Coalition Control Model: A Dynamic Resource Distribution Method Based on Model Predicative Control

Abstract: Optimization of resource distribution has been a challenging topic in current society. To explore this topic, we develop a Coalition Control Model(CCM) based on the Model Predictive Control(MPC) and test it using a fishing model with linear parameters. The fishing model focuses on the problem of distributing fishing fleets in certain regions to maximize fish caught using either exhaustive or heuristic search. Our method introduces a communication mechanism to allow fishing fleets to merge or split, after which new coalitions can be automatically formed. Having the coalition structure stabilized, the system reaches the equilibrium state through the Nash-Bargaining process. Our experiments on the hypothetical fishing model demonstrated that the CCM can dynamically distribute limited resources in complex scenarios.