Research On Traffic Information Service Grid Based On WSRF

The grid technology has been developed rapidly since it was proposed in the mid-1990s, which is regarded as the third revolution of the Internet. A grid is a kind of infrastructures and has many characteristics including distributivity, dynamicality, heterogeneity, cooperativity, and high QoS. The goal of the grid is to integrate resources and to solve a problem cooperatively. Computing grids have achieved great successes in scientific and engineering computing. With the continuous development and maturing of grid technology, grids have been widely applied to many different fields. Due to the rapid development of modern highway and municipal transportation systems, research on applying grid technology to traffic management decision-making support has become a hot spot.For the problems in highway and urban traffic management decision support systems, the building of a traffic information service Grid (TISG) can provide an effective solution. The aim of TISG is to connect a large-scale geographically distributed traffic hardware and software infrastructure together by high speed Internet, manage them integratedly, implement resources sharing and cooperative working, and form a high performance traffic service platform that is transparent to traffic participants. TISG can help eliminate traffic information islands, improve the traffic environment, and increase the capacity of urban traffic, to satisfy multi-faceted and multi-level traffic requirements.Based on thorough analysis on the requirements and survey of a large number of related references, this dissertation focuses on the investigation of the TISG architecture, the system design of TISG, key technologies, and traffic flow simulations.The architecture of TISG is based on OGSA, adopts hierarchical structure, and follows WSRF——the new kernel specifications, which makes TISG positioned at the open and international standardization OGSA at the beginning. Based on the Red Had Linux operating system and Globus Toolkit 4.0 grid platform, the strategy is discussed for the implementation of TISG by using the development tools such as PostgreSQL, PBS, MPICH-G2, and Java CoG Kits, and the hierarchical structure of TISG and the system design of TISG are provided.There are some key technologies researched, such as aggregation of heterogeneous databases, information integration, combination and encapsulation of services, and the extended portal. The simulation of the traffic flow in roads is one of the important tasks of traffic management decision-making support. Under TISG environment, the traffic flow simulation method for a single lane is investigated. Because the urban road network is a large-scale and complicated system, the GeoCA theory is adopted to handle the roads and environment in the traffic flow simulation of a single lane and the concept of agent is introduced in order to simulate the genuine behavior of man and vehicle combination.Under the grid environment, a multi-agent system is built for the simulation experiments on the single lane traffic flow. By using the GT4 grid development tools, the simulation experiment system is implemented in Eclipse environment. Simulation experiment results show that the experiment of traffic flow simulation can be accomplished efficiently under TISG.In this dissertation, some problems in the building of a traffic information service grid and the traffic flow simulation are investigated. The simulation experiments on the traffic flow show that the overall design of TISG is feasible, applicable, and high efficient. The dissertation attempts to apply grid technology to traffic management decision support. Valuable results have been achieved, but more efforts are still needed to consummate and improve the current system.