A discrete particle swarm optimisation algorithm to operate distributed energy generation networks efficiently

This paper deals with the optimisation of the operating costs in a distributed electric and heating energy generation network. The network considers different options to supply the electric and heating demand of a large consumer building: the electricity can be directly bought from the grid, can be taken from renewable energy sources or can be produced from gas using a combined heat and power system. In the same line, the heating can be taken from a thermal solar renewable system, from the boiler or from the combined heat and power system. In addition, the large consumer has batteries to store electricity excesses and thermal storage systems to store the heating excess. The multicommodity flow mathematical formulation of the problem couples both electric and thermal models by considering cogeneration systems. The model is solved by a particle swarm optimisation (PSO) algorithm that is compared to the optimal solutions provided by Gurobi optimisation commercial software and a Monte Carlo algorithm. The PSO algorithm proved a very efficient performance in the available short time to provide the energy commands to the systems outperforming the alternative approaches.