Traffic flow operation studies are essential for understanding and optimizing the functioning of transportation systems. These studies investigate the movement of vehicles on road networks, examining factors such as congestion, capacity, and overall efficiency. The goal is to enhance traffic flow, reduce delays, and improve overall transportation system performance.

The data science and statistical inference methods play a crucial role in the empirical analysis and modeling aspects of traffic flow operations. Accurate and comprehensive data exploration is fundamental to understanding the intricacies of traffic patterns, vehicle behaviors, and the impact of various factors on flow dynamics. Statistical inference methods play a crucial role in extracting meaningful insights from the data, enabling the development of predictive models and informed decision-making.

Driver maneuvers encompass longitudinal actions, guiding vehicles along their path, and lateral maneuvers, typically involving lane changes within motorway networks.

Lane-changing maneuvers (LCMs) involve gap acceptance decisions and require the vehicle to temporarily occupy two lanes, resulting in reduced capacity, decreased speeds, and increased accident risk. LCM maneuvers are especially prevalent in complex motorway segments like weaving areas, making such areas congestion and accident hotspots.

The emergence of connected and automated vehicles presents promising opportunities to enhance traffic flow through real-time guided and cooperative management strategies. An in-depth empirical analysis of LCMs can unveil the circumstances under which drivers perform these maneuvers, aiding researchers in modeling these decisions and understanding the relationship between LCMs, congestion formation, congestion discharge (capacity drop), and accidents. This deeper understanding can inform the development of more effective traffic management strategies based on well-defined hypotheses. Furthermore, while microsimulation software plays a significant role in traffic flow studies, uncalibrated simulations lack accuracy and may lead investigations astray. Research efforts have predominantly focused on calibrating these microsimulators for longitudinal movements, while calibration for lateral movements has been largely overlooked. Comprehensive empirical studies can address this gap. Developing an LCM prediction model relies on thoroughly understanding why and when LCMs occur and their associated descriptive variables. Such insights can only be gained through empirical analysis, making lane-change and lateral maneuver studies essential research areas.

Modeling Lane-changes

This study presents a novel multi-class macroscopic lane change model constructed within a Nested Logit discrete choice framework. The model underwent estimation and validation processes using trajectory data collected from a busy weaving section in Antwerp, Belgium, revealing average and maximum rates of inaccurate predictions at 11% and 16%, respectively. The model's estimation of the total number of lane change maneuvers showed a margin of underestimation ranging from 4.1% to 7.7%. Addressing a gap in macroscopic lane change models primarily reliant on theoretical foundations, this research integrates empirical data derived from analyzing over 31,000 observed maneuvers. It introduces a unique alternative-specific discrete choice framework that distinguishes among three maneuver intents, allowing for the consideration of systematic taste variations among drivers. In contrast to data-driven models relying heavily on costly video-based data encompassing the entire traffic flow, this study proposes an innovative approach that estimates the position and frequency of maneuvers using a minimal fraction (1-2%) of vehicle trajectories supplemented by loop detector data. Furthermore, the model's practical usability is demonstrated in real-world traffic scenarios, and its integration into a macroscopic simulation ensures a highly consistent lane balance based on Scalable Quality Value statistics.

(Papers: Trans. Res. Part C-2024)

The Effect of Lane-changes on the Fundamental Traffic Flow Diagram

Lane Change Maneuvers have been demonstrated to impact road capacity negatively (Coifman et al., 2005). When a vehicle executes a lane change, it essentially occupies two lanes. Likewise, larger vehicles like buses and trucks exert a more pronounced influence on traffic flow than passenger cars. Over the years, traffic specialists have employed Passenger Car Equivalent (PCU) coefficients to account for this impact; however, a parallel effort has not been undertaken to quantify the effect of LCMs. The approximated trajectories facilitated the initiation of such an exploration.

Extreme Lane Change Behavior

Late lane changes, where a driver attempts to join a queue on an off-ramp or at a merging point, are often associated with traffic disruptions and can lead to safety hazards. Another example is radical (sudden) lane changes. Radical or sudden lane changes, where a driver makes abrupt and unexpected maneuvers to change lanes, can lead to shock waves in traffic flow and potentially result in accidents. The effect of such specific behavior might be a shock wave or, even more severe, lead to an accident.

Lateral Propagation of the Congestion Waves

Congestion might happen for different reasons and propagate longitudinally upstream or laterally from one lane to another. LCMs are the reason for lateral congestion propagation. The reach dataset of reconstructed trajectories facilitates this research.