Predicting vehicle fuel consumption in energy distribution companies using ANNs

摘要

Energy distribution companies use various vehicles such as Outage Elimination Maintenance Vehicles (OEMVs), Maintenance Vehicles (MVs) and Supervision Vehicles (SVs) in order to resolve network faults, to conduct inspection, service and maintenance activities and to monitor the integrity of operations. These vehicles have different levels of fuel consumption which concerns companies with regards to the cost and environmental hazards. Hence, it is important to know that a distribution company how many and what types of vehicles do need to predict the fuel consumption of these vehicles. Extra fuel consumption may be required owing to carry unnecessary equipment or used vehicles with large engine volume. Therefore, in this paper, the fuel consumption levels of outage elimination maintenance, maintenance, and supervision vehicles are considered. Then, given the fact that fuel consumption increases with the increase of vehicle weight and Engine Displacement (ED), the extra fuel consumption has been predicted using Artificial Neural Networks (ANNs). In a case study, the total annual fuel cost in the network under consideration is estimated to be $66750. In the case of a 40% increase in vehicle weight, the fuel consumption cost will rise by $24500, and when the engine displacement increases for a 500 CC, the cost rises up $17400 per year.