Multivariate Forecasting of Electricity Consumption for Sustainable Energy Planning

Abstract

Accurate electricity consumption forecasting is vital for energy security in a rapidly developing region. Existing research has established the positive correlation between electricity consumption, population, and economic growth. However, only few studies have applied the multivariate model to a regional-provincial context for long-term electricity forecast. This research aims to bridge the gap by providing a reliable forecast to support regional sustainable energy planning in Lampung Province. The methodology includes data preprocessing, integration and cleaning, and model training and validation using time-series data. The Vector Autoregressive (VAR) was employed to predict electricity consumption from 2024 to 2030 based on historical data 2010 – 2023. The model demonstrated a robust predictive performance, with a low MAPE of 0.57%, RMSE of 37.74, and a high R² value of 0.998. This results instilling confidence in the research findings and prospect of utilizing the VAR for electricity forecasting. By using the model, trend of electricity consumption in Lampung Province show the gradually increasing. The study also underscores the importance of integrating renewable energy sources to address future electricity needs sustainably, ensuring energy infrastructure aligns with the socio-economic growth and energy transition agenda in the context of Lampung Province development.

Affiliated Institutions

• Lampung University ID

Related Publications