Proactive Integrated Control for Relieving Freeway Congestion

• Author / Creator

  Wang, Xu

• Active bottlenecks limit traffic flow on freeway corridors. To relieve bottleneck severity, ramp metering (RM), variable speed limit (VSL) and their integration are often implemented to control the on-ramp and mainline input flow. Currently, freeway operation has become proactive based on short-term prediction. Macroscopic traffic flow models are often applied as prediction models in proactive traffic control strategies. Prior to field implementation, the models need to be calibrated and validated carefully to ensure that they represent real-life traffic situations. This study proposes modifications for METANET model to adapt it to the unpredictability of bottleneck activation during peak hours. The modified model is calibrated and verified its improvement of model prediction accuracy from segment-specific parameters. The modified model is validated that it can replicate traffic state evolutions during peak hours and be applicable in proactive traffic control practice. Weaving maneuvers (i.e., intensive lane changes) are a major cause of bottlenecks during high-demand periods. To consider weaving impacts in RM, this study introduces a proactive optimal RM algorithm that uses dynamic weaving capacity at weaving segments. Sensitivities of capacity and capacity drop are applied to dynamically estimate weaving capacity within a macroscopic traffic flow model. The proposed traffic flow model conducts estimation in a model predictive control (MPC) frame-work. The proposed RM algorithm is evaluated in macro-simulation and its effectiveness is enhanced by real-time estimated weaving capacity. The RM control research reveals a need of theoretical methods for weaving capacity estimation. This study then defines a linear optimization problem to solve weaving capacity and then establishes a lane-changing model to constrain the weaving flows. The proposed method is evaluated and analyzed for sensitivity with field data from two weaving segments. The capacity estimates from the proposed model are consistent with those from the HCM 2010 model and with field observations. Moreover, the weaving capacity is sensitive to weaving maneuvers. The proposed method is finally applied to estimate the real-time maximum discharge flow rate; the estimates match field measurements. Next, this study presents a proactive integrated control of RM and VSL, with goals to improve network-wide travel time and traffic flow. By decoupling the traffic prediction and simulation models, the possible control error sources are analyzed. The evaluation reveals the proactive integrated control achieves an amelioration in total time spent (TTS) up to 13.65% and an increase in total travel distance (TTD) up to 3.41%. The isolated and integrated controls benefit the traffic network in different extent under different demand scenarios. In addition, control rate profiles are analyzed in detail and found that RM is activated during slight congestion and the most congested situation to assist VSL. Through the integration, the infrastructure utility is maximized. Speed transition zones are complex when dynamically created and shifted by VSL. This study then attempts to represent speed limit effect and estimate real-time driver compliance at speed transition zones. The field data from two speed transition zones are investigated for temporal and spatial variations of speed and driver compliance using statistical tests. After selecting several key factors from statistical tests, a linear regression is established to rank the contributions of the selected factors and other general factors proposed by previous research. The regression results confirm speed limit value, surrounding traffic speed and existence of activated speed enforcement or education devices contribute more to driver compliance. Finally, this study reports the preliminary VSL test and details its implementation procedure on Whitemud Drive, Edmonton, Canada. DynaTAM-VSL software is designed to realize all necessary functions for VSL filed implementation. The preliminary test is conducted, and the VSL control performance and reliability are evaluated. The results for before-and-after VSL control are finally analyzed in depth. The analysis compares average traffic speed, standard deviation of speed, total travel time and total travel distance. The results from this study confirm that VSL can relieve recurrent traffic congestion.

• Subjects / Keywords

  • Freeway operation
  • Active traffic and demand management
  • Variable speed limit
  • Ramp metering
  • Integrated control

• Graduation date

  Fall 2015

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/R34X54N4D

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.