Research Of Data Process In Intelligent Traffic System Based On Cloud Computing

In the present urban traffic, user terminals such as on-board computers, smart phones,are faced with the huge conflicts of storage and processing power with the rapidly increase of data and application functions.The current transport system can’t meet the requirement of intelligent transportation in many aspects, such as information, openness, scale, intelligence, flexibility and real-time. Cloud computing, as a new concept of the network technology, has great advantage in the convenient user service, large capacity data storage, powerful computing ability, excellent service quality, high scalability, and cheap environment configuration, providing a good platform to solve the intelligent transportation and other types of large data storage and processing. With the improving of the network bandwidth, intelligent information service mode based on cloud computing becomes a hot research topic.This paper is focused on intelligent transportation services that are based on cloud computing data parallel processing technology. The data used in the intelligent transportation system is known as the traffic data of time and space, which includes the time and spatial attribute. this paper collected data of time and space in Shanghai grid project real GPS. In view of the current Intelligent transportation Service demand such as the huge amounts of data and huge computation, we proposed a kind of Intelligent transportation Service model based on Cloud Computing and Intelligent transportation Service Platform.For mass storage and processing of GPS spatio-temporal data, distributed storage and parallel preprocessing solution of GPS spatio-temporal data was proposed, which was aimed at bicycle vehicle navigation path. Mean while, to solve the linear program of vehicles guidance in intelligent transportation, the model of data parallel in large-scale sparse linear equations which based on Block Wiedemann algorithm and Block Lanczos algorithm was proposed. The main work is as followings:(1) Intelligent transportation services models based on the cloud computing. Architectures based on cloud computing is a feasible technique to solve the problem of intelligent transportation. On the basis of analyzing the advantages of the intelligent transportation cloud-computing-based service, compared to the existing urban traffic system, we proposed the framework of intelligent transportation service platform based on cloud computing(ITSPCC) and service based on ITSPCC mode. At the same time, we designed four kind of intelligent transportation service mode, which included the “service request pattern”. And through the vehicle path navigation services, we analyzes the key technology, such as the pattern of intelligent transportation services based on ITSPCC.(2) GPS spatio-temporal data preprocessing based on cloud computing. One of the core content of data processing is to find the relevant data and processing in real-time, effective and quickly from the vast amounts of traffic space-time data,according to different service requirements.For the Service oriented bicycle vehicle routing navigation, we proposed a series of technology of the spatio-temporal data for GPS structured parallel processing, distributed storage and abnormal data screening.(3) Traffic spatio-temporal data model in parallel For sparse matrix. Parallel technology is an important way to improve the efficiency of data processing. In traffic spatio-temporal data processing, due to the sparse characteristic of empty data, linear programming problem such as vehicle navigation services ultimately boils down to large sparse system of linear equations to solve problems.Take the large-scale sparse linear equations as an object,and typical Block Lanczos algorithm and Block Wiedemann algorithm as the key point, this paper studied the data parallel processing technology on the basis of ITSPCC. By analyzing the parallel characteristics of algorithm, we proposed three kinds of data parallel model, based on the row, column, line dividing the sparse coefficient matrix, and has carried on the comparative analysis with experiments.This paper studies the architecture and service model of the intelligent transportation system based on cloud computing, solving problems in a better way,such as the bicycle GPS trajectory data oriented distributed structured storage and abnormal data screening, solving problems through the data parallel of large-scale sparse linear equations of the parallel,carrying beneficial exploration on the efficient processing of.spatio-temporal data traffic. All this have better promotion and application value.