Development of Integrated Meso/Microscale Traffic Simulation Software for Testing Fault Detection and Handling Algorithms in AHS: Final Report

Executive Summary A large portion of PATH's e.ort in the area of Advanced Vehicle Control Systems (AVCS) during the past several years has focused on the development of theory and technologies that may signi.cantly increase highway safety and capacity through the use of Automated Highway Systems (AHS). To continue with the analysis of AHS feasibility, it is necessary to perform exhaustive large scale simulations that will provide information on the impact of the proposed technologies upon system safety and capacity. In particular, it is necessary to study and evaluate how faults in a vehicle impact the overall AHS performance and capacity, how the roadside control systems can react to these faults and perform degraded-mode activities at the higher hierarchical levels, and how the roadside control system can detect faults either in the vehicle or in its own infrastructure. When the response of an AHS is being simulated, di.erent degrees of precision are required. In some sections of the highway the detailed behavior in each vehicle in the section must be simulated to model vehicle or infrastructure faults, or test vehicle level fault detection and handling algorithms or degraded mode maneuvers. This level of AHS simulation, where the dynamics of each vehicle is simulated, is called microsimulation. The AHS microsimulation software currently under development at PATH is SmartAHS. This software package is based on the hybrid systems language SHIFT [7]. In most sections, it is convenient to model the highway using .uid-like conservation models of tra.c .ow, applied to the average tra.c characteristics in that section, such as density, .ow, and velocity. This level of AHS simulation is called mesosimulation. Mesoscale simulators do not provide information about particular vehicles but are instead capable of simulating large highway networks. The mesosimulation software used by PATH for analyzing Automated Highway Systems is SmartCAP [5]. Although it is in principle possible to microsimulate a large scale AHS, the computational cost is prohibitively high and, in many cases yields no advantage over mesosimulation, except for the few sections of the highway where it is needed. The aim of this project is to bridge the simulation gap between SmartAHS and SmartCAP, by implementing an integrated AHS micro-meso simulation environment for simulating a large scale AHS network, where both SmartCAP and SmartAHS run simultaneously and interact with each other. In this simulation environment, most of the highway sections in the AHS are simulated by the SmartCAP mesosimulation software, except for one or more sections, which are simulated by the SmartAHS microsimulation software. Highway sections that are microsimulated using SmartAHS are referred to as u-windows. The topology of the entire AHS, as well as the location and length of each u-window is set up by the user, by means of user interface software, named mminterface. The bene.ts of the meso-micro simulation environment are best realized when u-windows are placed in the regions where microscopic e.ects are of greatest interest; by allowing the remainder of the AHS to be simulated by the more computationally e.cient SmartCAP, the meso-micro software is able to simulate the AHS at a lower computational cost than stand-alone SmartAHS. This .nal report documents the theoretical design and software implementation of the integrated micro-meso simulator. It also provides a brief tutorial on its use, including a simple AHS simulation example. The following tasks were executed in this project: 1. The capabilities of SmartAHS were extended by developing a simpli.ed sensor architecture, a simpli.ed vehicle-roadway environment processor, and a simpli.ed set of regulation-layer components. Under certain conditions, the simpli.ed set produces identical results as the full components, and increases simulation speed 4 to 5 fold. 2 2. The SmartCAP activity model was extended to include platooning and join/split maneuvers. 3. The interface between SmartCAP and SmartAHS was designed and implemented in SHIFT. A SmartAHS component was created to schedule and monitor all aspects of the interface between SmartCAP and SmartAHS. 4. A batch compiler and a MATLAB-based visual interface were created to allow the user to input simulation parameters, de.ne the highway topology, and view both mesoscale and microscale simulation outputs simultaneously. 5. The developed integrated meso/microscale simulation software was tested for di.erent scenarios. The PATH hierarchical control architecture, speci.cally the link, coordination, regulation, and physical layers, was tested in the meso-microscale simulator. The meso-micro simulator is well-suited for testing the response of an AHS to situations that are characterized by localized microscale phenomena, such as: the stalling of an individual vehicle, oversaturation of a particular artery due to external factors (e.g. a sporting event), emergency vehicle maneuvers, and changes in highway capacity due to di.erent on-ramp metering policies. The use of the software is described in Sec. 6 using a simple simulation example. The example shows a tra.c density wave propagating forward from the mesoscale region into the microscale region, thus demonstrating the functionality of the software.