Alexander Skabardonis

Detecting Errors and Imputing Missing Data for Single-Loop Surveillance Systems

Single-loop detectors provide the most abundant source of traffic data in California, but loop data samples are often missing or invalid. A method is described that detects bad data samples and imputes missing or bad samples to form a complete grid of clean data, in real time. The diagnostics algorithm and the imputation algorithm that implement this method are operational on 14,871 loops in six districts of the California Department of Transportation. The diagnostics algorithm detects bad (malfunctioning) single-loop detectors from their volume and occupancy measurements. Its novelty is its use of time series of many samples, instead of basing decisions on single samples, as in previous approaches. The imputation algorithm models the relationship between neighboring loops as linear and uses linear regression to estimate the value of missing or bad samples. This gives a better estimate than previous methods because it uses historical data to learn how pairs of neighboring loops behave. Detection of bad loops and imputation of loop data are important because they allow algorithms that use loop data to perform analysis without requiring them to compensate for missing or incorrect data samples.

Travel-Time Reliability as a Measure of Service

Statistics from a corridor along Interstate 5 in Los Angeles show that average travel time and travel-time variability are meaningful measures of freeway performance. Variability of travel time is an important measure of service quality for travelers. Travel time can be used to quantify the effect of incidents, and incident information can help reduce travel-time uncertainty. Predictability of travel time is a measure of the benefits of intelligent transportation systems. These measures differ from those defined in the Highway Capacity Manual and other aggregate measures of delay.

Freeway performance Measurement System. Mining loop detector data

Performance Measurement System (PeMS) is a freeway performance measurement system for all of California. It processes 2 GB/day of 30-s loop detector data in real time to produce useful information. At any time managers can have a uniform, comprehensive assessment of freeway performance. Traffic engineers can base their operational decisions on knowledge of the current status of the freeway network. Planners can determine whether congestion bottlenecks can be alleviated by improving operations or by minor capital improvements. Travelers can obtain the current shortest route and travel time estimates. Researchers can validate their theory and calibrate simulation models. PeMS, which has been in stable operation for 18 months, is a low-cost system. It uses the California Department of Transportation (Caltrans) network for data acquisition and is easy to deploy and maintain. It takes under 6 weeks to bring a Caltrans district online, and functionality can be added incrementally. PeMS applications are accessed over the World Wide Web; custom applications can work directly with the PeMS database. Built as a prototype, PeMS can be transitioned into a 7 × 24 production system. The PeMS architecture and use are described.

Real-Time Monitoring and Control on Signalized Arterials

An analytical model for real-time estimation of travel times along signalized arterials was developed. The application of the model on two arterial sites and comparisons of the estimated travel times with simulated and field data show that the model accurately predicts travel times on the selected sites. In this article, we present several important extensions and refinements to the model including treatment of long queues and queue spillovers, and algorithms for signal priority to transit vehicles. We also describe the integration of the model into an archival data management system for continuous measurement and monitoring of traffic performance along arterials.

GUIDELINES FOR CALIBRATION OF MICROSIMULATION MODELS: FRAMEWORK AND APPLICATIONS

The past few years have seen a rapid evolution in the sophistication of traffic microsimulation models and a consequent major expansion of their use in transportation engineering and planning practice. Researchers and practitioners have employed an extensive array of approaches to calibrate these models and have selected a wide range of parameters to calibrate and a broad range of acceptance criteria. A methodical, top-down approach to model calibration is outlined; it focuses the initial effort on a few key parameters that have the greatest impact on model performance and then proceeds to less critical parameters to finalize the calibration. A three-step calibration/validation process is recommended. First, the model is calibrated for capacity at the key bottlenecks in the system (the capacity calibration step). Second, the model is calibrated for traffic flows at nonbottleneck locations in the system (the route choice calibration step). Finally, the overall model performance is calibrated against field-measured system performance measures such as travel time and delay (the system performance calibration step). This three-step process is illustrated in an example application for a freeway/arterial corridor.

SYSTEMATIC IDENTIFICATION OF FREEWAY BOTTLENECKS

The algorithm presented identifies bottleneck locations, the times for which each bottleneck is active, and the delay they cause. The bottlenecks are ranked in terms of their frequency of recurrence and the magnitude of their delay impact. The algorithm works with 5-min loop detector data. It uses speed difference as an indicator of bottleneck activation. The algorithm is applied to 3 months of data from 270 mi of seven freeways in San Diego. It identifies 160 locations whose bottlenecks cause 64% of the total delay on these freeways. The top 10 account for 61% of the delay from all bottlenecks. The method can be used in any area in which large amounts of data are available. Transportation authorities may use it to identify bottlenecks and to track their impact over time.

I-880 Field Experiment: Data-Base Development and Incident Delay Estimation Procedures

The collection and processing of field data on freeway incidents and operating conditions are described, and the development and application of a methodology for estimating incident delays are presented. The data were collected on a section of the I-880 freeway in the San Francisco Bay Area before and after the implementation of freeway service patrols (FSPs). Incident characteristics were obtained through observations of probe vehicle drivers traveling at an average of 7-min headways. Travel times were obtained from the specially instrumented probe vehicles. Speeds, flows, and occupancies at 1-sec intervals were collected from closely spaced loop detectors on the freeway main line and the ramps. Software was developed to process the data and create a computerized data base. The I-880 data base consists of 276 hours of field data that are uniquely linked to provide a complete representation of the freeway operating conditions at the test site. Improved procedures for estimating incident delay and other pe...

Freeway Performance Measurement System: Operational Analysis Tool

The freeway performance measurement system (PeMS) is a system for all of California. It processes 2 gigabytes/day of 30-s loop detector data in real time to produce useful information. Managers at any time can have a uniform and comprehensive assessment of freeway performance. Traffic engineers can base their operational decisions on knowledge of the current state of the freeway network. Planners can determine whether congestion bottlenecks can be alleviated by improving operations or by minor capital improvements. Travelers can obtain the current shortest route and travel time estimates. Researchers can validate their theory and calibrate simulation models. The use of PeMS in conducting operational analysis, planning, and research studies is described here. The advantages of PeMS over conventional study approaches are demonstrated from case studies on conducting freeway operational analyses, bottleneck identification, level of service determination, assessment of incident impacts, and evaluation of advanced control strategies.

Understanding Stop-and-go Traffic in View of Asymmetric Traffic Theory

Stop-and-go traffic is a frequently observed phenomenon in congested highway traffic, but it has not been accurately modeled in existing traffic models. Car-following models based on kinematic flow theory cannot model stop-and-go traffic. Other approach assumed traffic states deviating from the equilibrium curve in the fundamental diagram, and the transitions between them, but no explanation was provided on the reason for the existence of different states. There is a need to understand the mechanism of stop-and-go traffic in terms of generation, propagation and dissipation in order to accurately model traffic dynamics. We propose an asymmetric traffic theory and explain the stop-and-go traffic phenomenon in light of the developed theory. The proposed theory is verified using individual vehicle trajectories from two freeway sites in California, US, collected as part of the Next Generation Simulation (NGSIM) project.

Prediction of Arrival Profiles and Queue Lengths Along Signalized Arterials by Using a Markov Decision Process

An analytical methodology for prediction of the platoon arrival profiles and queue length along signalized arterials is proposed. Traffic between successive traffic signals is modeled as a two-step Markov decision process (MDP). Traffic dynamics are modeled with the use of the kinematic wave theory. The MDP formulation allows prediction of the arrival profiles several signals downstream from a known starting flow. This modeling approach can be used to estimate queue lengths and predict travel times, even in cases in which data from loop detectors are unknown, inaccurate, or aggregated. The proposed model was applied to two real-world test sites. The queues estimated with the model are in close agreement with the results from microscopic simulation.