Model Order Estimation for Low-Frequency Oscillations Using Optimized Density Peak Clustering with K-Nearest Neighbor and Weighted Similarity.

Saved in:

Bibliographic Details
Title:	Model Order Estimation for Low-Frequency Oscillations Using Optimized Density Peak Clustering with K-Nearest Neighbor and Weighted Similarity.
Authors:	Li, Luzhuang¹ hnliluzhuang@163.com, Zhang, Zhao² zhangzhao333@hotmail.com, Zhou, Hongyan¹ zhou321yan@163.com, Chen, Xue-Bo¹ xuebochen@126.com
Source:	Engineering Letters. May2026, Vol. 34 Issue 5, p1865-1873. 9p.
Subjects:	Clustering algorithms, Parameter estimation, Robust statistics, Frequencies of oscillating systems, Electric power systems, Signal processing
Abstract:	To address the challenge of accurately estimating model order in the analysis of low-frequency oscillation signals in power systems, we propose a density peak clustering with k-nearest neighbor and weighted similarity (DPC-KWS). This algorithm enhances eigenvalue density-based automatic order determination by incorporating nearest neighbor assignments and weighted similarity strategies. By utilizing a k-nearest neighbors relative density calculation, the algorithm effectively adapts to the characteristic eigenvalue distribution, where larger eigenvalues are sparse while smaller ones are dense, thereby enabling precise identification of the signal-noise subspace boundary. Coupled with dynamic neighborhood selection and weighted similarity allocation, this approach significantly enhances clustering robustness and reliability. Simulation results indicate that DPC-KWS achieves superior estimation accuracy and improved robustness under various noise conditions compared to conventional algorithms. Furthermore, when applied to total least squares estimation of signal parameters via rotational invariance techniques parameter estimation, the algorithm produces results that closely align with the true values, thereby demonstrating its effectiveness and practical value in advancing low-frequency oscillation analysis. [ABSTRACT FROM AUTHOR]
	Copyright of Engineering Letters is the property of International Association of Engineers (IAENG) and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)
Database:	Engineering Source

Description
Abstract:	To address the challenge of accurately estimating model order in the analysis of low-frequency oscillation signals in power systems, we propose a density peak clustering with k-nearest neighbor and weighted similarity (DPC-KWS). This algorithm enhances eigenvalue density-based automatic order determination by incorporating nearest neighbor assignments and weighted similarity strategies. By utilizing a k-nearest neighbors relative density calculation, the algorithm effectively adapts to the characteristic eigenvalue distribution, where larger eigenvalues are sparse while smaller ones are dense, thereby enabling precise identification of the signal-noise subspace boundary. Coupled with dynamic neighborhood selection and weighted similarity allocation, this approach significantly enhances clustering robustness and reliability. Simulation results indicate that DPC-KWS achieves superior estimation accuracy and improved robustness under various noise conditions compared to conventional algorithms. Furthermore, when applied to total least squares estimation of signal parameters via rotational invariance techniques parameter estimation, the algorithm produces results that closely align with the true values, thereby demonstrating its effectiveness and practical value in advancing low-frequency oscillation analysis. [ABSTRACT FROM AUTHOR]
ISSN:	1816093X