Institutional Repository
Technical University of Crete
Technical University of Crete
EN
|
EL
| URI: | http://purl.tuc.gr/dl/dias/610A99D3-5177-484C-ACA4-E577E3505B09 | ||
| Year | 2023 | ||
| Type of Item | Diploma Work | ||
| License |
|
||
| Bibliographic Citation | Alexandros-Konstantinos Valtatzis, "Highway Traffic State Estimation and Control in Presence of Connected Automated Vehicles", Diploma Thesis, School of Electrical and Computer Engineering, Technical University of Crete, Chania, Greece, 2023 https://doi.org/10.26233/heallink.tuc.96811 | ||
| Appears in Collections |
It is expected that the widespread adoption of Vehicle Automation and Communication Systems in the upcoming years will have a strong impact on freeway traffic performance. In addition to ensuring comfort and safety, other key objectives are to estimate and reduce traffic congestion, which constitute two related, substantial and demanding issues of the modern world. The aim of this diploma thesis is twofold. Its first objective is to demonstrate the efficiency of a macroscopic model-based approach, based on a Kalman filter, in estimating the total density and flow of vehicles in a real road network in Antwerp, Belgium, using real data and assuming that all vehicles are Connected Automated Vehicles (CAVs). The proposed filter utilizes only real speeds and a limited amount of real flow measurements from spot-sensors, on a 48Km motorway stretch that starts from A13 2100 in Boterlaar-Silsburg neighborhood and ends at A13 3945 in Ham municipality, Antwerp, Belgium. The accuracy and reliability of the estimated traffic states are assessed through comparison with ground truth measurements. The second goal of this diploma thesis is to show the effectiveness of utilizing an Adaptive Cruise Control (ACC) traffic control strategy in boosting the motorway traffic flow at active bottleneck locations by adjusting the time-gap of ACC-equipped vehicles in selected motorway sections in real-time, based on current and estimated traffic conditions. The proposed deployment and application scenario is implemented successfully in a toy road network in the Aimsun Next microsimulator, using the Gipps car-following model, which is known for its limitations in relation to the capacity drop phenomenon. The simulation results are displayed over different ACC penetration rates (the percentage of the CAVs present in the traffic) and prove that there is a significant improvement in the total time spent in the road network and in the average vehicle delay, as the penetration rate increases.
Copyright © DIAS 2013
