Microscopic Car Modeling for Intelligent Traffic and Scenario Generation in theUCF Driving Simulator. Year 2

摘要

A multi-year project was initiated to introduce autonomous vehicles in the UCFDriving Simulator for real-time interaction with the simulator vehicle. This report describes the progress during the second year. In the first two years, a traffic network consisting of several one and two-way roads, an outer loop and intersections with traffic control devices was created and populated with vehicles moving in a random manner. The network is completely defined in terms of a set of nodes and links. Using a Windows PC, the scientific and engineering program MATLAB was used to calculate positions and headings of the simulated vehicles with respect to the link/node coordinate system of the network. MATLAB was chosen because of its powerful data analysis, visualization and programming environment, extensive library of mathematical functions and its capacity to produce easily customizable GUI's. During the last year, some intelligence has been added to the driving patterns of the ambient traffic. This report describes the algorithms employed to accomplish this. It concludes with a brief discussion of the remaining work to achieve a fully interactive driving simulator with intelligent traffic.