RUS  ENG
Full version
JOURNALS // University proceedings. Volga region. Physical and mathematical sciences // Archive

University proceedings. Volga region. Physical and mathematical sciences, 2016 Issue 1, Pages 23–36 (Mi ivpnz254)

Mathematics

Numerical simulation of replicator systems

T. Yakushkinaa, I. V. Eriklintsevb

a National Research University “Higher School of Economics”, Moscow
b Institute of Computer Aided Design of RAS, Moscow

Abstract: Background. Replicator systems often arise when evolution is concerned. Mathematical models of population dynamics, game theory, economics and biological and molecular evolution lead to systems of partial differential equations. Due to the absence of analytical solutions for a vast majority of such problems, approximate solutions obtained via numerical simulation are required. Hence, construction of efficient algorithms for solving spatial and time-dependent replicator systems is crucial for understanding the dynamics and properties of evolution. Results. We gave an overview of the existing approaches to numerical simulation of replicator systems arising in various fields. We described a mathematical model of population dynamics with explicit space in game theory setting with an asymmetric conflict and a model of biological evolution in presence of a mutator-gene. Both models led to nonlinear systems of partial differential equations that we reduced to the same general form. Then we described the numerical method based on the finite volume framework to solve the system, and provided some numerical examples that demonstrate the method's validity. Conclusions. We conclude that the constructed numerical method is suitable for simulation of replicator systems of general form.

Keywords: replicator equations, reaction-diffusion systems, bimatrix games, Crow-Kimura model, numerical simulation, finite volume method, partial differential equations.

UDC: 519.63, 519.837.4, 51-76



© Steklov Math. Inst. of RAS, 2024