Article Details

Abstract

In recent years, the development of smart laboratories has been very rapid, with an increasing variety and scale of equipment in the laboratory, leading to increasingly prominent energy consumption issues. This study takes the laboratory of Zhengzhou Aviation Institute as the research object, exploring a Machine Learning based energy consumption prediction method, aiming to optimize energy allocation through intelligent means. The article uses Python programming language to build Machine Learning models such as XGBoost (Extreme Gradient Boosting Tree) and Random Forest, and conducts training. The models are evaluated using three performance indicators: MAE, RMSE, and R2, and the two models are mixed to improve prediction accuracy. By comparing the performance indicators and actual prediction results of various models, the most suitable energy consumption prediction model is analyzed. The research results can respond to the national “dual carbon” strategy needs and the school's policy of opening up resources and reducing costs, while providing technical support for the sustainable development of university laboratories.

Keywords

smart laboratory; energy consumption prediction; Machine Learning; XGBoost; Random Forest

References

1. [1] Wang Lan Research on the Prediction and Implementation Path of Carbon Peak Scenarios in Universities [D]. Jinan: Shandong Jianzhu University, 2024
2. [2] Wang Rong Comparative analysis of four machine learning methods for short-term prediction of energy consumption in public buildings [D]. Hohhot: Inner Mongolia University of Finance and Economics, 2023
3. [3] Zhang Zhenjie, Cheng Qiuming, Yang Jie, etc Machine Learning and Mineral Prediction: Taking the Prediction of Iron Polymetallic Deposits in Southwest Fujian as an Example [J]. Frontiers in Geology, 2021, 28 (3): 221-235
4. [4] Zhong Xin Automatic classification of nighttime cloud images for all day cameras based on PSO+XGBoost [D]. Wuhan: Wuhan University of Light Industry, 2024
5. [5] Zhang Xinyue Research on the Application of Machine Learning Models in Tax Risk Identification in City B [D]. Baotou: Inner Mongolia University of Science and Technology, 2024
6. [6] Deng Wei Simulation of Water, Heat, and Carbon Processes in Small Watersheds in the Gully Region of the Loess Plateau Based on Machine Learning [D]. Xi'an: Xi'an University of Technology, 2024
7. [7] Ren Wenzong, Liu Jie, Deng Yuting, etc Research and Prediction of Air Pollution Coupling Based on RF Regression and LSTM Neural Network: A Case Study of Beijing [J]. Science and Technology Innovation, 2025 (10): 22-26
8. [8] Huakangmin, Yin Xusong Design and Implementation of NetEase Cloud Music Comment Data Analysis System Based on Python [J]. Modern Information Technology, 2024, 8 (24): 72-76
9. [9] Hua Kangmin, Feng Guilin, Yuan Luyan Design and Implementation of the "Civil Aviation School" Learning and Examination System Based on LNMP Architecture [J]. Science and Technology Innovation, 2021 (5): 69-71
10. [10] Huo Yujiao Research on Predicting Second hand Housing Prices Based on Single and Combination Models [D]. Beijing: China University of Geosciences (Beijing), 2024

Funding

Henan Higher Education Research Project (2025SXHLX036); Philosophy and Social Sciences Planning Project of Henan Province (2024CLH018)

About the authors

Hua Kangmin (May 1990-), male, Han nationality, from Huaxian County, Henan Province, lecturer, master's student, research direction: educational informatization, computer science and technology, big data application; Liu Zongzhe (May 2003-), male, Han ethnicity, from Shangshui, Henan Province, is currently pursuing a bachelor's degree with a research focus on computer science and technology; Corresponding author: Yin Xusong (October 2001-), male, Han ethnicity, from Huaxian County, Henan Province, currently pursuing a master's degree in deep learning.