A Crop Growth Prediction Model Using Energy Data Based on Machine Learning in Smart Farms

Area of Science:

• Agricultural Science
• Data Science
• Artificial Intelligence

Background:

• Smart farms utilize data analysis and AI, but high operating costs stem from inefficient energy use.
• Accurate estimation of agricultural energy usage and environmental factors is crucial for crop growth control in smart farms.
• Crop growth sequences are directly linked to energy usage and consumption in smart farm environments.

Purpose of the Study:

• To develop and validate a machine learning algorithm for interpreting crop growth rate responses to environmental and solar energy factors.
• To evaluate the developed algorithm's accuracy against a baseline model.
• To identify key growth and environmental factors influencing paprika crop development.

Main Methods:

• Comparative experiment using three machine learning techniques: Random Forest (RF), Support Vector Machine (SVM), and Gradient Boosting Machine (GBM).
• Focus on energy usage for environmental control, specifically its association with paprika crop growth.
• Utilized real-world data from a paprika smart farm in South Korea.

Main Results:

• The multi-level Random Forest (RF) model achieved an accuracy of 0.88 in predicting paprika growth.
• The model effectively analyzed data related to solar energy factors.
• Identified key growth factors (leaf length, leaf width) and environmental factors influencing crop development.

Conclusions:

• The proposed machine learning algorithm, particularly the multi-level RF, accurately predicts paprika growth in smart farms.
• The algorithm's ability to analyze environmental and solar energy data contributes to efficient energy usage and cost reduction.
• This approach can be extended for big data analysis of crop growth in various smart farm settings.