A Feature Engineering Framework for Smart Meter Group Failure Rate Prediction

Yihong Li, Xia Xiao (), Zhengbo Zhang and Wenao Liu
Additional contact information
Yihong Li: School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
Xia Xiao: School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
Zhengbo Zhang: School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
Wenao Liu: School of Electrical and Electronic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China

Mathematics, 2025, vol. 13, issue 15, 1-20

Abstract: Smart meters play a significant role in power systems, but their condition assessment faces challenges such as inconsistent evaluation criteria and inaccurate assessment results. This paper proposes feature engineering including feature construction and feature selection for smart meter group failure rate prediction. First, the basic structure and common fault types of smart meters are introduced. Smart meters are grouped by batch and distribution area. Next, 25 condition features are constructed based on failure mechanisms and technical specifications. Then, an evolutionary multi-objective feature selection algorithm combining NSGA-II, Jaccard similarity, and XGBoost is developed, where feature subsets are encoded as binary individuals optimized for three objectives: MSE, 1 − R 2 , and the number of features. The experimental results demonstrate that the proposed method not only reduces the number of features (25→7) but also improves the prediction accuracy (MSE: 0.0049 → 0.0042, R 2 : 0.6638 → 0.7228) of smart meter group failure rates. Comparative studies with other feature selection methods further confirm the superiority of our approach. The optimized features enhance interpretability and computational efficiency, providing a practical solution for large-scale smart meter condition assessment in power systems.

Keywords: smart meter group; condition assessment; feature construction; evolutionary algorithm; multi-objective; feature selection (search for similar items in EconPapers)
JEL-codes: C (search for similar items in EconPapers)
Date: 2025
References: Add references at CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/2227-7390/13/15/2472/pdf (application/pdf)
https://www.mdpi.com/2227-7390/13/15/2472/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jmathe:v:13:y:2025:i:15:p:2472-:d:1714502

Access Statistics for this article

Mathematics is currently edited by Ms. Emma He

More articles in Mathematics from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().