Effects of Traffic-Actuated Signal Control Strategies on Fuel by HC121003135446


									Effects of Traffic-Actuated Signal Control Strategies on Fuel Consumption

                              Thesis Proposal by
                              Stefan J. Widomski

                     Thesis Advisor Professor Feng-Bor Lin

          Submitted to Clarkson University Honors Research Committee

                      Date of Submission: March 14, 2008


       There are numerous signalized intersections in the United States. Most of these

intersections are controlled by traffic-actuated signals. Traffic-actuated signals are capable of

altering the green duration for a given signal phase through embedded control logic. Traffic-

actuated control logic relies on information collected from vehicle sensors. It also relies on

timing design settings to determine how right-of-way should be allocated in a given cycle.

Given their complexity, most traffic-actuated signals do not operate at maximum efficiency.

This leads to an unnecessary waste of fuel, as vehicles will likely be idling longer while waiting

for a green light. Improving the efficiency of traffic-actuated signals may save a considerable

amount of fuel.

       There are various traffic-actuated signal control strategies. The most common strategy

employs inductive loop detectors embedded in the pavement, a set of timing settings, a phasing

plan, and a control logic. These four components collectively define the control strategy. The

two main types of detectors currently used in traffic-actuated signal controls are motion detectors

and presence detectors. Motion detectors are relatively short in length; they send a signal to the

controller when a vehicle moves across a detection area. Presence detectors are longer in length;

they send signals to the controller when a vehicle is inside the detection area. Timing settings

are comprised of the minimum green, vehicle interval, and maximum green. The phasing plan is

the grouping of movements through and intersection at different time durations to avoid

conflicts. Control logic processes all of the input signals and dictates whether the controller

should stop or extend the green duration.

       The main objectives of this proposed research are as follows:

           1. To identify the effects of phasing plan, detector loop length, and respective timing

               settings on fuel consumption.

           2. To identify combinations of detector loop length and timing settings that can yield

               high levels of fuel economy.

       Past traffic signal control research has mainly been concerned about operating signals to

reduce vehicle delays or to maximize intersection capacity. There is very little understanding

about the link between traffic-actuated signal control strategies and fuel consumption. The

relationships between signal control strategies and fuel consumption must be understood in order

to facilitate control improvement to reduce fuel consumption.

                                        Literature Review

       Past research on signal control strategies has been focused on minimizing vehicle delays.

For example, Webster [1] has developed the following model for determining the optimal pre-

timed signal control cycle length:

                                        1.5L  5
                                 Co                            (1)
                                          1 Z

where Co = optimal cycle length for delay minimization (s);
      L = lost time per cycle (s); and
      Z = the sum of critical volume to saturation flow ratios of individual signal phases.

Cheng, Tian, and Messer [2] have also developed an alternative model for determining the

optimal pre-timed signal control cycle length:

                                 C o  1.5Le 1.8 Z              (2)

where all variables are previously defined.

       Practically all research relating to traffic-actuated signal controls is focused on operating

the controls so that vehicle delays are minimized. Lin and Percy [3] analyzed the interactions

between loop detectors and vehicles from the perspective of vehicle delay. In addition, Lin [4]

also investigated optimal timing settings and detector lengths of traffic-actuated signal control.

These studies focused on vehicle delay, not fuel consumption, as the measure of effectiveness

(MOE) of an intersection.

       More recent research has led to the development of analytical fuel consumption models.

The Canadian Capacity Guide for Signalized Intersections [5] employs a fuel consumption

model based on vehicle delay:

                                                                                        
                  N sij u stop , v    d sij q sij u idle    q ij u cruise 
              U                                                                        
                   j i                     j i                        j i

where U = estimate of fuel consumption in the zone of influence of the signalized intersection
      Nsij = number of stops in lane i during phase j;
      ustop,v = additional unit of passenger car fuel consumption caused by stopping and
      resuming a given cruise speed (g/stop);
      qij = arrival flow in lane i during phase j (pcu/h);
      dsij = average stopped delay in lane i during phase j (s/pcu);
      uidle = unit of passenger car fuel consumption per second of idling (g/s); and
      ucruise = passenger car fuel consumption over a distance of 100 m at a given cruise speed
      on level ground (g/100 m).

From the above variables that define the model, it is easy to see that the model takes cruising,

idling, acceleration, and deceleration into account for a passenger car moving through the

intersection. This model is not directly linked to signal control strategies. Therefore, additional

models are needed to estimate the variables that are needed as inputs to the model. Liao and

Machemehl [6] estimated fuel consumption using 22 variables. Some of these variables include

the number of stopped vehicles, the number of moving vehicles, effective red time, effective

green time, and vehicle speed. Like the Canadian model, though, fuel consumption is estimated

by taking the sum of nine different models. Similarly, the model developed by Liao and

Machemehl is only applicable to pre-timed control strategies.

       The operation of traffic-actuated signals is too complex to be modeled analytically.

Computer simulation is the only practical tool for examining the potential relationships between

actuated signal control and fuel consumption. Some of the simulation models available are

NETSIM, TEXAS, and SIDRA. In the United States, NETSIM (Network Simulation) model [7]

is the most widely used computer simulation software for simulating traffic on surface (urban)

streets. Developed by the Federal Highway Administration, this model is very flexible, because

it allows simulated scenarios to be changed over a wide range of conditions. NETSIM is capable

of outputting many MOEs, such as vehicle delay, fuel consumption, and queue length. TEXAS

(Traffic Experimental and Analytical Simulation model) [6] is computer simulation software

developed at the University of Texas at Austin that provides estimates of fuel consumption and

emissions. These estimates are obtained by consulting large EPA fuel tables. However, this

software is not as widely used as NETSIM in the United States. SIDRA [8] is a simulation

software package developed in Australia capable of simulating different types of intersections.

SIDRA uses a four-mode elemental model for estimating fuel consumption. Like TEXAS, this

software is not as widely used as NETSIM.

                                      Research Approach

       Because computer simulation software is the only practical way of identifying and

analyzing the complex relationships between fuel consumption and signal control strategies, this

proposed research will use NETSIM as the tool of analysis. The research will encompass the

tasks described below.

   Task 1      Conduct an In-Depth Literature Review

            Two reasons warrant the completion of a literature review. The first reason is to

            document relevant research conducted in the past. The second reason is to help

            define the scope and the direction of the proposed research.

Task 2      Design Experiments (Scenarios to be Analyzed)
            Several factors may affect fuel consumption. These include intersection

            geometric design, traffic demand pattern, and signal control strategy.

         a. Intersection Geometric Design
                    Most intersections are either four-way or T-intersections. These two types

                    will be analyzed in the research. Each intersection approach will have two

                    lanes. Exclusive left-turn lanes will be considered when an analysis

                    involves protected left-turn phases.

         b. Traffic Demand Pattern
                    Traffic demand involves vehicle arrival patterns and vehicle flow rates

                    through an intersection. For this research, only random arrivals will be

                    considered. This is because most traffic-actuated signals are not

                    coordinated. Each intersection approach will be simulated for low,

                    moderate, and heavy flow rates.

         c. Signal Control Strategy
                   The grouping of traffic movements from different directions for right-of-

                    way allocations is called the phasing plan. The timing design elements of

                    each phase consist of minimum green time, vehicle interval, and

                    maximum green time. The minimum green time is the minimum time

                    allotted for a phase to display a green light in a given cycle. The vehicle

                    interval is the amount of time the green light is to be extended when a

                    vehicle actuates the detector. It is also used to determine when a green

                interval for another phase should be terminated. Each vehicle that actuates

                the detector during the current vehicle interval extends the green time until

                another phase calls for service and/or maximum green time is reached.

                The maximum green time is the set time limit for the green light to be lit

                for a given phase. However, this time limit can be exceeded for fully-

                actuated controls if no other phase immediately calls for service. Phasing

                plans to be considered will include both two- and four-phase controls with

                protected and permitted left turn phases. The National Electrical

                Manufacturers Association (NEMA) [9] has developed and published

                standards for signal control logic. NEMA control logic will be used as the

                basis of analysis.

         Test simulation runs will be conducted prior to the production runs so that the

         number of runs needed to analyze each scenario with a high level of reliability can

         be estimated. The length of each run should simulate signal operation for at least

         15 minutes. The initial phasing plan will consist of a minimum green time, a

         vehicle interval being close to zero, and a maximum green time long enough to

         ensure no queue spillback. The flow rates for each scenario will be low, so that

         an optimal timing strategy may be determined when flow is increased.

Task 3   Perform Simulation
         Scenarios will be developed and simulated using NETSIM. Multiple runs will be

         completed for each scenario. In each group of runs, one element will be adjusted

         at a time until the output shows a timing design at which the fuel consumption is

near its optimal level. The simulation process is clearly outlined in the flow chart


                        Geometric Design
                      Traffic Demand Pattern

                           Phasing Plan, i

                         Detector Length, j

                      Minimum Green Time, k

                         Vehicle Interval, m

                      Maximum Green Time, n


                    Optimal Values Achieved?
                 No                         Yes

                           Select Variable
                            to increase:
                             m=m+1                  Terminate Simulation;
                              k=k+1                   Analyze Another
                              j=j+1                       Scenario

                Figure 1: Simulation Flow Chart

                     Data will be collected from the output files at the end of each group of runs.

                     These data will include vehicle delay, queue length, fuel consumption, and

                     percent of vehicles stopped.

      Task 4         Analyze Simulation Data
                     One purpose of analyzing the output data is to identify the relationship between

                     signal control strategies and fuel consumption. Another purpose is to identify

                     efficient ways of operating actuated signals in order to achieve optimal or near-

                     optimal fuel economy. Graphical representations of fuel consumption versus

                     different variables will be plotted and analyzed. For example, plotting fuel

                     consumption versus detector length will enable an optimal detector length at the

                     lowest fuel consumption level to be determined. The simulated data may also be

                     analyzed through regression analysis. If necessary, neural network modeling may

                     also be employed.

                                              Approximate Timeline

                                      2008                                                         2009









                Task 1

                         Task 2

                                            Task 3

                                                                  Task 4

                                                                           Organize, Write, & Present Thesis


1. Webster, F. V. Traffic Signal Settings. Road Research Technical Paper No. 39, Her

   Majesty’s Stationery Office, London, 1958

2. Cheng, D., Z. Z. Tian, and C. J. Messer. Development of an Improved Cycle Length

   Model over the Highway Capacity Manual 2000 Quick Estimation Method. Journal of

   Transportation, Vol. 131, No. 12, American Society of Civil Engineers, 2005, pp. 890-


3. Lin, F.B. and Percy, M.C. Vehicle-Detector Interactions and Analysis of Traffic-Actuated

   Signal Controls. In Transportation Research Record: Journal of the Transportation

   Research Board, No. 971, TRB, National Research Council, Washington, D.C., 1984, pp.


4. Lin, F. B. Estimating Average Cycle Lengths and Green Intervals of Semiactuated Signal

   Operations for Level-of-Service Analysis. In Transportation Research Record: Journal

   of the Transportation Research Board, No. 1287, TRB, National Research Council,

   Washington, D.C., 1990, pp. 119-128

5. Allingham, D. I., Richardson, D. B., Stephenson, B. W., and Teply, S. Canadian

   Capacity Guide for Signalized Intersections, Second Edition. Institute of Transportation

   Engineers. 1995, pp. 70-71

6. Liao, T. Y., and Machemehl, R. B. Development of an Aggregate Fuel Consumption

   Model for Signalized Intersections. In Transportation Research Record: Journal of the

   Transportation Research Board, No. 1641, TRB, National Research Council,

   Washington, D.C., 1998, pp. 9-18

7. Federal Highway Administration. Appendix A: Introduction to CORSIM Theory

   http://ops.fhwa.dot.gov/trafficanalysistools/tat_vol4/app_a.htm. Accessed April 6, 2008.

8. SIDRA INTERSECTION and SIDRA TRIP software for Operating Cost, Fuel

   Consumption and Emissions.

   www.sidrasolutions.com/traffic_resources_cost_fuel_emissions.htm. Accessed March 10,


9. The National Electrical Manufacturers Association. Traffic Controller Assemblies with

   NTCIP Requirements, Version 02.06. http://www.nema.org/stds/ts2.cfm. Accessed April

   6, 2008.

To top