Application of EPSO to designing a contract model of weather derivatives in Smart Grid

This paper proposes an efficient method for designing a contract model of the weather derivatives between energy utilities in Smart Grid. It is well-known that the weather conditions bring a profit decline or the increase of expenses to do damage to sound management. Weather derivatives are useful for solving such a problem. One of the ideas is to use the complementary relationship between electric power and gas companies in a sense that electric power companies are apt to make profits in hot summer while gas companies are inclined to reduce revenue. This paper focuses on how to create a reasonable contract model of the weather derivative. In this paper, EPSO (Evolutionary Particle Swarm Optimization) of meta-heuristics is applied to designing a contract model of the weather derivative. The proposed method aims at equalizing the mean and the variance of the payoffs between the power and gas companies. To enhance the model accuracy, DA clustering of global clustering is used to classify the historical data into clusters. The effectiveness of the proposed method is demonstrated for the real data in Tokyo, Japan.