LIC-free topology identification of complex dynamic networks
Yuzhi Kong, Yi Zheng, Ling Lei, Hao Liao, Xiaoqun Wu · Europhysics Letters (EPL) · 2025
Abstract Over the past decade, synchronization-based identification methods have become a commonly used approach for identifying unknown network topologies. Most existing identification methods rely on the linear independence condition (LIC), which has been shown to have certain limitations. Recently, several new methods have been proposed to overcome this constraint. Building on existing methods, this paper proposes a series of novel predefined signals as drive network. Based on the proposed predefined signals, we develop a novel topology identification method and establish a new proof framework to theoretically validate the effectiveness of our method. Through the designed controllers and update laws, synchronization between the drive and auxiliary networks is achieved, thereby enabling the estimated matrix to accurately identify the unknown topology matrix. Numerical simulations demonstrate the effectiveness of the proposed method and validate the impact of strength selection for the predefined signals on identification performance.