Aplicação do algoritmo de enxame de partículas na otimização do custo energético de bombas de velocidade variável

Abstract

The strategies of operation of the water supply systems (WSS) aim at the uninterrupted public supply, although they commonly neglect the oscillation of the daily energy tariff. This provides a significant increase in the energy cost of the pumps. However, in addition to efficiently using the tariff, it is also necessary to consider the broad horizon of feasible combinations of operational strategies, water varying demand and the maintenance of water levels in the reservoirs and the pressure in the extremities within pre-established intervals. In this scenario, this work was carried out, whose objective is to determine daily speed profiles of the pumps of an SAA, aiming at minimizing the daily energy cost. For this, a computational model was developed that operates in conjunction with the EPANET hydraulic simulator. The model is based on the meta-heuristic technique PSO (Particle Swarm Optimization), being the decision variable the speed of the pumps. This technique looks for strategies with reduced energy costs while EPANET analyzes hydraulic performance. The results evidenced the effectiveness of the PSO in determining a satisfactory amount of operating strategies and reducing the time required in this task. Thus, it is considered that the methodology developed can help the operation of WSSs, aiming at energy optimization.