基于 时间卷积注意神经网络 时间序列预测-光伏功率预测实例.完整项目分享 毕设

## TCAN TensorFlow TensorFlow implementation of multivariate time series forecasting model introduced in Lin, Y., Koprinska, I., & Rana, M. (2021). Temporal Convolutional Attention Neural Networks for Time Series Forecasting. In 2021 International Joint Conference on Neural Networks (IJCNN) (pp. 1-8). IEEE. [https://doi.org/10.1109/IJCNN52387.2021.9534351](https://doi.org/10.1109/IJCNN52387.2021.9534351). ### Dependencies ```bash pandas==1.3.4 numpy==1.19.5 tensorflow==2.7.0 tensorflow_probability==0.14.1 tensorflow_addons==0.15.0 plotly==5.3.1 kaleido==0.2.1 ``` ### Usage ```python import numpy as np from tcan_tensorflow.model import TCAN # Generate two time series N = 1000 t = np.linspace(0, 1, N) e = np.random.multivariate_normal(mean=[0, 0], cov=[[1, 0.25], [0.25, 1]], size=N) a = 40 + 30 * t + 20 * np.cos(2 * np.pi * (20 * t - 0.5)) + e[:, 0] b = 50 - 40 * t + 30 * np.sin(2 * np.pi * (20 * t - 0.5)) + e[:, 1] y = np.hstack([a.reshape(- 1, 1), b.reshape(- 1, 1)]) # Fit the model model = TCAN( y=y, x=None, forecast_period=100, lookback_period=300, quantiles=[0.01, 0.1, 0.5, 0.9, 0.99], filters=2, kernel_size=3, dilation_rates=[1, 2], dropout=0, alpha=1.5 ) model.fit( regularization=0.5, learning_rate=0.01, batch_size=64, epochs=200, verbose=1 ) ``` ```python # Plot the in-sample predictions predictions = model.predict(index=900) fig = model.plot_predictions() fig.write_image('predictions.png', width=750, height=650) ```  ```python # Plot the out-of-sample forecasts forecasts = model.forecast() fig = model.plot_forecasts() fig.write_image('forecasts.png', width=750, height=650) ```