A TIME-SERIES FORECASTING OF POWER CONSUMPTION AND FEATURE EXTRACTION IN AGRICULTURE SECTOR USING MACHINE LEARNING, 1-11.

Megha Sharma, Namita Mittal, Anukram Mishra, and Arun Gupta

Keywords

Agriculture sector, electrical power, time series-based load forecasting, XGBoost

Abstract

Energy plays a vital role in the economic development, growth, and productivity of any country. Commercial, industrial, residential, and agriculture sectors require reliable and adequate energy services. The paper emphasises the significance of accurate electric power forecasting in the agriculture sector in India to avoid power outages and impact on productivity. The paper aims to forecast medium- term load using time series models and identify important features for power prediction in agriculture. The time series model is used to find the peak season in the year, while the XGBoost technique is used to identify the feature importance and load forecasting. Statistical approaches, such as correlation matrix and scatter plots, are also used for feature extraction. The results of the study show that the addition of exogenous and endogenous data on the historical load improves the accuracy in terms of mean absolute percentage error (MAPE) and R-squared (R2). The research demonstrates the potential of using machine learning techniques to enhance the accuracy of medium-term agricultural electrical load forecasting.

Important Links:

Go Back