A Game Theoretic Approach for Demand-Side Management Considering Generation, Storage and the Combinatorial Nature of Load Scheduling

In this paper an energy consumption scheduling for residential area considering generation, storage and the combinatorial nature of shiftable device scheduling has been proposed. The smart grid consists of the traditional users as well as users with smart meter who participate in the day ahead optimization process to reduce their energy bill by storing, generating and shifting their device to non-peak hours. These entities by participating in optimization process reduce the per unit energy cost of the grid for all users. It is assumed that all of the participating users own a storing, generating or both devices having same characteristic. A billing scheme has been adopted to encourage the user to shift their load to non-peak hour. The users will compete to shift their load as well as to charge their storage device to low load period, therefore the resulting day ahead optimization problem has been formulated as non-cooperative game between the users. The iterative distributed algorithm can be run on user's smart meters to solve the formulated problem. The strategy set for the scheduling of time shiftable devices are discrete and combinatorial in nature, hence particle swarm algorithm for optimizing the individual users pay off function has been used. The obtained results demonstrated the effectiveness of the proposed method in terms of reduced energy price and system peak while considering the user's privacy and comfortability.