Algorithm based on particle swarm applied to electrical load scheduling in an industrial setting

In this work we propose the development of a novel particle swarm-based heuristic to solve a discrete mathematical problem. Such a problem is present in allocating electrical loads throughout the day in an industrial setting. Data on the total installed load and energy demand throughout the day at 15-min intervals were collected in five industrial facilities. The loads were randomly distributed and the developed algorithm was applied to balance and optimize the energy demand throughout the day. The performance of the proposed algorithm was compared to a standard binary Particle Swarm Optimization and a mathematical model, which was also implemented to solve the problem. Our results demonstrate that the proposed algorithm is more efficient for all the considered scenarios, regardless of the amount of loads and constraints applied.