Derivation of Van Aerde traffic stream model and studying the effect of Ramp Metering

With increase in population, demand for traffic is increasing tremendously. As a consequence congestion on urban freeways or highways during the morning and evening peak hours is always increasing. This research examines the traffic condition on Highway 2 (i.e Deerfoot Trail), Calgary, Canada by providing ramp metering (funded by Government of Canada in Collaboration with MITACS Globalink Program). In order to accommodate increasing traffic demand, constructing new highways and adding lanes is not always the best option. Instead proper management of transportation system can help in increasing its efficiency. It is therefore important to increase our understanding of traffic flow models to help estimating the efficiency of different road facilities and therefore prepare us to make enlightened decisions in the design process of new road facilities or improving older ones. This research examines the traffic condition on Highway 2 (i.e Deerfoot Trail), Calgary, Canada by providing ramp metering. The research Project consists of 2 phases: 1. Calibration of Deerfoot Trail with Van Aerde Model using Microsimulation tool Quadstone Paramics, and 2. Analyzing the efficiency of Ramp Metering (RM) by examining its impact on the Macroscopic Fundamental Diagrams (MFD) for Deerfoot Trail. (This research is funded by Government of Canada in Collaboration with MITACS Globalink Program). ANALYSIS 22(71), August 1, 2014 Discovery ISSN 2278–5469 EISSN 2278–5450


INTRODUCTION
With more vehicles on the roads the interest in enhancing

OBJECTIVE OF THE RESEARCH
The objective of this project is:

APPLICATION OF THE RESEARCH
The prime application of this research is in the off-line calibration of speed-flow relationships.In the past, poor fits of speed-flow relationships have often resulted in a lack of an ability to properly calibrate the entire simulation.Other main application of this research is to analyze the efficiency of Ramp Metering (RM) in case of Deerfoot Trail.

A. Van Aerde Car-Following Model
The entire document should be in Times New Roman or Times font.Type 3 fonts must not be used.Other font types may be used if needed for special purposes.

RAMP METERING
A ramp meter, ramp signal or metering light is a device, usually a basic traffic light or a two-section signal (red and green only,

B.2. Probe Vehicle Detection Technology
The Section 4

Diagrams comparing with and without RM case
Section 5 The results of the isolated ramp metering analysis showed an ANALYSIS overall improvement in traffic condition (i.e.average travel time, flow and speed) for both the freeway and network-wide.The results of the analysis indicated a superior performance of the algorithm under high and medium traffic conditions on both the ramp and on the freeway and for high demand of the freeway and moderate demand on the ramp.As expected, no improvement was shown for the case of high demand on the ramp and moderate demand on the freeway.

MACRSCOPIC FUNDAMENTAL DIAGRAMS
The accessibility of large cities is under high pressure.As an example, average speeds on major arterials in the city of Athens, Greece, appeared to be three to five times smaller during peak hours than during off peak hours, while in many other European cities speeds in the city centre are less than 10 km/h during peak hours.Nowadays, developing countries start to face similar problems.
Constructing new infrastructure or extending the existing infrastructure is less likely to occur in the coming years due to the high costs and the societal impacts.In the short term, the only way to improve accessibility is to make better use of shows that the shape of the MFD is not independent of the demand.

CONCLUSION
The summery of conclusions is as follow: 1. Data obtained from Quadstone Paramics was analyzed and modelled according to the Van Aerde microscopic model.
Data for 30 second intervals were analyzed and the average speed, flow and density were estimated.The Van Aerde model knowledge of microscopic simulation of traffic streams has become more important.Steady-state traffic stream modelling reveals the speed/flow/density relationship of a road facility and can provide us with answers to questions such as what parameters dictate maximum roadway throughput (capacity) as well as the speed-at-capacity.Traffic detectors collect accurate data regarding fundamental traffic stream parameters which can be used to present real data from different road facilities.By comparing different traffic stream models with real field data it is possible to obtain a measure of the model capability of revealing data from different facilities.
1. To compare real traffic data obtained from Quadstone Paramics Software with a four parameter steady-state traffic stream model, known as the Van Aerde carfollowing model (Van Aerde Calibration -Quadstone Paramics software) 2. To check the efficiency of Ramp Metering (RM) by examining its impact of RM on the Macroscopic Diagrams (MFD) for Deerfoot Trail.3.To develop MFD diagrams with and without RM to examine the impact of freeway control on the shape of MFD.

A
number of different traffic car-following models exist and most of them are based on two independent variables.Usually these variables include the distance headway and the speed differential between the leading and the followin vehicle (Rakha & Crowther, 2002)[1].One of the most popular of these models is Greenshield's model which requires the free speed vf and jam density kj as the only parameters.Greenshield's relationship has been found to be too simple and real data often deviate from the theory for both the uncongested as well as the congested regime [1][2].One recent model is the steady-state four parameter Van Aerde car-following model [4].
To use the Van Aerde model one has to start by estimating four parameters c1, c2, c3 and k.These parameters are related to the free-flow speed Vf; the speed-at-capacity vo; capacity qm and jam density Kj.The equations for the parameters are given as: c1 = m*c2 Where: c1 = fixed distance headway constant[Km] c2 = first variable distance headway constant[km^2/h] c3 = second variable distance headway constant[h] m = a constant used to solve for three headway constants[h/km]The constant m is calculated by the equation:According to Van Aerde Headways can be described by:The Van Aerde model can be seen in Figure:

ANALYSISFrom
Fig. : The Van Aerde Traffic stream Model no yellow) light together with a signal controller that regulates the flow of traffic entering freeways according to current traffic conditions.It is the use of traffic signals at freeway on-ramps to manage the rate of automobiles entering the freeway.Ramp metering systems have proved to be successful in decreasing traffic congestion and improving driver safety [3].Types of Ramp Metering strategies Ramp metering may operate either in pre-timed or adaptive: 1) Pre-timed Ramp Metering: have fixed cycle time and phase length.The ramp metering rate is mainly determined as function of historical traffic information.2) Adaptive Ramp Metering: Adaptive ramp metering control varies their rates minute-by-minute or second by second based on real-time traffic parameters.Few examples of Adaptive Ramp metering models are : a) Dynamic Ramp Metering Model b) Advanced Real Time Metering System (ARMS) Detection Technologies for traffic condition From the above literature, it is clear that most ramp metering algorithms optimize the traffic based on occupancy, approximate density or flow information measured in the vicinity of the ramp.The various traffic detection methods are reviewed and compared to the vehicle probe detection method.B.1.Point Detection Technology Point detection technology is currently the most widely used sensor in traffic detection systems.These sensors are mostly capable of measuring flow and occupancy, and are able to estimate speeds.
probe vehicle technology is used in Intelligent Transportation System (ITS) applications for real time traffic data collection.-Automatic Vehicle Location(AVL) -Automatic Vehicle Identification(AVI) -Global Positioning System(GPS) probes -Vehicular Probes that collect travel time data by discreetly tracking cell phones within cellular network.The detection technology used for in case of Deerfoot Trail was Vehicular probes.A C++ .DLL code was developed which can be used as a plug-in for Quadstone Paramics, as the Paramics software doesn't support dynamic ramp metering strategies[9].
existing road capacity and to improve the production efficiency of large transportation networks.The latter can be done by applying Dynamic Traffic Management (DTM) measures.The precondition for efficient implementation of networkwide DTM is to describe the state of the network, that is, a macroscopic description of the traffic flow.This can be used to predict the states of large scale networks more efficiently, while maintaining its accuracy and to assess the network wide performance quality.Daganzo and Geroliminis [8] developed the macroscopic fundamental diagram (MFD) which describes the network-average relation between the number of vehicles in the network and the performance / outflow [8].While monitoring the state of traffic continuously when the MFD is known, transportation managers can see if their traffic system is in the preferred state, that is, nowhere in the network the desired accessibility levels are exceeded.Using MFD, existing traffic management strategies can be refined and applied with more accuracy.However, because of the complex layout of the network and behavioral differences of different traffic types, the shape of the MFD changes and scatters may occur, which may lead to less accurate predictions of the traffic state.If the traffic state in a network cannot be predicted accurately, the quality of the traffic control based on these predictions may decrease as well.Therefore as a part of this research internship, MFDs for Deerfoot trail with and without RM case was also developed by using the data obtained from Paramics for different congestion levels.At the end of the MFD, we can see that lower congestion levels lead to lower accumulation.Levels with similar congestion can be identified, where more congestion corresponds to higher accumulation and higher flow.FIGURE6b, the MFD without ramp metering, shows that more congestion leads also to higher accumulation, but the corresponding flow is lower.This indicates a less efficient traffic situation.The shape of the MFD is therefore not only a property of the network and the demand, but also of the applied traffic control measures.So it can be seen, Ramp-metering has a direct impact on the shape of MFD.The reduction of the inflow leads to higher flows and lower accumulation.Also, different congestion levels appear to lead to only a slight increase in both the accumulation and the flow, while in the situation without ramp metering the accumulation significantly increases while the flow decreases (heavy congestion).It can also be seen that, by a fast decrease in demand leads to an inefficient use of the links: the links are able to process more traffic than the current demand, while the outflow is still low due to the congestion on the link.This clearly