A large-scale equilibrium model of energy emergency production: Embedding social choice rules into Nash Q-learning automatically achieving consensus of urgent recovery behaviors

A large-scale energy emergency production plan driven by extreme events in energy supply chain networks is the low-probability/high-consequence event that is difficult to prepare. One of the most prominent challenges is efficiently computing the equilibrium point characterized with more frequently involved in common resource conflicts due to different player behaviors in energy supply chain networks. In this paper, a novel large-scale equilibrium model of energy emergency production: embedding social choice rules into Nash Q-learning automatically achieving consensus of urgent recovery behaviors, is proposed to tackle this challenge. The main contributions of this work are that firstly set up a large-scale equilibrium model of energy emergency production to formulate energy emergency production plans by modifying the large-scale energy equilibrium model, and the computational limitations of Generalized Nash Equilibrium are overcame by combination of Nash Q-learning methods and individuals’ preferences reaching a collective decision which guarantees uniqueness of the large-scale Nash equilibrium to achieve both system-level efficiency and maximum fairness. Simulations results show that the generalized Nash bargaining solution can be implemented by the proposed large-scale equilibrium model of energy emergency production, in which outcome of the game is the emergency production stable equilibria alternative with no chance moves in a given consensus level, and compared with the existing techniques considering non-cooperative behaviors, it has a significantly lower minimisation of time for the energy restoration by twenty-nine percent, and reduces minimisation of cost for the energy restoration by seventeen percent and minimisation of carbon dioxide emissions by twenty-three percent with disaster recovery.