Geometric random walk of finite number of agents under constant variance. (8th May 2017)

• Record Type: Journal Article
• Authors: Yano, Ryosuke
• Is Part Of: Journal of statistical mechanics. (2017:May)
• Publication Date: 2017-05-08
• Subjects: 4 -- 15; Statistical mechanics -- Periodicals; Mechanics -- Statistical methods -- Periodicals; 530.1305
• DOI: 10.1088/1742-5468/aa6ce6
• Languages: English
• ISSNs: 1742-5468

Abstract

Abstract: The characteristics of the 1D geometric random walk of a finite number of agents are investigated by assuming constant variance. Firstly, the characteristics of the steady state solution of the distribution function, which is obtained using the extended geometric Brownian motion (EGBM), are investigated in the framework of the 1D Fokker–Planck type equation. The uniqueness and existence of the steady state solution of the distribution function requires the number of particles to be finite. To avoid the divergence of the steady state solution of the distribution function at the mean value in the 1D Fokker–Planck type equation, the hybrid model, which is a combination of EGBM and normal BM, is proposed. Next, the steady state solution of the distribution function, which is obtained using the geometric Lévy flight, is investigated under constant variance in the framework of the space fractional 1D Fokker–Planck type equation. Additionally, we confirm that the solution of the distribution function obtained using the super-elastic and inelastic (SI-) Boltzmann equation under constant variance approaches the Cauchy distribution, when the power law number of the relative velocity increases. Finally, dissipation processes of the pressure deviator and heat flux are numerically investigated using the 2D space fractional Fokker–Planck type equations for Lévy flight and SI-Boltzmann equation by assuming their linear response relations.