A Study on the Impact of Driver Behavior on the Energy Consumption of Electric Vehicles in a Virtual Traffic Environment

Abstract

Mild driving, i.e., accelerating slowly and smoothly, braking less frequently, and increasing spacing between vehicles to avoid harsh braking, has been shown to be effective in the improvement of fuel economy, especially for conventional vehicles with an internal combustion engine. For electric vehicles (EVs), it is implied that the energy consumption can also be improved by driving less aggressively. However, the extent in which the driver behavior reduces the energy consumption of an EV in various traffic environments has not been fully explored. A simulated environment can create a greater variety of driving cycles and conditions, thereby providing more insight as to how driving aggressiveness affects the vehicle’s energy consumption. The objective of this study is to evaluate the impact of the driving behavior on the energy consumption a battery electric vehicle (BEV) under various traffic scenarios. To simulate the driver behavior, a driver model is typically required to replicate the behavior of a human driver while traversing a given route (e.g., maintaining a safe distance from the preceding vehicle). Various driver models can be found in literature. Among these, the widely used Intelligent Driver Model is chosen in this study to characterize the different levels of driver aggressiveness. To that end, the microscopic traffic simulator, PTV Vissim, is used to simulate various realistic traffic environments in which a human driver’s behavior can be evaluated. The co-simulation of the PTV Vissim Component Object Model (COM) interface in conjunction with MATLAB allows the energy consumption performance on an EV to be determined for various levels of driving aggressiveness. The results obtained from the co-simulation with a virtual traffic environment are compared to those from single-lane car-following scenarios created using EPA (Environmental Protection Agency) standardized driving schedules. The results of the single-lane car-following scenario shows that there is a slight increase (<1.5%) in energy usage per kilometer by changing from a mild driving style to an aggressive driving style. For the city driving cycle created in Vissim, aggressive driving can lead to a 6.6% decrease in the average energy usage per kilometer driven than mild driving if it allows the vehicle to avoid red traffic signals and general vehicle traffic. However, driving at medium-level aggression is not quick enough to avoid these obstacles and consequently increases the average energy usage per kilometer by 1.1% over mild driving. For the highway driving cycle, the benefits of driving milder can be realized, as switching from aggressive to mild driving results in a 3.4% decrease in average energy usage per kilometer. The results of these driving tests demonstrate that the level of driving aggressiveness cannot be fixed and should instead adapt to the traffic environment in order to maximize the battery life and range of an EV.