Lane Choice Research Articles

Toll Plaza Lane Choice and Lane Configuration Strategy for Autonomous Vehicles in Mixed Traffic

Toll plazas are one kind of highway node that pose a challenge to the flow of autonomous vehicles (AVs). Some toll roads have open road tolling (ORT), which makes it easier and more efficient for AVs, but most toll plazas do not have ORT. The authors developed a decision model to assist AVs in choosing the right lane that involves two main indicators, the rod speed (SRi) and the decision-making value (DSRi). The decision model was applied to a six-lane toll plaza with manual toll collection (MTC), electronic toll collection (ETC), and electronic toll collection for AVs toll types. The suitable lane configurations for the varying percentages of AVs in traffic were as follows: at 10% AV, the optimal lane configuration was 4 MTC and 2 ETC/AV (4M2EA), at 30% AV it was 3M3EA, at 40%–60% AV it was 3M2EA1A, and at 70% AV it was 2M2EA2A. Thus AVs should share toll lanes with ETC manned vehicles before the population of AVs reaches 30%; after that dedicated lanes may be introduced. It was also concluded that a 10% increase in AVs reduces delays by 9% if the populations of ETC and MTC vehicles are equal.

Performance Analyses of Information-Based Managed Lane Choice Decisions in a Connected Vehicle Environment

Connected vehicle (CV) technology can connect, communicate, and share information between vehicles, infrastructure, and other traffic management systems. Recent research has examined and promoted CV and connected automated vehicle (CAV) technology on managed lane systems to increase capacity and reduce congestion, as managed lane systems could be equipped with advanced infrastructure relatively quickly. However, the effect on travel considering, information-based managed lane choice decisions in a CV environment is not clear. Therefore, this research analyzed the potential effects on a managed lane system with connected vehicles considering several travel behavior elements, including drivers’ willingness to reroute and their choice of managed lanes based on individual travel time savings. This study analyzed the potential effects on a managed lane system by assigning different market penetration rates (0%, 10%, 50%, 100%) of CVs and informing CV drivers about travel time savings for a 10-mi stretch at 5-min intervals. How the traffic performance measurements (i.e., throughput, travel time saving, average speed and average travel time) vary under different market penetration rates of CVs is then investigated. Two major conclusions are reached: (i) although information exchange was assumed to be instantaneous between vehicles and the system, there existed a response time (or time delay) in the macroscopic traffic reflection; (ii) managed lane use may decrease, when travel time information becomes available, since drivers perceive they are saving more travel time than they actually do save.

Personalized Choice Model for Managed Lane Travel Behavior

This paper presents a methodology for enhancing discrete choice models for managed lane travel behavior with personal trip history. We refer to this process as personalization and the enhanced model as a personalized choice model. With the objective of better understanding managed lane choices and improving the model’s prediction capability, personalization was carried out at two levels. First, we used each traveler’s habits and travel experiences before each trip for constructing a set of explanatory variables that could be used with any model structure. Second, under a logit mixture framework, the distribution of random parameters was updated with Bayesian inference according to personal trip history. The structure of the parameter distribution explicitly considered preference variations across individuals (interpersonal heterogeneity), as well as preference variations across trips performed by the same individual (intrapersonal heterogeneity). The proposed methodology is especially relevant for modeling revealed preference (RP) data from automatic vehicle identification sensors, for which limited socioeconomic characteristics of travelers are available. An empirical study was conducted on an operational managed lane corridor near Dallas/Fort Worth Airport in Texas. Available trip records over a 5-month period were utilized. A hierarchical Bayes estimator was adopted for efficient model estimation. The results suggest significant inter- and intrapersonal heterogeneity and that the proposed personalization method improves the model’s explanatory power and prediction capability. To the best of our knowledge, this paper represents the first introduction of personalization in managed lane choice behavior modeling and the first attempt to estimate intrapersonal heterogeneity with RP data.

Dynamic lane restrictions on congested arterials

Buses operating in mixed traffic experience problems with delay and travel time variability in peak periods. Numerous transit priority measures are found in the literature but most rely on separating buses from other vehicles. Complete separation such as dedicated bus lanes might underutilize the system capacity when the frequency of buses is not high. We propose a “dynamic bus lane policy”, where control adjusts the proportion of cars to buses in one lane of a multi-lane arterial instead of enforcing complete separation of vehicle types. This strategy is analyzed with aggregated macroscopic dynamics for different models of lane choice. It is shown to benefit both cars and buses by maintaining higher outflow during the peak period and helping the system to recover more quickly at the end of the peak. Greater reductions in travel time are possible with increased information available to vehicles. Analytical derivations and non-linear optimization are then used to obtain the system optimal allocation of cars, as well as the toll needed to achieve it. The system optimal tolling would be difficult to implement in practice. Another optimization which includes tolls paid in its objective is nearly as good at reducing vehicle hours traveled and produces a much simpler pricing structure amenable to feedback control. A microsimulation with feedback control confirms the results of the macroscopic model and shows additional benefits attributable to driver behavior. Future work will look into varying control parameters to achieve a global minimum travel time and will incorporate user heterogeneity to test pricing strategies.

Driving with and without automation on the motorway – an observational study

User-related assessment of a level 3 automated driving system, providing functions such as lane- and distance-keeping, stop & go driving, lane change and overtaking, was carried out on a motorway in Germany with the aim to assess user-related issues of automated driving, i.e., behavior when driving with automation on motorways, user experiences, reactions and acceptance. Twenty-one persons drove twice along the test route once with the system switched off and once with the system active. Driving data were logged and driving behavior was observed by two observers in the car and the drivers answered questionnaires. The results revealed that the drivers used the system as it was intended to be used and that the system affected driving positively in several ways, resulting in better speed adaptation, less speed variation, better distance keeping, better lane choice, better indicator usage and fewer dangerous lane changes. No differences with regard to subjective workload were shown. The system was perceived as being both useful and satisfactory. However, it was found that the system did not react to other drivers’ intention to make a lane change, especially in situations where they were attempting to merge onto the motorway. Further development is needed to improve the system’s ability to recognize other vehicles’ intention to enter the vehicle’s own lane. Also, the phenomenon of the driver “feeling guilty” because of the system’s reckless behavior by not allowing other drivers to merge onto the motorway or hindering other cars behind due to lengthy overtaking should be investigated further.

Psychological items: a useful addition in modeling travel behavior on managed lanes

Managed lanes (MLs) are a tool to more efficiently operate segments of a freeway. As ML prevalence increases in the United States of America, it is important to understand travel behavior in a ML setting (i.e., lane choices and carpooling decisions). Socio-demographic and trip data, along with travel time and toll, have been commonly used in this endeavor. However, there are some travelers who pay to use the ML despite there being little to no improvement in travel time over the adjacent general purpose lanes. This gives rise to the possibility that psychological traits and characteristics are a greater influence on ML use than even travel time savings for some travelers. This research examined this issue through a set of largely transportation-framed psychological items. After an initial creation and refining process, 25 psychological items were included in a survey advertised in five cities with MLs. In addition to psychological items, trip and demographic questions, and three SP questions were included in the online survey. Mixed logit models were estimated based on survey responses obtained from three study areas. Models that included psychological items performed better (in terms of adjusted rho squared value and percent correctly predicted values) than models with only trip and demographic variables. Likewise, models including psychological items plus trip and demographic data performed best. This information may be useful for traffic and revenue estimating firms interested in potentially including psychological items in future ML surveys intended to facilitate better estimation of ML use.

Application of machine learning to characterize uneconomical managed lane choice behaviour

Models of tolled Managed Lane (ML) travel assume that people pay the toll to experience a shorter travel time than using the toll-free general-purpose lanes. However, there are times users pay to travel on the MLs but go slower than the toll-free lanes. This research examined these ‘uneconomical trips’ on MLs to discover potential similarities among these trips and help understand the unexpected lane choice behavior. Some potential factors considered were toll rate, traffic flow, and past trip experience.Random forest and logistic regression methods were proposed to study the impact and significance of variables on the probability of a user making an uneconomical ML trip. The results indicated toll rate, traffic flow, travel time variability, and trip route (start and end points) are the crucial elements in predicting uneconomical ML trips.This study is the first to examine uneconomical ML trips and attempt to understand these trips. This provides the first step towards being able to predict these trips and provides additional understanding of travel on MLs. Also, the findings of this study are the first step towards being able to include uneconomical ML trips into lane choice modeling and toll feasibility studies. Finally, this study proposed an unconventional tool (random forest) to predict the travel behavior on MLs and showed that such tools can perform better than currently used logistic regression models in predicting these uneconomical ML trips.

Heterogeneity aware urban traffic control in a connected vehicle environment: A joint framework for congestion pricing and perimeter control

Real-time control of large-scale urban networks has been attracting significant research attention. This paper, using the information provided by connected vehicles, proposes a novel control approach based on the concept of perimeter control to maximize the social welfare of all vehicles. The contributions of this paper are threefold. First, we consider vehicle heterogeneity (i.e. corresponding to different transportation modes, or with different occupancies, values of time, priority levels, etc.) and integrate a priority scheme into perimeter control to improve both the traffic performance and the social welfare. This is achieved by installing priority lanes at some of the perimeter intersections. Unlike the existing research works that provide priority to certain traffic modes, we dynamically identify the groups of vehicles that we should prioritize, in order to maximize the social welfare. Second, we develop a model predictive control approach to simultaneously optimize the toll for using the priority lanes and the traffic signal timings at the perimeter intersections. This approach can explicitly handle the constraint of the storage capacity of each intersection link. Third, we propose a recursive estimation algorithm to update our knowledge on the distribution of the value of times (VOTs), using the lane choice information of connected vehicles. The proposed approach is tested in a simulated network which resembles the main features of the city center of Zurich, Switzerland. By using the proposed strategy, the traffic accumulation inside the network is still stabilized, and the monetary costs due to delay are significantly reduced at the entire network (up to 25.8%) compared to the strategy without priority. The distribution of the combined cost (including cost due to delay and tolls) is more uniform across VOT groups than that resulting from the strategy without priority. It is also shown that the proposed recursive estimation algorithm quickly converges and further improves the social welfare.

Comparing Route Choice Models for Managed Lane Networks with Multiple Entrances and Exits

Developing an appropriate route choice model for managed lanes with multiple entrances and exits is critical for the success of managed lane planning and operations. This research focuses on route choice models for managed lane networks with stochastic and time-varying tolls and travel times. In the model, a traveler receives real-time information about the tolls and travel times upon arrival at each diverge node and makes a dynamic lane choice decision that minimizes the total expected cost. The online route choice model is formulated as a Markov decision process and solved using a backward recursion algorithm. The model is compared against four other routing models: a binary logit model, a model based on decision routes, a model that chooses paths a priori, and a model with routes chosen randomly. The study also models irrational driver behavior with parameters like driver’s inclination toward making optimal lane choices and preference for certain lanes. Findings show that the expected costs from the routes chosen using the decision route model from the literature are close to the optimal cost with an average percentage error of 0.93%. The binary logit model is shown to have a high average error of 50% in the expected cost when a driver is assumed to behave rationally, but the same model shows optimal prediction for certain irrational driver behaviors. The proposed routing model forms a basis for future work in the area of managed lane pricing and planning.

Close Following Behavior: Estimation of Desired Gap Headway Using Loop Detector Data

The desired gap headway of drivers, while close following, represents the main parameter in determining the following distance between vehicles. This paper uses the raw individual vehicles data taken from loop detectors for millions of vehicles used M25 and M42 in order to estimate the gap headway distributions between successive pairs of vehicles. The data used in this paper were filtered so as to focus on the cases of close following behavior only and more than quarter million pairs of close following cases is used to presents the results. Such huge sample size taken from loop detectors will increase the results reliability as previous research used limited sample size. The results presented the cumulative distribution of drivers’ gap headway and suggested that the mean gap headway of drivers is about 1.1 s with standard deviation of 0.42 s. The lane choice found to be significantly influencing the desired gap headway for speed higher than 70km/h only. The effect of the follower vehicle type of the desired gap headway was also examined and the results suggested that such effect is insignificant. The findings of this paper are suggested to be used as inputs for traffic micro-simulation models.

Empirical Observations and Formulations of Tri-Class Traffic Flow Properties for Design of Traffic Signals

Recognizing the popularity of scooters and buses in many developing countries and the significant impact of their physical maneuverability on the properties of mixed traffic flows, this study presents our findings from empirical observations of tri-class traffic flows, including the complex interactions among scooters, passenger cars, and buses from their discharging to the formation of stop queues at downstream intersections. Using camcorders mounted either on high-rise structures or drones, our empirical data include individual vehicles and their trajectories over time. Formulations based on empirical findings have also been proposed to model the behaviors of such tri-class flows on queue discharging, formation, lane choice in propagation, and filtering processes. Evaluations of field data, with respect to time-varying throughputs, queue clearance times, and queue length evolution, confirms the effectiveness of the proposed formulations. These can serve as the basis for developing mesoscopic tri-class flow simulation models or for designing signals for arterials plagued by such mixed flows.

The effect of information on drivers’ toll lane choices and travel times expectations

This paper reports on a study of drivers’ toll lane choices and the effect of the type and content of information they are provided with on their travel time expectations and lane choices. In the data collection experiment, participants were asked about their travel time expectations. A mixed-effects regression model is developed to predict these expected travel times depending on the toll rates and type and content of travel information they are provided with. Then, a model to predict the choice whether or not to use the toll lane, using the expected travel times as explanatory variables, is formulated and estimated. The results show that drivers’ expected travel times are affected by the information provided to them on the VMS. In particular, in the absence of precise travel time information drivers use the toll rate itself as an indicator to the expected travel times. In the choice model, there is a significant heterogeneity in preferences and the related values of time among drivers. In particular, there are large differences in values of time between drivers who pay the toll themselves and those whose employers pay the tolls.

Unbalanced Multiple Left Turn Lane Usage Modelling: From Individual Choice to Aggregate Volume

Diverse lane preferences of left-turn drivers lead to unbalanced traffic distribution on multiple left-turn lanes. The preferences can be measured in terms of lane usage at macroscopic level and individual lane choice at microscopic level. The data of lane volume and individual lane choices are collected at eight dual or triple left-turn lanes equipped in signalized intersections in China. Linear regression model with dummy variables and discrete choice model were applied to analyse drivers’ lane choosing patterns at macroscopic and microscopic levels, respectively, and results of the two studies are mutually verified and complemented. The drivers’ lane preferences are found to vary with approach configurations, traffic control, and the number of lanes available. Static influential factors, such as turning radius inside the intersection, the design of shadowed lane, and intersection skewedness, as well as dynamic influential factors, including queue length, heavy vehicle in queue back and subject vehicle type, could affect the drivers’ lane preferences. The findings of this study have important implications for intersection design and traffic control in practice.

A state dependent mandatory lane-changing model for urban arterials with hidden markov model method

The inherent intention and decision process of lane changes are complex. Though the external environment and traffic conditions are changing along the traveling direction, the drivers’ characteristics and preferences may lead to persistence of preferable lane choice. The lane-changing decision process is commonly separated into two steps after the decision to consider a lane change: lane choices and acceptance of available gaps, while the outcome of first step is unobservable. To better represent the mandatory lane-changing behaviors on arterials, a two-stage lane-changing model with Hidden Markov Model (HMM) method is proposed. The proposed model is estimated and validated using detail Next Generation Simulation (NGSIM) vehicle trajectory data from an urban arterial. Comparison between generated lane position sequences and original trajectories validated the model’s capability of representing mandatory lane changes.

Dynamic pricing strategy for high occupancy toll lanes based on random forest and nested model

To utilise high occupancy vehicle lanes better, high occupancy toll (HOT) lanes are introduced to counter against congestion on the urban highways. In such system, low occupancy vehicles (LOVs) are allowed to pay a toll and access to HOT lanes from general purpose (GP) lanes, so that the toll rate plays a key role in dynamically allocating LOVs over the HOT and GP lanes to improve the overall system performance. First, this study presents an improved random forest (RF) method to build the lane choice behaviour prediction model. Then, by using the 5 min historical traffic and toll data collected from Interstate 405 in the USA, the improved RF combined with cross-validation and grid search shows the highest accuracy of 88.7%, which is better than other four methods. Furthermore, a novel nested model with two levels is proposed to optimise the toll rates under different real-time traffic conditions. For the nested model, the experimental results show that the proposed dynamic pricing strategy can decrease the total delay and improve the efficiency significantly. To realise the pricing strategy, some Intelligent Transportation System technologies for HOT lane systems are described in detail and designed as the fundamental of the pricing strategy.

Dynamic pricing for managed lanes with multiple entrances and exits

Priced managed lanes are increasingly being used to better utilize the existing capacity of the roadway to relieve congestion and offer reliable travel time to road users. In this paper, we investigate the optimization problem for pricing managed lanes with multiple entrances and exits which seeks to maximize the revenue and minimize the total system travel time (TSTT) over a finite horizon. We propose a lane choice model where travelers make online decisions at each diverge point considering all routes on a managed lane network. We formulate the problem as a deterministic Markov decision process and solve it using the value function approximation (VFA) method for different initializations. We compare the performance of the toll policies predicted by the VFA method against the myopic revenue policy which maximizes the revenue only at the current timestep and two heuristic policies based on the measured densities on the managed and general purpose lanes (GPLs). We test the results on four different test networks. The primary findings from our research suggest the usefulness of the VFA method for determining dynamic tolls. The best-found objective value from the method at its termination is better than other heuristics for all test networks with average improvements in the objective ranging between 10% and 90% for revenue maximization and 0–27% for TSTT minimization. Certain VFA initializations obtain best-found toll profiles within first 5–50 iterations which warrants computational time savings. Our findings also indicate that the revenue-maximizing optimal policies follow the “jam-and-harvest” behavior where the GPLs are pushed towards congestion in the earlier time steps to generate higher revenue in the later time steps, a characteristic not observed for the policies minimizing TSTT.

Evaluation of Downstream Merge Behaviors Resulting from Driver Lane Choice: A Driving Simulator Study

With the ever-increasing demand to add roadway capacity in a safe and efficient manner, the application of auxiliary through lanes (ATLs) at intersections has increased in recent years. Despite the potential capacity-related benefits, ATLs also introduce the potential for unique driver operation tactics, leading to challenging merge scenarios downstream of signalized intersections. Furthermore, the benefits of increasing capacity are only realized when drivers are making decisions that balance the utilization of the ATL and the adjacent continuous through lane. Moreover, balancing lane utilization upstream of an ATL signalized intersection may introduce an improvement in safety and operations in merging conflicts downstream. Previous research investigated the effectiveness of ATLs through microsimulation and field studies. This research employed two full-immersion driving simulation studies. The initial experiment was conducted using simulation technology to evaluate aspects of driver behavior associated with lane utilization at ATLs. Lane utilization of the ATL significantly increased when the driver was informed of the fundamental design concept, ultimately improving capacity. The second experiment was based upon the results of the initial experiment, coupled with ongoing research evaluating alternative merge signage. Specifically, the second experiment aimed to evaluate downstream merge signage that could mitigate some of the behaviors resulting from the less than optimal lane utilization for ATLs. With an emphasis on improving intersection performance, the results from this paper suggest that the implementation of alternative merging signage may lead to improved merging behavior.

Choice of High Occupancy/Toll Lanes under Alternative Pricing Strategies

High occupancy/toll (HOT) lanes typically vary tolls charged to single occupant vehicles, with the toll increasing during congested periods. The toll is usually tied to time of day or to the density of vehicles in the HOT lane. The purpose of raising the toll with congestion is to discourage demand sufficiently to maintain travel speeds in the HOT lane. However, it has been demonstrated that the HOT toll may act as a signal of downstream congestion (in both general purpose (GP) and HOT lanes), causing an increase in demand for the HOT lane, at least at lower prices. This paper develops a model of lane choice to evaluate alternative HOT lane pricing strategies, including the use of GP density, to more accurately reflect the value of the HOT lane. In addition, the paper explores the potential effect these strategies would have on the HOT lane vehicle share through a partial equilibrium analysis. This analysis demonstrates the change in demand elasticity with price, showing the point at which drivers switch from a positive to negative elasticity.

Effects of Weather Conditions on Motorway Lane Flow Distributions

Several factors affect the lane choices made by motorway drivers. According to the driving rules, the nearside lane is the one that is primarily used. The main reasons for lane changes are overtaking, congestion, or restrictions on other lanes. The empirical research presented in this paper presents comprehensive traffic characteristics observed in different traffic lanes on four-lane motorways in Slovenia. The research was focused on the influence of adverse weather conditions on the lane flow distribution, and on the speed of vehicles in different lanes. The lane flow and speed distributions both directly affect road capacity and safety; therefore, estimating these characteristics could improve the reliability of active traffic control when traffic flow perturbation is detected. Field test results show that lane flow distributions and lane speed distributions at a particular site vary depending on weather conditions, namely, dry, wet (rain), low-visibility, and snow conditions.

Empirical study of lane-changing behavior on three Chinese freeways.

Lane-changing (LC) behavior is investigated on Chinese freeways, where the driving circumstances are relatively aggressive. Three data sets were collected from urban expressways and an intercity highway in the form of traffic videos. Different aspects of LC behaviors are analyzed, i.e., the LC rate, motivation, target lane choice and impact on traffic. The results suggest that LC is a transient behavior that randomly occurs with high frequency, which is the main feature of aggressive driving. Several LC patterns and the combination effect of ramps, fast lanes and various vehicle types are presented. The influence of LC on local traffic endures for approximately 15 to 30 s, which rapidly increases and slowly declines. LC behavior will increase the risk of high-speed car-following. All results are obtained from the empirical data; they will be useful for traffic management and traffic modeling.