More than ever, the climate crisis requires the transport sector to significantly reduce energy demand and the respective CO2 equivalents. With the help of predictive optimization methods for longitudinal vehicle control, automated and connected vehicles can make a remarkable contribution to this. To ensure traffic safety, the optimization methods may only select the vehicle speed within certain limits. Previous research mostly omits complex situations with crossing traffic at intersections not controlled by traffic signals or assumes idealized knowledge of solution space limits. This is usually not sufficient for operation in real traffic situations. In the scope of this work, it is shown that the prediction of traffic boundary conditions can be robustly extended to numerous traffic situations of the particularly challenging urban environment and can be utilized to reduce energy demand. The prediction algorithm developed for this purpose estimates the future behavior of surrounding traffic within the prediction horizon based on data from environmental perception technologies (digital map, vehicle connectivity as well as environmental sensors) and derives time-dependent speed and position constraints for the optimized vehicle. In order to evaluate the accuracy of the developed prediction algorithm, it has been investigated in real traffic scenarios. To this end, a measurement campaign has been carried out in the urban environment of the Aldenhoven Testing Center and numerous random traffic situations have been recorded. The measurement data was then processed in the MATLAB/Simulink development environment and each of the participating vehicles has been virtually equipped with the prediction algorithm. To evaluate the performance, the predicted values has been compared with the real vehicle behavior over the prediction horizon and for different traffic situations and parameter variations. To quantify the energy saving potential, the algorithm has been combined with a model predictive control. The performed simulation study shows the general saving potential of the prediction algorithm in combination with an energetic optimization. Depending on the choice of parameters, savings of between 12% and 23% can be achieved compared to a reference vehicle whose driving behavior is based on that of a human driver. In summary, the work performed can support the activities of bringing automated vehicles one step closer to higher automation levels. Additionally, the prediction algorithm can be used in further technologies to increase energy efficiency. Overall, a large potential for reducing energy demand and lowering greenhouse gas emissions in the transport sector can be tapped, thus contributing to meeting climate protection targets.
Prädiktion von Verkehrsrandbedingungen zur Effizienzsteigerung vernetzter und automatisierter Fahrzeuge
Prediction of traffic boundary conditions to increase the efficiency of connected and automated vehicles
2023-01-01
1 Online-Ressource : Illustrationen, Diagramme pages
Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023; Aachen : RWTH Aachen University 1 Online-Ressource : Illustrationen, Diagramme (2023). = Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023
Sonstige
Elektronische Ressource
Deutsch
TIBKAT | 2023
|British Library Conference Proceedings | 2018
|Handbuch Einsatz automatisierter und vernetzter Fahrzeuge für den öffentlichen Verkehr
DataCite | 2021
|