⚡ Power Plant Energy Output Predictor (PyTorch)

This project implements an Artificial Neural Network (ANN) using PyTorch to predict the net hourly electrical energy output ($EP$) of a Combined Cycle Power Plant. By analyzing ambient environmental factors, the model provides high-precision regression for energy forecasting.

---

🚀 Features

• PyTorch Core: Built using torch.nn for modular and efficient deep learning.
• Deep Regression: Designed to capture non-linear relationships between weather data and power output.
• Standardized Pipeline: Includes full data preprocessing with StandardScaler for optimized neural network training.
• Performance Tracking: Uses Mean Squared Error (MSE) to monitor and minimize prediction variance.

---

📊 Model Methodology

• Data Preprocessing: Features ($T$, $V$, $AP$, $RH$) are scaled to a standard normal distribution before being converted into PyTorch tensors.
• ANN Architecture: * Input Layer: 4 neurons receiving environmental data.
  • Hidden Layers: Two dense layers (nn.Linear) with 6 neurons each, utilizing ReLU activation for non-linearity.
  • Output Layer: 1 neuron providing the continuous predicted value for $EP$.
• Optimization: The model is trained using the Adam optimizer to find the global minimum of the loss function efficiently.

---

🛠 Tech Stack

• Language: Python
• Framework: PyTorch (torch, torch.nn)
• Libraries: Scikit-Learn, Pandas, NumPy
• Environment: Jupyter Notebook / Anaconda

---

📈 Business Impact

• Grid Efficiency: Enables more accurate energy supply forecasting for grid operators.
• Cost Reduction: Minimizes operational overhead by automating output estimations.
• Environmental Adaptation: Helps plants adjust operations based on fluctuating ambient temperature and pressure.

---

📁 Project Structure

ANN_Regression/
├── regression.ipynb       # PyTorch Model training and evaluation
├── app.py                 # Application script
├── prediction.py          # Prediction utilities
├── requirements.txt       # Python dependencies
└── README.md              # Project documentation