Parameters and order identification of fractional-order epidemiological systems by Lévy-PSO and its application for the spread of COVID-19. (March 2023)

• Record Type: Journal Article
• Authors: Xie, Bing; Ge, Fudong
• Is Part Of: Chaos, solitons and fractals. Volume 168(2023)
• Publication Date: 2023-03
• Subjects: Parameters identification -- Order identification -- Nonlinear optimization problem -- Fractional-order epidemiological systems -- Lévy-PSO; Chaotic behavior in systems -- Periodicals; Solitons -- Periodicals; Fractals -- Periodicals; Chaotic behavior in systems; Fractals; Solitons; Periodicals; 003.7
• DOI: 10.1016/j.chaos.2023.113163
• Languages: English
• ISSNs: 0960-0779

Abstract

Abstract: To identify the knowledge about parameters and order is very important for the modeling of fractional-order epidemiological systems. In this paper, such an identification problem is formulated as a nonlinear optimization problem. For solving this, the Lévy-PSO algorithm, which is obtained by applying Lévy flight to generalize the classical particle swarm optimization (PSO), is used. More precisely, we first utilize Lévy-PSO to identify the constant parameters and the order of fractional-order SIR, SEIR systems with simulated data to show the effectiveness of our proposed identification strategy. Then, we continue employing Lévy-PSO to solve the parameter estimation problem of fractional-order SEAIR model under the real data of COVID-19 in Shanghai from 2 / 26 / 2022 to 4 / 27 / 2022 . Numerical examples and associated comparisons with other existing methods allow us to achieve that our proposed identification strategy can generate a good performance with high accuracy and rapid convergence. Highlights: We aim to identify both parameters and order of fractional epidemiological systems. A novel Lévy-PSO that prevents particles from being trapped in local minima is used. Parameters and order identification for fractional SEAIR with real data is tackled.