guest ::
TY - THES AU - Sohn, Christian TI - Vorausschauende Fahrzeuglängsführung verbrennungsmotorisch und batterieelektrisch betriebener Fahrzeuge PB - Rheinisch-Westfälische Technische Hochschule Aachen VL - Dissertation CY - Aachen M1 - RWTH-2023-09900 SP - 1 Online-Ressource : Illustrationen, Diagramme PY - 2023 N1 - Veröffentlicht auf dem Publikationsserver der RWTH Aachen University N1 - Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023 AB - Automated longitudinal control systems undertake the task of planning and adjusting vehicle velocity. Adapting vehicle velocity to route, other road users and switching times of traffic lights can lead to energy savings in the double-digit percentage range. The challenge here is the computationally efficient velocity planning over a long-term horizon. Existing approaches based on discrete dynamic programming are either computationally expensive or suboptimal. Another challenge is the dedicated exploitation of vehicle characteristics in velocity planning. Facing the increasing penetration rate of mild-hybrid electric vehicles, in particular the use of sailing and recuperation increase in importance. In most cases, the sailing function remains unused if not explicitly considered in the longitudinal control design. The mild-hybrid technology also enables the fuel-saving pulse and glide (PnG) strategy, consisting of alternating acceleration and deceleration phases. The fuel savings of this strategy have already been investigated several times; however, the aspect of customer acceptance has been neglected. The thesis consists of two parts. In the first part, ride comfort and fuel savings of PnG are investigated in detail. The second part presents a predictive longitudinal control system that considers the above levers for reducing energy consumption, including the PnG strategy, and that can be used for both internal combustion engines and battery electric vehicles by switching off specific functions. For long-term planning (LTP), an algorithm is developed that finds the optimal velocity trajectory (in terms of energy demand at the wheels) to reach the target green phase of the next traffic light with little computational effort, and that does not suffer from the drawback of discrete dynamic programming with two-dimensional state space. The algorithm is combined with a technique that adapts the sailing-to-recuperation ratio of deceleration profiles to the switching times of traffic light and an efficient traffic light approach. For short-term planning, a rule-based approach is used that implements the deceleration profiles already known to be optimal for the basic driving maneuvers as well as a PnG operation explicitly considering ride comfort. The longitudinal control system is evaluated by simulated round trips within the city Paderborn in dynamic traffic scenarios. Reference for evaluation is a longitudinal control system employing model-predictive control (MPC), which navigates the vehicle within a solution space, penalizing acceleration and deceleration maneuvers as well as deviations from a maximum velocity. For the mild-hybrid vehicle (P0 hybrid), activating the LTP achieves average fuel savings of about 20 LB - PUB:(DE-HGF)11 DO - DOI:10.18154/RWTH-2023-09900 UR - https://publications.rwth-aachen.de/record/971921 ER -