PRIMAVERA: INTEGRATED ATT STRATEGIES FOR URBAN ARTERIALS


F Biora*, KA Fox+ and FO Montgomery+

*MIZAR Automazione SpA

 Via Vincenzo Monti 48, Torino, 10126, Italy

+Institute for Transport Studies

 University of Leeds, Leeds, LS2 9JT, UK


ABSTRACT

The main aim of DRIVE II project PRIMAVERA is to evaluate the effect of the implementation of integrated ATT traffic management techniques on urban arterial roads. The main research tool used in the project is a micro simulation package with calibrated models of the field trial sites. A common framework has been developed to evaluate each strategy both individually and in combination. The best strategies are being tested in field trials in Leeds and Torino. This paper describes the process of generating and simulating the strategies and presents some of the results from the simulations.


1. INTRODUCTION

Telematics has enabled us to collect and influence both traffic flows and vehicle speeds on urban arterial roads. Traffic flows on the arterial and surrounding network can be collected automatically and passed to computers which rapidly calculate signal plans to minimise the overall travel time. Public transport vehicle location systems can also send data to these computers to let them adjust the signal plans to ensure that public transport vehicles get priority when they pass through intersections. If it is important that vehicles obey a speed limit, vehicle speeds can be detected and speeding vehicles warned to slow down via a VMS. This advice can be enforced using speed cameras.

 These methods can conflict with one another. Giving priority to public transport might cause severe disruption to other vehicles. Slowing vehicles down appears to be in direct opposition to schemes to reduce vehicle delay. However, it is possible to envisage schemes that use these methods in a co-operative way. Making vehicles move with similar speeds can produce well behaved platoons that are more easily controlled. Speeding the progress of public transport vehicles can also help following private traffic.

 PRIMAVERA's aim is to develop co-operative strategies combining these methods and to assess the benefits to efficiency, safety and the environment that can be obtained. It is not the aim of PRIMAVERA to test and evaluate individual ATT elements, but rather to evaluate the integrated effect of several ATT techniques. The ATT elements which will be combined are largely from the queue management sector and are implemented via the SCOOT and SPOT systems together with other measures such as selective bus detectors. In addition use will be made of a speed camera (Leeds site) and VMS signs (Leeds and Turin). These ATT "tools" can then be used to develop different strategies together with queue management techniques, for example different gating procedures to relieve congestion. In this paper only the simulation results using the SPOT based system on the Dewsbury Road trial site in Leeds are presented.


2. THE SIMULATION TEST BED

Micro-simulation models are becoming an increasingly important tool for traffic control research and development. They permit the traffic engineer to have a "bird's eye" view of the traffic in an urban area and an instant feel for current congestion problems and possible solutions. Controversial or new techniques can be tried and tested without any disruption to traffic in the real network.

 NEMIS was designed as a specific solution to the problem of on-street testing. Its ability to model urban networks in microscopic detail (individual vehicles, single intersections or road sections) makes it a valuable tool for testing traffic control strategies or techniques at local and area levels.

 Vehicle movement within the network is determined by:

  • the car-following rule
  • the possible manoeuvres within the link
  • the choice of turning at the next junction
  • traffic signal regulation and right-of-way rules
  • implemented traffic control strategies and/or techniques.

  •  An interface program has been produced to link the NEMIS model to either the UK SCOOT system or the Italian SPOT system, allowing them to interact. The interface package is installed on a MS-DOS PC connected by a serial line with the NEMIS computer and by another serial line with the appropriate UTC system as shown in the following diagram.

    Click here for Figure 2.1

    Figure 2.1 : The UTC System-NEMIS Interface

    An arrangement as depicted above is suitable to enable simulation of a region while also driving the real traffic signals on street, in the other regions on line. The UTC system is configured complete with data for the whole area. The NEMIS package only needs to be configured with the simulation region data. This set-up produces more realistic simulation results than many other simulation packages which have to simulate the operation of the appropriate UTC system as well as the traffic.


    3. THE EVALUATION FRAMEWORK

    The Evaluation Framework used for the simulations is described fully in [1]. Both a Cost-Benefit approach and a Multi-Criteria method have been used to assess the simulations. In the Cost Benefit Analysis (CBA) monetary values have to be given to each impact evaluated. The rates used are those given in the DRIVE I EVA manual [2], they are as follows:

    Table 3.1 : The cost of each impact used in the CBA

    Impact                     Cost

    Travel Time (UK)           14.26 ECU/person hour

    Travel Time (Italy)        18.28 ECU/person hour

    Fuel Consumption           0.36 ECU/l

    CO Emissions               3 ECU/ton

    NOx Emissions              443 ECU/ton

    Hydrocarbon Emissions      348 ECU/ton

    Fatal Casualty             744,177 ECU

    Serious Casualty           105,593 ECU

    Slight Casualty            7,080 ECU
     

    For the Multi-Criteria Analysis (MCA) two different sets of weights were used

    Table 3.2 : Weights and targets used for the MCA

    Impacts                    Units    Target     Official     Environmental
    
    sorted by                             %        Weights        Weights
    
    goals                               change
    
    Efficiency
    
    Car travel time saving     k veh s   -15        5.5455         3.0248
    
    Bus travel time saving     k veh s    -5      138.6388        106.645
    
    Travel time sd reduction   k veh s   -15        2.7728         1.5124
    
    Bus time sd reduction      k veh s   -10       69.3194        53.3225
    
    Stops                         k      -10          0             100
    
    Speed sd                     m/s      -1          0             -0.3
    
    
    
    Environment
    
    Fuel consumption saving   k litres    -5         360           3600
    
    NOx emissions                kg      -10       -0.443         -4.43
    
    HC emissions                 kg      -10       -0.348         -3.48
    
    CO emissions                 kg      -10       -0.003         -0.03
    
    Visual intrusion by queues  veh      -10          10            15
    
    
    
    Safety
    
    Mean speed                   m/s       0          0            -10
    
    Excessive speed time         k s      -5          0             50
    
    Fatal casualty reduction      -       -5       744177        1284910
    
    Serious casualty reduction    -       -5       105593         211186
    
    Slight casualty reduction     -       -5        7080          12390
    
    

    4. THE STRATEGIES

    The strategies that have been tested by simulation are described below. They are listed in the following table together with a short abbreviation that will be used in the rest of the paper to refer to the strategy.
     
     
    Table 4.1: The strategy components used at the Dewsbury Road test site
    Strategy code Strategy name
    Q Horizontal Queue Model
    A1 Auto-gating 1 - The MX strategy
    A2 Auto-gating 2 - Local Feedback Control
    B Bus priority with TIRIS
    S Speed Advice

    Horizontal Queue Model (Q)

     This strategy improves the model within the SPOT system so that queue lengths of traffic are estimated accurately. As well as improving the overall performance of the UTC system this also allows strategies such as the auto-gating strategies, to be implemented as they need an accurate estimate of the amount of space available on each link to store queues.

     Auto-gating 1 - The MX Strategy (A1)

     Auto-gating 2 - Local Feedback Control (A2)

     Gating stores vehicles upstream of a critical bottleneck, in a pre-determined section of road with plenty of storage capacity. Auto-gating or metering is a new form of gating whereby each link stores vehicles without blocking-back. This is achieved by funnelling the green times according to the downstream queues or space left. These methods involve the calculation of an upper limit to the total length of green time to be allotted to a link. This value is then used to overwrite temporarily the maximum green time for the stage which is green to the link; in cases where the link receives green time in more than one (consecutive) stage, the upper limit of green time is distributed between the relevant stages.

     Bus Priority using TIRIS (B)

     If information on the approach of a bus at a set of signals is available then this information can be used by the SPOT optimiser in order to benefit buses. This benefit may take three different forms:

     (i) prevent an early termination of the stage which benefits the bus;

     (ii) extend the stage which benefits the bus;

     (iii) recall early the stage which benefits the bus.

     Four junctions in the inbound direction have been equipped with TIRIS readers. Priority is given to bus services 2, 24 and 46.

     Speed Advice (S)

    VMSThe installation of a speed camera on the southern section of Dewsbury Road, provides a mechanism for enforcing slower progression speeds. This will be combined with a VMS sign warning drivers if they are travelling too fast to slow down. In the NEMIS model it has been assumed that this mechanism will be successful in limiting the maximum speeds of the vehicles.


    5. RESULTS

    For the SPOT runs the base environment uses the standard SPOT system. This allows the new strategies to be tested against the current state-of-the-art system. The key to the strategy abbreviations used in the following tables and figures can be found in Table 4.1. Table 5.1 contains the % change in the impact over the base case. The travel times of the bus services given priority treatment (Bus TT) are also shown.
     
    Table 5.1 : % changes in Network Impacts over standard SPOT - AM Peak
    Impact Q+S Q+B B+S Q+B+S
    Mean Speed (m/s) 2.27 1.08 2.49 3.06
    Speeding Time (s) -8.24 0.91 -7.94 -8.18
    Blocking Back (s) 5.39 17.22 32.37 6.74
    Stops (veh) -1.67 -0.73 -2.46 -1.64
    Delays (s) -1.13 -1.28 -2.45 -4.33
    Travel Time (s) -2.26 -1.22 -2.44 -2.95
    Fuel Consumption (l) -0.79 -0.38 -1.06 -1.02
    CO Emissions (g) -1.81 -1.17 -2.40 -2.66
    NOx Emissions (g) -1.04 -0.78 -1.54 -1.76
    HC Emissions (g) -1.68 -1.08 -2.01 -2.38
    Bus Travel Time (s) -2.10 -1.21 -1.50 -1.66
    Bus Time Route 2 (s) -2.23 -4.32 -2.96 -2.75
    Bus Time Route 24 (s) 0.38 1.46 2.51 -1.85
    Bus Time Route 46 (s) -3.07 -2.83 -1.49 -2.07

    Table 5.2 indicates the results of the CBA. All values refer to the two hour AM peak data collection period. All the monetary values are in ECU. The rates used for the CBA can be found in Table 3.1. For the SPOT based strategies in the AM Peak, the integrated strategy containing the Horizontal Queue Model, bus detection with TIRIS and Speed Advice using VMS is easily the most beneficial integrated strategy. Giving priority to buses always improves a strategy when combined with other components.

    Table 5.2 : SPOT based strategies for the AM Peak - Leeds

    Strategy       Cost (ECU)   Benefit over  % Benefit over
    
    (see Table                    standard        standard      
    
    4.1 for codes)                  SPOT            SPOT  
    
    
    B               26606.83       901.13          3.28      
    
    Q + B + S       26991.82       516.14          1.88
    
    Q + S           27042.12       465.84          1.69      
    
    B + S           27059.62       448.33          1.63      
    
    Q + B           27232.93       275.02          1.00      
    
    Q               27395.91       112.04          0.41      
    
    S               27829.06      -321.11         -1.17      
    
    A1              27878.04      -370.09         -1.35      
    
    A2              28188.33      -680.38         -2.47      
    
    
    The travel time figures dominate the costs associated with each strategy. Typically they are about ten times larger than the accident costs, twenty times larger than the fuel costs and a thousand times larger than the pollution emission costs. The CBA will therefore favour strategies which reduce travel time, with little regard for environmental or safety factors. Safety is a problem on the Dewsbury Road, having an accident rate of nearly double the national average for roads of its type. The CBA does not adequately reflect any improvements in safety. For this reason, the MCA is preferred.

     A multi-criteria analysis has been carried out using the MASCOT program [3]. This program comes with three sets of weights built-in that reflect the opinions of the values of "Official", "Environmental" and "Commercial" groups. This set of weights has been supplemented by some additional values for impacts not originally considered. As the results of the "Commercial" weights are very similar to those using the "Official" weights, only the "Official" weights are considered here. The weights and targets used can be found in Table 3.2.

     For each integrated strategy three scores are obtained that relate to each of the three DRIVE goals of Safety, Efficiency and the Environment. These three scores can define a point on a three dimensional graph whose axes are the three DRIVE goals. By plotting each strategy in this way it becomes relatively easy to see the effects of integrating strategy components and to isolate families of strategies that have similar consequences. To aid strategy recognition, all the individual strategies are represented by a hollow symbol and all the integrated strategies by a solid or filled symbol.

    Click here for Figure 5.1

    Figure 5.1 : 3-D plot of the MCA of the SPOT based strategies

    For the SPOT based strategies the Q+B+S strategy comes out on top when using the "Environmental" weights and second with the "Official" weights. There is a cluster of strategies which are all good on safety, efficiency and the environment, these being Q+S, B+S and Q+B+S. For all the strategies, integration with Speed Advice improves safety without adversely affecting efficiency or the environment. Integration with bus priority does not have much effect. A sensitivity analysis has been carried out which reveals that it is very difficult to remove Q+B+S from its top position by changing the weights. The most likely change is to reduce the excessive speed weight by a factor of three, which puts the B strategy on top. Changes in scores can also change the rankings, however within the uncertainties associated with each score it is only possible to get either Q+S or B+S to replace Q+B+S as the top strategy.


    6. CONCLUSIONS

    PRIMAVERA is looking for ways of integrating strategies to enhance their overall effect. The project objectives of efficiency, safety and environmental improvement clearly have conflicts and so do the policies of queue management, public transport priority and traffic calming.

     The simulation runs have shown that traffic can be calmed to a slow but constant speed, resulting in predictable platoons of traffic which are more easily managed by the ATT queue management strategies. This results in smoother flows of vehicles which produce less pollution and are less likely to produce conflicts with vulnerable road users. Integrated strategies that also give priority to public transport have been successfully designed. Synergy has been achieved.


    7. REFERENCES

    [1] Watson, S. et al., "Evaluation Methodology", Deliverable No. 11, DRIVE II Project V2016 - PRIMAVERA, 1993.

    [2] "EVA Manual", Technische Universität München, Munich, 1991.

    [3] Bonsall, P.W. et al., "MASCOT - A Decision Support System to Help in the Definition, Development and Ranking of Scheme Options" , Proceedings of 19th PTRC Summer Conference, Seminar G, London, 1991, pp.123-135.



    ITS Home PageITS home page