Rakha becomes a fellow of the Canadian Academy of Engineering
The INTEGRATION 2.40 model is a trip-based microscopic traffic assignment, simulation, and optimization model that is capable of modeling networks of up to 10,000 links and 500,000 vehicle departures. The model is designed to trace individual vehicle movements from a vehicle's origin to its final destination at a level of resolution of one deci-second. The model simulates the departures of vehicles from a time-varying origin-destination (O-D) table. Vehicles are assigned to the network using a time-varying multi-class traffic assignment. Vehicles are then tracked at a deci-second level of resolution by modeling vehicle car-following, lane-changing, and gap acceptance behavior. Specifically, vehicle's select their speed based on link-specific steady-state car-following relationships that are calibrated using field loop detector data. A vehicle's movement from one steady-state to another is constrained its dynamics.
The model computes a number of measures of performance including vehicle delay, stops, fuel consumption, hydrocarbon, carbon monoxide, carbon dioxide, and nitrous oxides emissions, and the crash risk for 14 crash types.
The current domain of application of the commercial version of INTEGRATION includes pre-trip and en-route driver decisions, starting from the time a driver has elected to depart from a particular origin with the objective of reaching a particular destination, at a particular time, and by means of a specific vehicle class. This implies that, at present, the commercial version of INTEGRATION does not directly model the impact of drivers electing to depart at a different time, by means of a different mode, or to an alternate destination.
The model was initially developed in 1983 by the late Michel Van Aerde who continued its development until 1999. Dr. Hesham Rakha continues directing development of the model since 1999. The name of the model stems from the fact that the model was developed to integrate the modeling of freeway/arterial corridors and to integrate traffic assignment with traffic simulation.
INTEGRATION was initially developed as a mesoscopic model and then evolved into a microscopic model in 1995. Further enhancements to the model included the incorporation of a multi-modal dynamic traffic assignment, the modeling of adaptive signal control, the modeling of transit vehicles and transit priority, the inclusion of microscopic energy and emission models, the incorporation of a crash risk model, and the inclusion of tolls and High Occupancy Lanes (HOVs).
The model is commercially sold worldwide through M. Van Aerde and Assoc., Ltd. The model is also being distributed by McTrans in the US.
The INTEGRATION model is unique in a number of aspects. First, the model combines traffic assignment with microscopic simulation. Second, the model captures vehicle dynamics in terms of its acceleration behavior. Third, the model includes detailed microscopic energy, emission, and safety models. Finally, the INTEGRATION software provides flexibility in modeling numerous Intelligent Transportation System (ITS) applications.
Microscopic Traffic Assignment and Simulation
A unique feature of the INTEGRATION software is its integration of time-dependent traffic assignment capabilities with microscopic simulation. The INTEGRATION simulation model provides ten basic user equilibrium traffic assignment/routing options:
A vehicle within the simulation model computes its desired speed on the basis of the distance headway and speed differential between a vehicle and the vehicle immediately downstream of it within the same lane. The car-following model is calibrated macroscopically using loop detector data. The vehicle's ability to achieve its desired speed is constrained by its power to weight ratio, the aerodynamic resistance, the rolling resistance, and the grade resistance. A total of 25 vehicle types are incorporated in the model to capture different vehicle dynamics capabilities. The explicit modeling of car-following behavior means that the model captures the spatial and temporal formation of queues upstream of bottlenecks and the formation of shockwaves. The explicit modeling of lane changing behavior means that the model can capture the dynamic change in the capacity of a merge, diverge, or weaving section.
The model computes a vehicle's fuel consumption and HC, CO, CO2, and NOx emissions every second as it travels within the network. In addition, the model can disperse the vehicle emissions based on a user input wind speed and direction. The software can model 5 normal light duty vehicle categories, 2 normal light duty truck categories, 4 high emitter types, and 1 heavy duty truck. This unique feature of the model allows the model to accurately access the environmental impacts of operational-level transportation projects.
The INTEGRATION model also computes 14 different crash risks and the level of vehicle damage and passenger injury severity. The crash risk model was developed using the General Estimates System (GES) crash database.
Intelligent Transportation Systems
With its unique capabilities, the INTEGRATION software is capable of evaluating the efficiency, throughput, energy, environmental, and safety impacts of Intelligent Transportation Systems. The model is capable of modeling numerous ITS applications, including:
Numerous features of the model have been validated against standard traffic flow theory and field data. The features of the model that have been validated include:
The model has been and continues to be used extensively in Canada, the US, and the Netherlands. The model is also being used in Brazil, Korea, Singapore, Italy, and France. The model has been used by researchers, transportation planners, and traffic engineers from both the private and public sector. Some of the model applications include:
Download Small Version. We request that you acknowledge us in any publications or documents in which you use the the INTEGRATION software.
QueensOD (pronounced queens oh dee ) is a model for estimating origin-destination traffic demands based on observed link traffic flows, observed link turning movement counts, link travel times and, potentially, additional information on drivers' route choices. QueensOD has been created to support model the INTEGRATION traffic network simulation model. The two models, therefore, share the same data file structures and file formats. However, it is not necessary to have, or use, the INTEGRATION model in order to use the QueensOD model to estimate origin-destination traffic demands. Similarly, INTEGRATION can be used without QueensOD. QueensOD Release 2.10 is shipped with the User's Guide and the appropriate version of QueensOD software (e.g. QOD-S, QOD-M, QOD-L, QOD-X or QOD-H). In addition, the data input files to run the QNET sample network and 5 small sample networks are also provided.
Download the Small Version of the Software. We request that you acknowledge us in any publications or documents in which you use the QueensOD software by referencing the QueensOD user's guide and the two papers listed above.
The SPD_CAL software is an iterative heuristic procedure that was developed by Van Aerde and Rakha (1995) that fits the fundamental speed-flow-density relationship to loop detector data. The model uses the Van Aerde (Rakha and Crowther, 2002) functional form. The functional form requires the calibration of four parameters; free-speed, speed-at-capacity, capacity, and jam density. The solution algorithm starts with an initial set of parameters and computes the error function for this initial solution by projecting each observation on the three-dimensional functional form. The algorithm then proceeds by iteratively varying the values of the free-speed, speed-at-capacity, capacity, and jam density using a heuristic hill-climbing technique to select the optimum parameters that minimize the sum of squared orthogonal error. Unlike the traditional off-the-shelf optimization tools the software only solves for 4 decision variables with the estimated flow, speed, and density parameters computed by projecting the field observed data orthogonally on the functional relationship. Such an approach enhances the computational efficiency. The software, which runs on a standard personal computer, can also allow the user to input a feasible range for the four desired parameters, namely free-speed, speed-at-capacity, capacity, and jam density. For example, the user may request that the software only consider free-speeds in the range of 80 to 120 km/h.
Download Software . We request that you acknowledge us in any publications or documents in which you use the Spd_Cal software by referencing the papers listed above.
The VT-Micro software is a microscopic vehicle fuel consumption and emission modeling tool that allows users to estimate vehicle instantaneous fuel consumption and emission levels. The input to the model is a second-by-second vehicle speed profile.
Download Software. We request that you acknowledge us in any publications or documents in which you use the VT-Micro software by referencing the papers listed above. Note that the software has to be run in Matlab.
The VT-CPFM software is a microscopic vehicle fuel consumption and emission modeling tool that allows users to estimate vehicle instantaneous fuel consumption and emission levels. This fuel consumption model utilizes instantaneous power as an input variable and can be calibrated using publicly available fuel economy data (i.e., EPA published city and highway fuel ratings). Thus, the calibration of model parameters does not require gathering any vehicle-specific data.
Download Calibration Software. We request that you acknowledge us in any publications or documents in which you use the the VT-CPFM calibration tool by referencing the papers listed above. Note that the software has to be run in Matlab.