Toll Station Electricity Consumption Forecasting and Climate Classification

This project integrates meteorological analysis and machine learning to estimate monthly electricity consumption of highway toll stations in Taiwan, using geolocation, infrastructure data, and inferred climate zones. It also includes tools for climate classification based on historical temperature data.

---

Objectives

1. Predict Monthly Electricity Consumption (in kWh) of toll stations using:

• Location (latitude and longitude)

• Lane count

• Month

• Inferred climate type

2. Classify Regional Climate based on historical average monthly temperatures using:

• KMeans clustering

• Multinomial logistic regression for generalization

3. Aggregate and process raw electricity usage and temperature sensor data from CSV logs.

---

Power Consumption Model Performance

---

Requirements

• Python 3.8+
• pandas, numpy, scikit-learn, joblib

Notes

Project covers data from June 2024 to May 2025.