Parallel Simulation Of Traffic Flow Based On PETSc Toolkit

To model vehicular traffic, a common way is to treat the traffic stream as a one dimensional compressible fluid flow. A second order nonlinear PDE for this flow that has the general formis derived based on the conservationlaw =φ(x,t). Usually this PDE is unsolvableanalytically. Parallel computing is a main trend in the development of computer science and technology. In numerically solving the second-order nonlinear PDE proposed for traffic flow, parallelization has been achieved by using PETSc, a well designed parallel toolkit for accelerating parallel programming without losing performance. Numerical experiments with various initial and boundary conditions have been performed and the corresponding results are given.