Equilibrium Search Technique Using Genetic Algorithm and Replicator Dynamics and Its Application to Food Supply Chain Model

Although expressing a social system as a game model and searching for equilibrium states are useful in the design study of social systems, the calculation time is a problem in practical applications. In this study, we propose an equilibrium search technique that is based on genetic algorithm and introduced the concept of replicator dynamics, wherein strategies with a higher payoff are more likely to be employed in the next generation. Using the proposed method, the calculation time for a game that models an auction of products was reduced by 46%. Additionally, the effects of the selling time limit for manufacturers and supermarkets, one of the business practices, on the food (milk) supply chain were examined. Under the conditions implemented in our study, the relaxation of such limits reduced food loss and waste without harming the profits of stakeholders in the supply chain.