Thomas Urbanik Ii

Short-Term Freeway Traffic Volume Forecasting Using Radial Basis Function Neural Network

A radial basis function (RBF) neural network has recently been applied to time-series forecasting. The test results of an RBF neural network in forecasting short-term freeway traffic volumes are provided. Real observations of freeway traffic volumes from the San Antonio TransGuide System have been used in these experiments. For comparison of forecasting performances, Taylor series, exponential smoothing method (ESM), double exponential smoothing method, and backpropagation neural network were also designed and tested. The RBF neural network model provided the best performance and required less computational time than BPN. It seems that RBF and ESM can be a viable forecasting routine for advanced traffic management systems. There are some tradeoffs between RBF and ESM. Although the performance of ESM is inferior to RBF, the former does not need a complicated training process or historic database, and vice versa. However, even in the best performance case, 35 percent of the forecast traffic volumes showed 10 percent or more percentage errors. This means that we cannot heavily depend on the forecast traffic volumes as long as we are utilizing the models tested. Further work is needed to provide a more reliable traffic forecasting model.

Traffic Signal Optimization Program for Oversaturated Conditions: Genetic Algorithm Approach

Traffic signal optimization programs have been used widely among transportation professionals. However, none of the existing computer programs can optimize all four traffic control parameters (i.e., cycle length, green split, offset, and phase sequence) simultaneously, even for undersaturated conditions. In this paper, a genetic algorithm-based signal optimization program that can handle oversaturated signalized intersections is presented. The program consists of a genetic algorithm (GA) optimizer and a mesoscopic traffic simulator. The GA optimizer is designed to search for a near-optimal traffic signal timing plan on the basis of a fitness value obtained from the mesoscopic simulator. The proposed program is compared with the newly released TRANSYT-7F version 8.1 on the basis of CORSIM simulation program. Three different demand volume levels—low, medium, and high demand—are tested. For the low-demand and high-demand volume cases, the GA-based program produced statistically better signal timing plans than did TRANSYT-7F in terms of queue time. In the case of medium-demand volume level, the signal timing plan obtained from the GA-based program produced statistically equivalent queue time compared with TRANSYT-7F. Both programs are judged to provide superior capability for oversaturated conditions due to their queue blockage model when compared with previously available signal timing optimization software.

Enhanced genetic algorithm for signal-timing optimization of oversaturated intersections

Enhancements were provided to a previously developed genetic algorithm (GA) for traffic signal optimization for oversaturated traffic conditions. A broader range of optimization strategies was provided to include modified delay minimization with a penalty function and throughput maximization. These were added to the initial delay minimization strategy and were further extended to cover all operating conditions. The enhanced program was evaluated at different intersection spacings. The optimization strategies were evaluated and compared with their counterpart from TRANSYT-7F, version 8.1. A microscopic stochastic simulation program, CORSIM, was used as the unbiased evaluator. Hypothesis testing indicated that the GA-based program with average delay minimization produced a superior signal-timing plan compared with those produced by other GA strategies and the TRANSYT-7F program in terms of queue time. It was also found from the experiments that TRANSYT-7F tended to select longer cycle lengths than the GA program to reduce random plus oversaturation delay.

Evaluation of Lane-by-Lane Vehicle Detection for Actuated Controllers Serving Multilane Approaches

A comparison was conducted of traditional vehicle detection, in which all lanes on an approach are connected to an actuated controller through one amplifier card and input, and lane-by-lane detection, in which each lane is interfaced with the controller through separate amplifier cards and inputs. Both detection strategies were evaluated with various extension time values on the same traffic stream. A total of 430 h of data from two separate approaches was evaluated on an hour-by-hour basis. During certain time periods, lane-by-lane detection provided up to a 13% increase in efficiency. For all 930 h of observations, the median increase in efficiency was 5.2% on an approach with 51-ft-long detection zones and 3.5% on an approach with 38-ft-long detection zones. This increased efficiency corresponds to time that could be allocated to other movements or used to reduce the cycle length. The largest improvements associated with lane-by-lane detection occurred during periods with moderate volume-to-capacity ratios, with smaller benefits observed during periods of heavy or light traffic.

Green Extension and Traffic Detection Schemes at Signalized Intersections

This paper provides analyses of green extensions associated with two vehicle detection schemes for actuated signal control: the current single-channel detection and the emerging lane-by-lane detection. The current single-channel detection has all detectors across all lanes on a particular approach providing a single input to a signal phase. Lane-by-lane detection, however, monitors headways and gaps on a lane-by-lane basis. A simulation model was developed to analyze both detection schemes. With the simulation model, green extensions by the two detection schemes were compared over a wide range of traffic scenarios. On the basis of study results, it was found that the two detection schemes do not produce significantly different green extensions under normal traffic flow conditions. For the various factors examined, maximum allowable headway (also passage time) is found to be more sensitive compared with other factors such as arrival headway patterns and lane volume distribution. Although the difference between the two detection schemes in average green extension is generally minimal, large differences do exist among certain cycles, and the actual impact on signal operations could be more significant; this would need further evaluation with other standard traffic simulation models.

Effectiveness of Alternative Detector Configurations for Option Zone Protection on High-Speed Approaches to Traffic Signals

Safety and efficiency are both prime issues of concern at high-speed isolated signalized intersections. However, although safety problems at such intersections can be mitigated by the application of detector configurations with option zone (also called decision zone) protection, such configurations can themselves create problems. Although option zones are often called dilemma zones, the term “dilemma” is reserved for issues associated with yellow and red clearance intervals. In this study, four detector configurations for option zone protection features—the single-detector configuration, the Southern District Institute of Traffic Engineers configuration, the Beirele configuration, and the Bonneson configuration—are compared through computer simulation for their effects relative to the number of vehicles in the option zone, max-out occurrences, and average vehicle delay. The comparison suggests that although each configuration has its own advantages and disadvantages, the Bonneson detector configuration, in most circumstances, yields the lower number of vehicles in the option zone per cycle and less average delay.

Advance Preempt with Gate-Down Confirmation: Solution for Preempt Trap

The preempt trap, a safety issue at some traffic signals near railroad grade crossings with advance preempt, can be addressed in several ways if the operating authority is aware of the potential problem. However, some solutions, such as not using advance preemption, result in other problems, including potentially unsafe clearance intervals during preemption. A simple solution, available from railroads under the standards of the American Railway Engineering and Maintenance-of-Way Association, that provides both safe and efficient operation is evaluated. The evaluation demonstrates the potential value of the use of advance preempt with gate-down confirmation.

Modeling Traffic Signal Operations with Precedence Graphs

A series of examples of traffic signal operations with the precedence graph model is presented to illustrate the interactions among phases, intervals, and overlaps. The examples are built up from the simple operation of a T-intersection by adding more complex behaviors, including advance “Walk,” delayed overlap termination indications, and advance flashing warning signals. In each of these extensions the required signal timing behavior is illustrated through the appropriate precedence graph model. The precedence graph approach provides a structured conceptual representation to support the analysis of operational behaviors of controller timing features such as added lost time and required fixed timing intervals that are induced when certain overlap features are enabled. It is believed that the precedence graph modeling approach may provide a mechanism to formalize the often ad hoc interaction between traffic signal controller software development, logic design, and field operations. This improved understanding may ultimately result in a better understood and more robust deployment of innovative signal control logic by utilizing a structure-modeling approach similar to that found in project management techniques such as the critical path method and the program evaluation and review technique.

Noncoordinated Phases in Coordinated Traffic Signal System: Evaluation of Alternative Permissive Periods on Performance

Currently, there is no standard nomenclature to describe coordination modes. Different traffic signal controller manufacturers use various terms, and the terms are not always easily understood. Furthermore, there is no documented understanding of the effectiveness of alternative approaches. The purpose of this study is to investigate the main issues related to noncoordinated movements of coordinated semiactuated traffic signals. A set of consistent terms and definitions is proposed. Based on this terminology, three coordination modes are presented, and their performance is evaluated for three different volume-to-capacity (v:c) ratios by using hardware-in-the-loop simulation. With average vehicle delay as the measure of effectiveness, results suggest that for lower v:c ratios, the modes perform differently. This paper provides some guidance on the use of coordinated semiactuated traffic signal operation by making traffic engineers aware of how different coordination modes can affect intersection performance.

Traffic Signal Phase Truncation in Event of Traffic Flow Restriction

This paper proposes an improvement to traffic signal controller logic. The concept consists of truncating a phase on which there is demand but with zero or minimal flow owing to a restriction of traffic. Flow restrictions occur for a variety of reasons including queue spillback, stalled vehicles, and railroad blockages. The logic presented in this paper recognizes such a flow restriction and, consequently, terminates the phase in favor of a conflicting call for service. The theory of the concept is addressed, and methods of implementation are explored. Although this research focuses on phase truncation, terminating a phase may not always be the most strategic decision; other priorities can override tactical detection of minimal flow. Therefore, this research considers the feasibility and potential benefit of including phase truncation logic as one component of a more intelligent and strategic controller algorithm. A limited experimental analysis was conducted to qualify the benefit of the phase truncation...