Development of Virtual Fuel Economy Trend Evaluation Process 2019-01-0510

With the advancement of the autonomous vehicle development, the different possibilities of improving fuel economy have increased significantly by changing the driver or powertrain response under different traffic conditions. Development of new fuel-efficient driving strategies requires extensive experiments and simulations in traffic. In this paper, a fuel efficiency simulator environment with existing simulator software such as Simulink, Vissim, Sumo, and CarSim is developed in order to reduce the overall effort required for developing new fuel-efficient algorithms. The simulation environment is created by combining a mid-sized sedan MATLAB-Simulink powertrain model with a realistic microscopic traffic simulation program. To simulate the traffic realistically, real roads from urban and highway sections are modeled in the simulator with different traffic densities. Other traffic elements which would affect the fuel consumption, such as speed limit information, traffic stop sign and a traffic lights with SPaT (Signal Phase and Timing) information, are a part of the simulator. In order to evaluate the performance of the developed algorithms, fuel consumption performance of the developed algorithms are compared with the fuel consumption performance of a simple driver model, the Intelligent Driver Model (IDM). Therefore, the IDM with different aggressiveness levels is integrated into the simulation environment. Finally, the developed simulation environment is integrated into CarSim for perception sensor simulation capabilities. This developed simulation environment aims to accelerate development of fuel-efficient algorithms.