Electric Power Profiles Contract via Probabilistic Optimization

Abstract

In this paper, we deal with a problem of how to make contracts of time series of electric energy, called electric power profiles, in electricity markets. We suppose that each energy supplier can deliver a time series of electric energy in a possible profile. In such situation, each electric power profile has its own value for suppliers and consumers and it should be traded in electric markets appropriately. However, different from a case of ordinary spot markets, which deal with a contract on each time slot independently, an electric power profile includes multidimensional variables according to the number of time slots, then to find a matching of a profile and its price between suppliers and consumers is a hard problem from computation complexity. In order to solve this difficulty, we propose a probabilistic algorithm to find a matching of an electric power profile and its price between suppliers and consumers. A kind of properness of the solution is guaranteed in a probabilistic sense and the method also can avoid the curse of dimensionality. The efficiency of the proposed algorithm and the necessary computation complexity are also verified by a numerical simulation.