Decentralized power economic dispatch by distributed crisscross optimization in multi-agent system

Abstract

This paper proposes a high-efficient crisscross optimization (CSO) solution to the multi-area economic dispatch (MAED) in both centralized and decentralized optimization manners. First, the CSO is first employed to solve the complex MAED problem by using two powerful search operators including horizontal crossover and vertical crossover. Second, a unique distributed crisscross optimization (DCSO) is put forward to address the MAED problem in a fully decentralized optimization manner, aiming to protect the data privacy, reduce the solving dimensions, and alleviate the heavy communication burden. Under the decentralized framework, the proposed DCSO allows several separate CSOs across the network to optimize the operation of each generation area in parallel. Third, the CSOs assigned to different generation areas are all implemented by multi-agent system that provides the underlying framework of communication between generation areas, which contributes to achieving the independent and asynchronous optimization in each area while minimizing the total operational cost of the entire multi-area power system. Finally, the proposed approach is validated on three different cases. The experimental results verify the superiority of the CSO over other methods in solving the conventional MAED problem and confirm the effectiveness of the proposed DCSO in solving the decentralized MAED problem.