New type of chaotic solutions found in Gravity model of network transport

TL;DR

This paper investigates the complex dynamics of gravity models on small networks, revealing new types of chaotic solutions and their stability through numerical simulations.

Contribution

It introduces the discovery of new chaotic solution types in gravity models on networks, especially in small ring networks with seven nodes.

Findings

Chaotic solutions exist in small networks, notably in rings with seven nodes.

New mixture of periodic solutions identified as a novel chaotic behavior.

Parameter regions for stable periodic and chaotic solutions mapped.

Abstract

The gravity model is a mathematical model that applies Newton's universal law of gravitation to socio-economic transport phenomena and has been widely used to describe world trade, intercity traffic flows, and business transactions for more than several decades. However, its strong nonlinearity and diverse network topology make a theoretical analysis difficult, and only a short history of studies on its stability exist. In this study, the stability of gravity models defined on networks with few nodes is analyzed in detail using numerical simulations. It was found that, other than the previously known transition of stationary solutions from a unique diffusion solution to multiple localized solutions, parameter regions exist where periodic solutions with the same repeated motions and chaotic solutions with no periods are realized. The smallest network with chaotic solutions was found to be a ring with seven nodes, which produced a new type of chaotic solution in the form of a mixture of right and left periodic solutions.