Increasing the accuracy of loop detector counts using adaptive neural fuzzy inference system and genetic programming

Loop detectors are devices that are most commonly used for obtaining data at intersections. Multiple detectors are usually required to monitor a location, and this reduces the accuracy of detectors for collecting traffic volumes. The purpose of this paper is to increase the accuracy of loop detector counts using Adaptive Neural Fuzzy Inference System (ANFIS) and Genetic Programming (GP) based on detector volume and occupancy. These methods do not need microscopic analysis and are easy to employ. Four approaches for one intersection are used in a case study. Results show that the models can improve intersection detector counts significantly. Results also show that ANFIS produces more accurate counts compared to regression and GP.     

Using stop bar detector information to determine turning movement proportions in shared lanes

Turning vehicle volumes at signalized intersections are critical inputs for various transportation studies such as level of service, signal timing, and traffic safety analysis. There are various types of detectors installed at signalized intersections for control and operation. These detectors have the potential of producing volume estimates. However, it is quite a challenge to use such detectors for conducting turning movement counts in shared lanes. The purpose of this paper was to provide three methods to estimate turning movement proportions in shared lanes. These methods are characterized as flow characteristics (FC), volume and queue (VQ) length, and network equilibrium (NE). FC and VQ methods are based on the geometry of an intersection and behavior of drivers. The NE method does not depend on these factors and is purely based on detector counts from the study intersection and the downstream intersection. These methods were tested using regression and genetic programming (GP). It was found that the hourly average error ranged between 4 and 27% using linear regression and 1 to 15% using GP. A general conclusion was that the proposed methods have the potential of being applied to locations where appropriate detectors are installed for obtaining the required data. Copyright © 2016 John Wiley & Sons, Ltd.     

Real-time estimation of turning movement counts at signalized intersections using signal phase information

A variety of sensor technologies, such as loop detectors, traffic cameras, and radar have been developed for real-time traffic monitoring at intersections most of which are limited to providing link traffic information with few being capable of detecting turning movements. Accurate real-time information on turning movement counts at signalized intersections is a critical requirement for applications such as adaptive traffic signal control. Several attempts have been made in the past to develop algorithms for inferring turning movements at intersections from entry and exit counts; however, the estimation quality of these algorithms varies considerably. This paper introduces a method to improve accuracy and robustness of turning movement estimation at signalized intersections. The new algorithm makes use of signal phase status to minimize the underlying estimation ambiguity. A case study was conducted based on turning movement data obtained from a four-leg signalized intersection to evaluate the performance of the proposed method and compare it with two other existing well-known estimation methods. The results show that the algorithm is accurate, robust and fairly straightforward for real world implementation.     

Real-time estimation of turning movement proportions from partial counts on urban networks

Modern traffic signal control systems require reliable estimates of turning flows in real time to formulate effective control actions, and accommodate disturbances in traffic demand without deteriorating the system performance. The more accurate the estimation is, the more effective the control plan is. Most of the previous research works assumed that a full set of detector counts is available and employed the least-squares methods to produce unbiased estimates of the turning movement proportions. However, in practice, such a dense detector configuration is expensive to install and maintain. Also, the least-squares estimates are not feasible when the travel time between inflows and outflows is significant, or when intervening traffic conditions change the travel time. This study proposes a nonlinear least-square (NLS) approach and a quasi maximum likelihood (QML) approach to recursively estimate turning movement proportions in a network of intersections where only a partial set of detector counts are available. Using large population approximation technique, a class of nonlinear, discrete-time traffic flow models are transformed into a linear state–space model tractable for on-line applications. The quality of estimates is demonstrated by implementing the proposed algorithms with simulation and real data. As a comparison, the NLS estimator shows less bias but with higher variance than the QML estimator. The QML estimator outperforms the NLS estimator in terms of total mean square error, due to an increase in bias being traded for a decrease in variance.     

An intersection turning movement estimation procedure based on path flow estimator

Estimation of intersection turning movements is one of the key inputs required for a variety of transportation analysis, including intersection geometric design, signal timing design, traffic impact assessment, and transportation planning. Conventional approaches that use manual techniques for estimation of turning movements are insensitive to congestion. The drawbacks of the manual techniques can be amended by integrating a network traffic model with a computation procedure capable of estimating turning movements from a set of link traffic counts and intersection turning movement counts. This study proposes using the path flow estimator, originally used to estimate path flows (hence origin–destination flows), to derive not only complete link flows, but also turning movements for the whole road network given some counts at selected roads and intersections. Two case studies using actual traffic counts are used to demonstrate the proposed intersection turning movement estimation procedure. Copyright © 2010 John Wiley & Sons, Ltd.     

Vehicle reidentification and travel time measurement across freeway junctions using the existing detector infrastructure

Conventional vehicle detectors are capable of monitoring discrete points along the freeway but do not provide information about conditions on the link between detectors. Knowledge of conditions on the link is useful to operating agencies for enabling timely decisions in response to various delay causing events and hence to reduce the resulting congestion of the freeway system. This paper presents an approach that matches vehicle measurements between detector stations to provide information on the conditions over the link between the detectors rather than relying strictly on the aggregate point measurements from the detectors. In particular this work reidentifies measurements from distinct vehicles using the existing loop detector infrastructure. Here the distinct vehicles are the long vehicles, but depending on the vehicle population or type of detector used, one might chose to use some other reproducible feature.This new methodology represents an important advancement over preceding loop based vehicle reidentification, as illustrated herein, it enables vehicle reidentification across a major diverge and a major merge. The examples include a case where the reidentification algorithm responded to delay between two detector stations an hour before the delay was locally observable at either of the stations used for reidentification. While previous loop based reidentification work was limited to dual loop detectors, the present effort also extends the methodology to single loop detectors; thereby making it more widely applicable. Although the research uses loop detector data, the algorithm would be equally applicable to data obtained from many other traffic detectors that provide reproducible vehicle features.     

Average duration and performance of actuated signal phases

This paper describes an approach for evaluating alternative traffic detection designs for a signalized intersection. The models described in this paper can be used to determine the average phase duration and frequency of phase “max-out” as a function of the detector loop layout, detector unit timing, traffic demand, and approach speed. Layout and timing are described by the number of detectors on each approach served by the phase, detector location on each approach, detector length, and detector unit and controller time settings. The authors have used the concept of maximum allowable headway (MAH) to combine the many possible combinations of layout and timing variables into one representative quantity, which greatly simplifies the modelling process. The performance models were used to examine the sensitivity of intersection performance to a range of design values. In general, both phase duration and cycle length increase with higher demands or larger MAHs. Multiloop (i.e. two or more detection zones per lane) detector designs typically have larger MAHs than designs with one detector loop per lane. Phase duration and cycle length also increase for very low demand levels. In terms of performance, the maximum green duration was found to have a contrary effect at higher flow conditions. Larger maximum greens were found to reduce delays to the phase in service by reducing the probability of max-out but they increased delays to drivers waiting for service.     

Data fusion algorithm for macroscopic fundamental diagram estimation

A promising framework that describes traffic conditions in urban networks is the macroscopic fundamental diagram (MFD), relating average flow and average density in a relatively homogeneous urban network. It has been shown that the MFD can be used, for example, for traffic access control. However, an implementation requires an accurate estimation of the MFD with the available data sources.Most scientific literature has considered the estimation of MFDs based on either loop detector data (LDD) or floating car data (FCD). In this paper, however, we propose a methodology for estimating the MFD based on both data sources simultaneously. To that end, we have defined a fusion algorithm that separates the urban network into two sub-networks, one with loop detectors and one without. The LDD and the FCD are then fused taking into account the accuracy and network coverage of each data type. Simulations of an abstract grid network and the network of the city of Zurich show that the fusion algorithm always reduces the estimation error significantly with respect to an estimation where only one data source is used. This holds true, even when we account for the fact that the probe penetration rate of FCD needs to be estimated with loop detectors, hence it might also include some errors depending on the number of loop detectors, especially when probe vehicles are not homogeneously distributed within the network.     

Length-based vehicle classification using event-based loop detector data

Length-based vehicle classification is an important topic in traffic engineering, because estimation of traffic speed from single loop detectors usually requires the knowledge of vehicle length. In this paper, we present an algorithm that can classify vehicles passing by a loop detector into two categories: long vehicles and regular cars. The proposed algorithm takes advantage of event-based loop detector data that contains every vehicle detector actuation and de-actuation “event”, therefore time gaps between consecutive vehicles and detector occupation time for each vehicle can be easily derived. The proposed algorithm is based on an intuitive observation that, for a vehicle platoon, longer vehicles in the platoon will have relatively longer detector occupation time. Therefore, we can identify longer vehicles by examining the changes of occupation time in a vehicle platoon. The method was tested using the event-based data collected from Trunk Highway 55 in Minnesota, which is a high speed arterial corridor controlled by semi-actuated coordinated traffic signals. The result shows that the proposed method can correctly classify most of the vehicles passing by a single loop detector.     

An Adaptive Neuro-Fuzzy Inference System for estimating the number of vehicles for queue management at signalized intersections

Queue management is a valuable but underutilized technique which could be used to minimize the negative impacts of queues during oversaturated traffic conditions. One of the main obstacles of applying queue management techniques along signalized arterials is the unavailability of a robust and sufficiently accurate method for measuring the number of vehicles approaching a signalized intersection. The method based on counting vehicles as they enter and exit a specific detection zone with check-in and check-out detectors is unreliable because of the likely systematic under or over counting and the resulting cumulative errors. This paper describes the application of the Adaptive Neuro-Fuzzy Inference System (ANFIS) in the development of a new fuzzy logic-based approach for estimating the Number of Vehicles in a Detection Zone (NVDZ) by using detector time-occupancy data (instead of detector counts). Microscopic simulation results are used to evaluate the accuracy of the NVDZ estimates. Tests were carried out to determine the transferability of a tuned Fuzzy Inference System (FIS) and to check the sensitivity of the calibrated FIS to detection coverage, the location of the detection zone relative to the signalized (bottleneck) intersection, the length of the detection zone, and different signal timings at the bottleneck intersection. Results show that the NVDZ estimation based on fuzzy logic seems to be a feasible approach. Although the primary objective of developing the NVDZ estimation technique has been queue management, other applications such as ramp metering and incident detection could potentially use the same technique.     

Macroscopic Fundamental Diagrams: A cross-comparison of estimation methods

This paper aims to cross-compare existing estimation methods for the Macroscopic Fundamental Diagram. Raw data are provided by a mesoscopic simulation tool for two typical networks that mimic an urban corridor and a meshed urban center. We mainly focus on homogenous network loading in order to fairly cross-compare the different methods with the analytical reference. It appears that the only way to estimate the MFD without bias is to have the full information of vehicle trajectories over the network and to apply Edie’s definitions. Combining information from probes (mean network speed) and loop detectors (mean network flow) also provides accurate results even for low sampling rate (<10%). Loop detectors fail to provide a good estimation for mean network speed or density because they cannot capture the traffic spatial dynamics over links. This paper proposes a simple adjustment technic in order to reduce the discrepancy when only loop detectors are available.     

Speed estimation and length based vehicle classification from freeway single-loop detectors

Roadway usage, particularly by large vehicles, is one of the fundamental factors determining the lifespan of highway infrastructure. Operating agencies typically employ expensive classification stations to monitor large vehicle usage. Meanwhile, single-loop detectors are the most common vehicle detector and many new, out-of-pavement detectors seek to replace loop detectors by emulating the operation of single-loop detectors. In either case, collecting reliable length data from these detectors has been considered impossible due to the noisy speed estimates provided by conventional data aggregation at single-loop detectors. This research refines non-conventional techniques for estimating speed at single-loop detectors, yielding estimates that approach the accuracy of a dual-loop detector’s measurements. Employing these speed estimation advances, this research brings length based vehicle classification to single-loop detectors (and by extension, many of the emerging out-of-pavement detectors). The classification methodology is evaluated against concurrent measurements from video and dual-loop detectors. To capture higher truck volumes than empirically observed, a process of generating synthetic detector actuations is developed. By extending vehicle classification to single-loop detectors, this work leverages the existing investment deployed in single-loop detector count stations and real-time traffic management stations. The work also offers a viable treatment in the event that one of the loops in a dual-loop detector classification station fails and thus, also promises to improve the reliability of existing classification stations.     

Improved vehicle classification from dual-loop detectors in congested traffic

Vehicle classification is an important traffic parameter for transportation planning and infrastructure management. Length-based vehicle classification from dual loop detectors is among the lowest cost technologies commonly used for collecting these data. Like many vehicle classification technologies, the dual loop approach works well in free flow traffic. Effective vehicle lengths are measured from the quotient of the detector dwell time and vehicle traversal time between the paired loops. This approach implicitly assumes that vehicle acceleration is negligible, but unfortunately at low speeds this assumption is invalid and length-based classification performance degrades in congestion.To addresses this problem, we seek a solution that relies strictly on the measured effective vehicle length and measured speed. We analytically evaluate the feasible range of true effective vehicle lengths that could underlie a given combination of measured effective vehicle length, measured speed, and unobserved acceleration at a dual loop detector. From this analysis we find that there are small uncertainty zones where the measured length class can differ from the true length class, depending on the unobserved acceleration. In other words, a given combination of measured speed and measured effective vehicle length falling in the uncertainty zones could arise from vehicles with different true length classes. Outside of the uncertainty zones, any error in the measured effective vehicle length due to acceleration will not lead to an error in the measured length class. Thus, by mapping these uncertainty zones, most vehicles can be accurately sorted to a single length class, while the few vehicles that fall within the uncertainty zones are assigned to two or more classes. We find that these uncertainty zones remain small down to about 10 mph and then grow exponentially as speeds drop further.Using empirical data from stop-and-go traffic at a well-tuned loop detector station the best conventional approach does surprisingly well; however, our new approach does even better, reducing the classification error rate due to acceleration by at least a factor of four relative to the best conventional method. Meanwhile, our approach still assigns over 98% of the vehicles to a single class.     

Short-term travel time prediction

Effective prediction of travel times is central to many advanced traveler information and transportation management systems. In this paper we propose a method to predict freeway travel times using a linear model in which the coefficients vary as smooth functions of the departure time. The method is straightforward to implement, computationally efficient and applicable to widely available freeway sensor data.We demonstrate the effectiveness of the proposed method by applying the method to two real-life loop detector data sets. The first data set––on I-880––is relatively small in scale, but very high in quality, containing information from probe vehicles and double loop detectors. On this data set the prediction error ranges from 5% for a trip leaving immediately to 10% for a trip leaving 30 min or more in the future. Having obtained encouraging results from the small data set, we move on to apply the method to a data set on a much larger spatial scale, from Caltrans District 12 in Los Angeles. On this data set, our errors range from about 8% at zero lag to 13% at a time lag of 30 min or more. We also investigate several extensions to the original method in the context of this larger data set.     

Estimating arterial link speed using conventional road detectors

Abstract

In response to an initiative to develop an advanced traffic information system in Bangkok, this paper explores practical guidelines for the optimal location of road sensors, such that the data collected on spot speeds reflect an entire link's average speed. In particular, the authors use microsimulation software to investigate optimal detector locations, using the sum of squared errors and root mean squared errors. The analysis hypothesizes that road segments are 0.4, 0.6, 0.8, 1.0, 2.0 and 3.0 km in length and are specially designed to replicate typical arterial streets in Bangkok. The results show that a single detector location can produce good estimates of link speed only for segments that are shorter than 1.0 km. For distances of 1.0 km or more, the results suggest that two detectors be used for good link speed estimates under all traffic conditions.     

A gradient boosting logit model to investigate driver’s stop-or-run behavior at signalized intersections using high-resolution traffic data

Driver’s stop-or-run behavior at signalized intersection has become a major concern for the intersection safety. While many studies were undertaken to model and predict drivers’ stop-or-run (SoR) behaviors including Yellow-Light-Running (YLR) and Red-Light-Running (RLR) using traditional statistical regression models, a critical problem for these models is that the relative influences of predictor variables on driver’s SoR behavior could not be evaluated. To address this challenge, this research proposes a new approach which applies a recently developed data mining approach called gradient boosting logit model to handle different types of predictor variables, fit complex nonlinear relationships among variables, and automatically disentangle interaction effects between influential factors using high-resolution traffic and signal event data collected from loop detectors. Particularly, this research will first identify a series of related influential factors including signal timing information, surrounding traffic information, and surrounding drivers’ behaviors using thousands drivers’ decision events including YLR, RLR, and first-to-stop (FSTP) extracted from high-resolution loop detector data from three intersections. Then the research applies the proposed data mining approach to search for the optimal prediction model for each intersection. Furthermore, a comparison was conducted to compare the proposed new method with the traditional statistical regression model. The results show that the gradient boosting logit model has superior performance in terms of prediction accuracy. In contrast to other machine learning methods which usually apply ‘black-box’ procedures, the gradient boosting logit model can identify and rank the relative importance of influential factors on driver’s stop-or-run behavior prediction. This study brings great potential for future practical applications since loops have been widely implemented in many intersections and can collect data in real time. This research is expected to contribute to the improvement of intersection safety significantly.     

Recursive estimation of average vehicle time headway using single inductive loop detector data

Vehicle time headway is an important traffic parameter. It affects roadway safety, capacity, and level of service. Single inductive loop detectors are widely deployed in road networks, supplying a wealth of information on the current status of traffic flow. In this paper, we perform Bayesian analysis to online estimate average vehicle time headway using the data collected from a single inductive loop detector. We consider three different scenarios, i.e. light, congested, and disturbed traffic conditions, and have developed a set of unified recursive estimation equations that can be applied to all three scenarios. The computational overhead of updating the estimate is kept to a minimum. The developed recursive method provides an efficient way for the online monitoring of roadway safety and level of service. The method is illustrated using a simulation study and real traffic data.     

A real-time adaptive signal control in a connected vehicle environment

The state of the practice traffic signal control strategies mainly rely on infrastructure based vehicle detector data as the input for the control logic. The infrastructure based detectors are generally point detectors which cannot directly provide measurement of vehicle location and speed. With the advances in wireless communication technology, vehicles are able to communicate with each other and with the infrastructure in the emerging connected vehicle system. Data collected from connected vehicles provides a much more complete picture of the traffic states near an intersection and can be utilized for signal control. This paper presents a real-time adaptive signal phase allocation algorithm using connected vehicle data. The proposed algorithm optimizes the phase sequence and duration by solving a two-level optimization problem. Two objective functions are considered: minimization of total vehicle delay and minimization of queue length. Due to the low penetration rate of the connected vehicles, an algorithm that estimates the states of unequipped vehicle based on connected vehicle data is developed to construct a complete arrival table for the phase allocation algorithm. A real-world intersection is modeled in VISSIM to validate the algorithms. Results with a variety of connected vehicle market penetration rates and demand levels are compared to well-tuned fully actuated control. In general, the proposed control algorithm outperforms actuated control by reducing total delay by as much as 16.33% in a high penetration rate case and similar delay in a low penetration rate case. Different objective functions result in different behaviors of signal timing. The minimization of total vehicle delay usually generates lower total vehicle delay, while minimization of queue length serves all phases in a more balanced way.     

A simulation-based approach in determining permitted left-turn capacities

A fundamental objective of traffic signal operations is the development of phasing plans that reduce delays while maintaining a high level of safety. One issue of concern is the treatment of left-turn phasing, which can operate as a protected movement, a permitted movement yielding to conflicting traffic, a combination protected–permitted movement or as a split-phase intersection. While protected-only movements can improve safety for the turning movement, they can also increase delays and congestion at the intersection. Most states maintain independent guidance for determining left-turn phasing; however, the most common identified guidance for protected left-turn phases is using a threshold based on the cross product of the left-turn volume and opposing through movements. The use of the cross product has been questioned recently as an indicator for determining phase selection. Based on simulation analysis within this research, the cross product is shown to be a poor indicator of left-turn capacity and congestion at the intersection.This research proposes a simplified single variable exponential model to determine left-turn capacity based on opposing volume and percent green time to determine left-turn capacity thresholds for protected left-turn phasing. The model is developed based on observed capacity from 450 VISSIM microsimulation scenarios which evaluated varying opposing volume, opposing number of lanes, cycle lengths and green time splits. Validation of the model based on complex Highway Capacity Manual procedures, indicates that the proposed model provides similar correlation to observed capacities. Finally, a nomograph is developed which presents the model in a simple form for interpretation and application by practicing traffic engineers, when required to determine left-turn phasing options. This procedure allows simple determination based on minimum input data needs similar to the cross product determination, without the need for complex hand calculations or computing requirements of the Highway Capacity Manual.     

Vehicle reidentification and travel time measurement on congested freeways

The paper presents an algorithm for matching individual vehicles measured at a freeway detector with the vehicles’ corresponding measurements taken earlier at another detector located upstream. Although this algorithm is potentially compatible with many vehicle detector technologies, the paper illustrates the method using existing dual-loop detectors to measure vehicle lengths. This detector technology has seen widespread deployment for velocity measurement. Since the detectors were not developed to measure vehicle length, these measurements can include significant errors. To overcome this problem, the algorithm exploits drivers’ tendencies to retain their positions within dense platoons. The otherwise complicated task of vehicle reidentification is carried out by matching these platoons rather than individual vehicles. Of course once a vehicle has been matched across neighboring detector stations, the difference in its arrival time at each station defines the vehicle’s travel time on the intervening segment.Findings from an application of the algorithm over a 1/3 mile long segment are presented herein and they indicate that a sufficient number of vehicles can be matched for the purpose of traffic surveillance. As such, the algorithm extracts travel time data without requiring the deployment of new detector technologies. In addition to the immediate impacts on traffic monitoring, the work provides a means to quantify the potential benefits of emerging detector technologies that promise to extract more detailed information from individual vehicles.