University of Twente Student Theses

Login

Autonomous and cooperative vehicles and highway capacity : a theoretical model approach for vehicle automation and communication on Dutch highway road segments, focused on non-platoon based care-following behavior

Rossen, V.G. (2018) Autonomous and cooperative vehicles and highway capacity : a theoretical model approach for vehicle automation and communication on Dutch highway road segments, focused on non-platoon based care-following behavior.

[img] PDF
16MB
Abstract:Autonomous driving is one of the main current developments in the field. Many aspects, including the technology and systems used to achieve autonomous driving, human interaction, safety, reliability and security as well as the effects on traffic, are involved. This study focuses on the effects on capacity of multiple levels and penetration rates of automated driving on Dutch highways. Automated driving consists of many forms and levels, with and without communication, that can be achieved on the short term as well as the long term. The capacities and network design according to the Dutch highway manual, are the base for the microscopic traffic simulation model VISSIM, in which longitudinal and lateral movements are parameterized. The network, namely 8 road segments, and driving behavior parameters are modeled and calibrated for 2 reference subscenarios, one for straight road segments (2-6 lanes) and one for merging and diverging road segments (an On-Ramp, an Off-Ramp and a Symmetrical Weaving Section). MATLAB is used to control the VISSIM simulations and change input. Validation is done with measured data of a Dutch highway. Alteryx and MATLAB are used to (pre)process the data and to produce output in the form of tables, boxplots and other figures. In total 1300 simulations have been executed in VISSIM during 100 hours of net simulation time and many more for preprocessing, processing and presenting output. Five main scenarios for vehicle automation are studied: • Adaptive Cruise Control (ACC) • Cooperative Adaptive Cruise Control (CACC) • Autonomous Vehicles (AV) • Cooperative Autonomous Vehicles (CAV) • Cooperative Autonomous Vehicles with fixed desired speed (CAV+) The simulations show a drop in capacity of 5-10% for road segments with straight lanes for scenarios with 100% market penetration of ACC in comparison to the reference scenario, while scenarios with 100% cooperative vehicles (both CACC and CAV) show an increase of around 10% to over 20% in capacity in comparison to the modeled reference. Scenarios with autonomous vehicles without communication show a small increase in capacity for these road segments. An On-Ramp shows similar results, though the capacity increase of cooperative vehicles is limited to around 10%. An Off-ramp barely shows any differences between scenarios in comparison to reference, only 100% ACC shows a drop of about 3% and the scenarios with 100% cooperative autonomous vehicles show an increase of 3.1% (CAV) to 7.3% (CAV+). Last, the Symmetrical Weaving Section is very difficult to calibrate properly. However, the cooperative scenarios again show higher capacities than the other scenarios, but since the road segment has not been modeled properly in the reference scenario, conclusions must be drawn very carefully. Scenarios with high capacities have high (left lane) speeds for higher intensities and approximate equal lane shares near congestion later or not even at all. This can be explained by the deviation in speed per lane. The scenarios with high capacities show less deviation in (mean) speed on the left lane between different intervals, as well as much smaller left lane standard deviation in speed within 5-minute intervals. To conclude, vehicle automation is likely to decrease capacity in the short-term since vehicles equipped with ACC use larger headways than human drivers, while in the long-term it will increase capacity since autonomous vehicles can use smaller headways and have smaller speed deviations, especially on the left lane. Communication between vehicles is hereby more important than vehicle automation itself.
Item Type:Essay (Master)
Faculty:ET: Engineering Technology
Subject:56 civil engineering
Programme:Civil Engineering and Management MSc (60026)
Link to this item:http://purl.utwente.nl/essays/76615
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page