Towards smart EVs energy scheduling for a logistic distribution center: cost, emissions and makespan optimizing

Electrifying the transport sector supplied by renewable energy is crucial for reducing emissions and addressing the challenges associated with climate change. This paper presents a novel multiobjective approach for transporting goods in a logistics distribution (LD) network, focusing on sustainable Vehicle-to-Grid (V2G) smart energy scheduling to simultaneously minimize three key performance indicators: total energy system costs, equivalent C O 2 e emissions, and operational makespan using two enhanced versions of advanced reference-point-based metaheuristics, the Nondominated Sorting Genetic Algorithm III (NSGA-III) and the Adaptive Reference-point MOEA (AR-MOEA), to address the many-objective optimization problem. The optimization determines the timing of each delivery and the charging or discharging power of the electric vehicles (EVs) when connected to the charger, while also accounting for emissions and the costs associated with battery degradation. A case study involving a photovoltaic system and five EVs completing 19 deliveries demonstrates a well-balanced trade-off among the three objective functions, achieving a makespan (time from the first EV departure to the last EV return to the LD) of 10.00 h, 53.87 kgC O 2 e of emissions, and a daily energy system cost of R$ 973.56. These results are made possible by the elimination of energy consumption during peak periods (moments of highest price and higher emissions) and the energy injection into the Distributed System Operator grid, in addition to the maintenance of a reduced maximum demand level.