Open Access

ARTICLE

Adaptive Multi-Objective Energy Management Strategy Considering the Differentiated Demands of Distribution Networks with a High Proportion of New-Generation Sources and Loads

Huang Tan¹, Haibo Yu¹, Tianyang Chen¹, Hanjun Deng², Yetong Hu^3,*

1 China Electric Power Research Institute, State Grid Corporation of China, Beijing, 100080, China
2 State Grid Hunan Electric Power Company Marketing Service Center (Metering Center), Changsha, 410000, China
3 College of Information and Electrical Engineering, China Agricultural University, Beijing, 100080, China

* Corresponding Author: Yetong Hu. Email:

Energy Engineering 2025, 122(5), 1949-1973. https://doi.org/10.32604/ee.2025.062574

Received 21 December 2024; Accepted 05 March 2025; Issue published 25 April 2025

Abstract

With the increasing integration of emerging source-load types such as distributed photovoltaics, electric vehicles, and energy storage into distribution networks, the operational characteristics of these networks have evolved from traditional single-load centers to complex multi-source, multi-load systems. This transition not only increases the difficulty of effectively classifying distribution networks due to their heightened complexity but also renders traditional energy management approaches—primarily focused on economic objectives—insufficient to meet the growing demands for flexible scheduling and dynamic response. To address these challenges, this paper proposes an adaptive multi-objective energy management strategy that accounts for the distinct operational requirements of distribution networks with a high penetration of new-type source-loads. The goal is to establish a comprehensive energy management framework that optimally balances energy efficiency, carbon reduction, and economic performance in modern distribution networks. To enhance classification accuracy, the strategy constructs a multi-dimensional scenario classification model that integrates environmental and climatic factors by analyzing the operational characteristics of new-type distribution networks and incorporating expert knowledge. An improved split-coupling K-means pre-clustering algorithm is employed to classify distribution networks effectively. Based on the classification results, fuzzy logic control is then utilized to dynamically optimize the weighting of each objective, allowing for an adaptive adjustment of priorities to achieve a flexible and responsive multi-objective energy management strategy. The effectiveness of the proposed approach is validated through practical case studies. Simulation results indicate that the proposed method improves classification accuracy by 18.18% compared to traditional classification methods and enhances energy savings and carbon reduction by 4.34% and 20.94%, respectively, compared to the fixed-weight strategy.

---

Keywords

---