Convolutional Neural Network (CNN) for Wind Speed Forecasting

This repository contains scripts and files for wind speed forecasting using a Convolutional Neural Network (CNN). The CNN model is trained on GRIB data and can generate iterative wind speed predictions.

CNN Model

The CNN model architecture is defined in the model.py file. The model uses a series of convolutional layers with periodic convolutions to capture spatial dependencies in the wind data.

Training

The training script (train.py) processes GRIB data, splits it into training, validation, and test sets, and trains the CNN model. The trained model is then saved to the specified output directory (MODEL_DIR in train_config.py).

Configuration

Before training, configure the necessary parameters in train_config.py, such as the path to the GRIB file (GRIB_FILE), lead time (LEAD_TIME), and batch size (BATCH_SIZE).

Inference

The inference script (infer.py) loads the trained model and performs wind forecasting on specified designated test data. The predictions are visualized and saved as intensity plots in the specified output directory (OUT_DIR in train_config.py).

Directory Structure

• data_transform.py: Contains functions for processing and transforming GRIB data.
• infer.py: Script for wind forecasting inference using the trained CNN model.
• infer_visualisation.py: Visualization functions for wind speed intensity.
• model.py: Defines the CNN model architecture.
• train.py: Script for training the CNN model on GRIB data.
• train_config.py: Configuration file for training parameters.

Usage

1. Install Dependencies:

   pip install -r requirements.txt
   

2. Execute Containerised:

   CMD ["sh", "-c", "python train.py && python infer.py"]