近年来，我国经济高速发展，车辆保有率的随之提升，一方面代表着我国居民生活水平提高，另一方面也为我国城市道路交通通行带来压力，各地的交管部门纷纷建设智能交通系统以加强区域交通管控。随着智能交通系统数量的增加，也诞生了海量的交通数据。由于其行业特殊性，交通大数据存在多源、异构、局部性、时空关联和信息稀疏性等特点，而交通系统对基础数据处理也有高时效性的要求，这就要求系统要在前期对繁杂、海量、质量不一的交通数据进行预处理，以提高系统工作效率。因此海量数据下，如何提高城市路网交通数据利用率及传统智能交通系统数据如何支撑融合业务协同则成为了制约智能交通行业发展的核心要素。本文针对智能交通系统发展新趋势下，融合数据比例高、种类多、管理难度大的问题，采用基于ETL数据融合及工程项目数据管理知识，提出一种基于ETL的MapReduce工作流与数据融合管理应用相结合的模型，并通过具体项目应用来检验该模型的有效性与可行性，最终实现场景应用。首先，本文对智能交通系统建设现状、发展趋势进行分析，整理并总结出系统对基础数据，尤其是融合数据的要求越来越高，需要一个新的、有区别于以往单一数据管理模型以满足新系统对数据管理的要求。其次，针对智能交通数据融合管理需求分析结论，结合融合数据管理的特点与相关数据管理规范，提出一种基于ETL的MapReduce工作流与数据融合管理应用相结合的模型，并阐述在典型交通业务场景下，融合数据管理模型的应用。再者，将提出的模型应用在智能交通系统建设工程项目的数据融合管理工作中，结合数据现状及融合管理需求，从整个交通管控平台系统架构中，规划数据架构及数据融合管理模型，如交通管控平台数据库进行分析及数据库设计等，完成项目中的数据融合管理工作。最后，将数据代入检验是在系统真实运行环境中，筛选四个典型系统页面响应效果以检验数据融合管理模型应用的有效性。经系统测试数据可验证四个页面均满足测试性能要求，证明模型能对智能交通大数据融合管理应用起支撑作用。该模型满足新时代下智慧交通发展大数据决策分析要求，为解决多源异构环境下交通数据融合处理提供了管理流程与方法。为交通治理工作向智慧化、智能化、高质量发展提供了参考与借鉴。
In recent years, with the rapid development of China‘s economy, the vehicle ownership rate has increased. On the one hand, it represents the improvement of the living standard of Chinese residents, and on the other hand, it also brings pressure on urban road traffic in China. Traffic management departments around the country have built intelligent transportation systems to strengthen regional traffic control. With the increase of the number of intelligent transportation systems, a large amount of traffic data has also been generated. Due to the particularity of the industry, the transportation big data has the characteristics of multi-source, heterogeneous, local, spatio-temporal correlation and information sparsity, and the transportation system also has a high timeliness requirement for basic data processing, which requires the system to preprocess the complicated, massive and different quality traffic data in the early stage to improve the system efficiency. Therefore, how to improve the utilization rate of urban road network traffic data and how to support the integration business collaboration of traditional intelligent transportation system data have become the core elements restricting the development of intelligent transportation industry under the massive data.Aiming at the problems of high proportion, multiple types and difficult management of fusion data in the new trend of intelligent transportation system development, this paper adopts ETL based data fusion and engineering project data management knowledge to propose a model combining ETL based MapReduce workflow and data fusion management application, and tests the effectiveness and feasibility of this model through specific project applications, Finally, the application of multi-source heterogeneous data fusion management of urban road traffic is realized.First of all, this paper analyzes the current situation and development trend of intelligent transportation system construction, collates and summarizes that the system has higher and higher requirements for basic data, especially fusion data, and needs a new, different from the previous single data management model to meet the requirements of the new system for data management.Secondly, according to the analysis conclusion of the demand for intelligent transportation data fusion management, combined with the characteristics of the fusion data management, a MapReduce workflow model based on ETL is proposed. Combining with the relevant data management specifications, a fusion management model that can meet the requirements of urban road traffic multi-source heterogeneous data is designed, and the application of the fusion data management model in typical traffic business scenarios is described.Furthermore, the proposed model will be applied to the data fusion management of the intelligent transportation system construction project. In combination with the data status and fusion management requirements, the data architecture and data fusion management model will be planned from the whole traffic control platform system architecture, such as the analysis and database design of the traffic control platform database, to complete the data fusion management of the project.Finally, the data substitution test is to screen the response effects of four typical system pages in the real operating environment of the system to test the effectiveness and feasibility of the data fusion management model. The system test data can verify that the four pages meet the test performance requirements, that is, the data fusion management model proposed in this paper can meet the application requirements of urban road traffic system for data. It is proved that the data fusion management model can support the application of intelligent transportation big data fusion management.The management model meets the requirements of big data decision analysis for intelligent transportation development in the new era, and provides management processes and methods for solving the traffic data fusion processing in multi-source heterogeneous environment. It provides reference and reference for the intelligent, intelligent and high-quality development of traffic management.
