Study On Several Key Technical Issues Of Seamless Transportation Information Grid

The fundamental objective of intelligent comprehensive transportation system is to take full advantage of information-based modern technologies to enable all kinds of transportation modes to effectively link up, and carry out integrated operations management, decision-making, planning, information and resource sharing. The nature of grid is to reach full resource sharing, therefore the development of grid technology provides an effective way for the construction of intelligent comprehensive transportation system. Seamless Transportation Information Grid(STIG) presented in this dissertation is an open system architecture, which can provide interconnection of information facilities and resources in highway, railway, waterway, aviation, etc transportation systems, and dynamically form virtual organizations of transportation for completion of each complex task, carry out real-time coordination and organization. Moreover, STIG can achieve rapid information collection, rapid information processing, rapid information sharing, and ultimately lead to implement comprehensive, anytime, anywhere information sharing and cooperative work of various transportation modes. For issues of architecture, data integration, semantic query, data mining and information coordination of multimodal transportation in Seamless Transportation Information Grid, the main research contents of this dissertation are as follows: (1) Through binding grid technology and information sharing, intelligent processing in seamless transportation, this dissertation introduces the concept of Seamless Transportation Information Grid. According to the features of seamless transportation as well as Service-centric of SOA and advantages of OGSA hierarchical model, this dissertation presents architecture of Seamless Transportation Information Grid. The architecture of STIG is composed of resource layer, basic SOA layer, grid services deployment layer, grid middleware layer, agent layer and applications layer, which provides plug and play services for the implementation and management of seamless transportation.(2) In view of the weakness of OGSA-DAI on data integration, this dissertation designs a data integration platform which is composed of uniform access interface for data integration, data processor, data accessor, basic services and components of OGSA-DAI, data sources in order to achieve all traffic and transportation information sharing in comprehensive transportation system. With the platform, users of Seamless Transportation Information Grid can uniformly access data from databases of different vendors. This integration platform introduces virtual database, database schema integration strategy and optimization model of distributed query for supporting data transparent access, heterogeneous data integration and improving the efficiency of distributed query.(3) For problem of traditional information query which was caused by lacking semantic support, this dissertation implements an ontology-based semantic query method in Seamless Transportation Information Grid. In this method, each data source corresponds to a local ontology, multiple partial ontology form global ontology; For user query, the method first builds user query vector, then adopts UQuery2GQuery algorithm to generate global query vector; For global query vector, the method first constructs query tree for decomposition, then uses GQuery2SQuery algorithm to map each query tree to sub-query on corresponding data source. Moreover, the dissertation presents a GSQO(Grid Semantic Query Optimizer) for improving performance of semantic query in STIG. GSQO mainly implements optimization of the following four modules:semantic extension of user query, SPARQL query, resource selection, parallel processing.(4) In order to implement parallel data mining in Seamless Transportation Information Grid environment, this dissertation designs a data mining architecture based on real-time virtual data warehouse. For the mapping between data mining nodes and data sources, the dissertation presents MapND mapping algorithm. Through using minimum response time as the key evaluation factor, MapND algorithm respectively estimated the sum of transmission time and data mining time between each mining node and each data block, then mapped data block to corresponding mining node according to the minimum sum of time.(5) The dissertation presents an information coordination model of multimodal transportation based on grid workflow, the model builds EDAG-based collaboration diagram for information grid services of multimodal transportation. The complicated synchronization of nodes in collaboration diagram can be described using converge structure and branch structure. In addition, an information collaborative optimization method is given in this dissertation, the method processed workflow applications of multimodal transportation according to following five phases:weighting phase, ranking phase, mapping phase, grouping phase and rescheduling phase, and reasonably arranged services scheduling of workflow tasks, at last obtained optimization of information coordination for multimodal transportation workflow application.