Quantification of Influencing Factors of an Energy-Efficient Longitudinal Control of Battery Electric Vehicles

Global warming forces the automotive industry to reduce real driving emissions and thus, its CO2 footprint. Besides maximizing the individual efficiency of powertrain components, there is also energy-saving potential in the choice of the driving strategy. Thus, model predictive control based advanced driver assistance systems to reduce the energy consumption during driving gains a significant interest in the literature. However, this results in a complex control system with many parameter dependencies that could possibly affect the energy efficiency of the vehicle. Most of these parameters are subject to uncertainties. Thus, the important question remains how these parameter uncertainties affect the energy efficiency of the system and how a driver assistance system should be designed to be robust against these uncertainties. To answer this question this thesis applies variance-based sensitivity analyses to design an appropriate driver assistance system and to quantify the influences of the uncertain system and controller parameters.
First, a detailed vehicle and powertrain model of a battery electric vehicle is evolved and verified on component test benches. The parameter uncertainties and their sensitivities were investigated on typical urban and interurban commuter routes using quantitative variance-based sensitivity analysis methods. Based on these findings an economic nonlinear model predictive control eco-cruise control is derived which takes the identified parameter dependencies into account. The developed economic nonlinear model predictive control system is evaluated on artificial drive cycles and compared to a linear model predictive control approach as often outlined in the literature.
Afterwards, the closed loop control system, consisting of the developed economic nonlinear model predictive control and the detailed vehicle model, is analyzed on typical urban and interurban commuter routes using variance-based sensitivity analysis. The findings and parameter dependencies are outlined and discussed. It has been shown, that vehicle parameters as well as controller parameters impact the energy consumption and the driving time of the vehicle. It has been outlined that if the as influential identified parameters are optimized, an average energy-saving potential on the investigated routes of 10.5% exists by only increasing the driving time of 0.7%.