Application of multi-agent scheduling optimisation algorithm in Finnish residential energy management system

Abstract

This thesis explores the possibility of using a scheduling optimisation algorithm, the Iterative Economic Planning and Optimised Selections (I-EPOS), to solve load balancing problems in the Finnish residential sector. The algorithm uses principles of self-adaptive collective learning to solve the decentralised combinatorial optimisation problems, which offers flexibility and adaptability to the energy management system structure in addressing future electricity needs, as well as ensuring security for all participants in the system.

The thesis contains the necessary base knowledge of the combinatorial optimisation problem and the working principles of the algorithm itself, as well as a short performance simulation of the algorithm and recommendations on deploying it in a hypothetical Finnish household.