.6667e-01,
        0.0000e+00, 0.0000e+00, 0.0000e+00, 0.0000e+00, 9.9338e-01, 0.0000e+00,
        0.0000e+00, 0.0000e+00]), tensor([5.]))

数据被划分为三部分：Dtr、Den以及Dte，Dtr用作训练集，Dte用作测试集。

ANN模型

1.模型训练

ANN模型搭建如下：

def ANN():
    Dtr, Den, Dte = nn_seq()
    my_nn = torch.nn.Sequential(
        torch.nn.Linear(38, 64),
        torch.nn.ReLU(),
        torch.nn.Linear(64, 128),
        torch.nn.ReLU(),
        torch.nn.Linear(128, 1),
    )
    model = my_nn.to(device)
    loss_function = nn.MSELoss().to(device)
    optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
    train_inout_seq = Dtr
    # 训练
    epochs = 50
    for i in range(epochs):
        print('当前', i)
        for seq, labels in train_inout_seq:
            seq = seq.to(device)
            labels = labels.to(device)
            y_pred = model(seq)
            single_loss = loss_function(y_pred, labels)
            optimizer.zero_grad()
            single_loss.backward()
            optimizer.step()
        # if i % 2 == 1:
        print(f'epoch: {i:3} loss: {single_loss.item():10.8f}')

    print(f'epoch: {i:3} loss: {single_loss.item():10.10f}')
    state = {'model': model.state_dict(), 'optimizer': optimizer.state_dict(), 'epoch': epochs}
    torch.save(state, 'Barcelona' + ANN_PATH)

可以看到，模型定义的代码段为：

my_nn = torch.nn.Sequential(
    torch.nn.Linear(38, 64),
    torch.nn.ReLU(),
    torch.nn.Linear(64, 128),
    torch.nn.ReLU(),
    torch.nn.Linear(128, 1),
)

第一层全连接层输入维度为38（前24小时风速+14种气象数据），输出维度为64；第二层输入为64，输出128；第三层输入为128，输出为1。

2.模型预测及表现

def ANN_predict(ann, test_seq):
    pred = []
    for seq, labels in test_seq:
        seq = seq.to(device)
        with torch.no_grad():
            pred.append(ann(seq).item())
    pred = np.array([pred])

    return pred

测试：

def test():
    Dtr, Den, Dte = nn_seq()
    ann = torch.nn.Sequential(
        torch.nn.Linear(38, 64),
        torch.nn.ReLU(),
        torch.nn.Linear(64, 128),
        torch.nn.ReLU(),
        torch.nn.Linear(128, 1),
    )
    ann = ann.to(device)
    ann.load_state_dict(torch.load('Barcelona' + ANN_PATH)['model'])
    ann.eval()
    pred = ANN_predict(ann, Dte)
    print(mean_absolute_error(te_y, pred2.T), np.sqrt(mean_squared_error(te_y, pred2.T)))

ANN在Dte上的表现如下表所示：

MAE	RMSE
1.04	1.46