公路交通生命线多源信息融合技术研究

Study on technology of  multi-source information of highway lifeline system fusion

研究生：陆纯

指导教师：郭忠印教授

二〇一四年五月

摘要

近年来，受自然环境和人类活动的影响，我国自然灾害不断增多，对人民生命线财产造成的损失也日趋严重。公路交通系统在重大自然灾害应急救援和重要活动保障中担负着重要作用，抢险救灾的成效与公路交通的通畅、运输效率有直接关系。快速获取灾后公路交通的状态信息是保障生命线畅通的关键，但灾后初期交通生命线灾情信息来源广泛，数据紊乱，因此，对多源公路交通生命线信息融合技术的研究至关重要。
     论文首先阐述了公路交通生命线的含义，根据相关文献资料和实地调研结果，归纳分析了典型重大自然灾害对公路交通生命线的破坏类型；基于混合分类法对公路交通生命线灾情信息进行分类；根据灾后应急救灾阶段的救灾侧重点对公路交通的要求，提出公路交通生命线灾情信息需求。

基于国内外现有的信息采集技术，构建满足灾情信息需求的灾后公路交通生命线信息采集网络，分析了各信息采集方式的实际应用过程。针对遥感图像公路特征信息提取进行深入研究，提出基于ENVI和ArcGIS的道路损毁信息提取方法。同时，分析了不同信息传输方式的适用性，保障灾情信息的快速传递。

分析信息采集网络获取的灾情信息特点，在信息融合定义的基础上，阐述了多源公路交通生命线灾情信息融合的含义。依据信息源特点及灾情信息需求，提出多源公路交通生命线信息融合模型；针对不确定信息的融合，提出一种改进的D-S证据理论融合算法，该算法通过信息源实时采集的数据对其可靠度进行量化。并且通过一个算例说明了该融合模型的具体应用。

最后，对本文的研究成果进行了总结，指出了不足之处，对今后的研究方向进行了展望。

关键词：公路交通生命线，信息需求，信息采集，遥感图像，多源信息融合，D-S证据理论

ABSTRACT

In recent  years, affected by the natural environment and human activities increasing  China's natural disasters，loss of property caused by the lifeline of the people are becoming  increasingly serious．Road transport system assumes significant role in major natural  disaster emergency relief and protection of important activities，the effectiveness of disaster  relief and smooth transport efficiency is directly related to road traffic．Quick access to state  disaster Highway Traffic flow information is critical to protect the lifeline  but the initial disaster extensive transportation lifeline disaster  Information sources, data disorders, therefore, multi-source information  fusion vital lifeline road transport technology research.

Firstly,  expounded the meaning of road transport lifeline，according to the relevant literature  and field research results，summarized and analyzed the typical type of major natural  disasters destroy Highway Traffic lifeline；classification method based on hybrid lifeline  disaster information highway classification according disaster emergency  focus on disaster relief phase of road traffic requirements，proposed highway lifeline  disaster information needs.

Based on existing  information collection technology at home and abroad to build to meet the  information needs of post-disaster disaster lifeline road traffic information  collection network，analyzed the actual application process each information  collection methods. Information extraction depth research on remote sensing  images road characteristics, proposed corrupted information extraction method  based on ENVI and ArcGIS road. Meanwhile, the analysis of the applicability of  the different methods of information transmission，ensure rapid delivery of  disaster information.

Characteristics  of disaster information network to get information collection based on  Information fusion defined elaborated multi-source information fusion  disaster lifeline road traffic implications. Based information source  characteristics and information needs of the disaster，road transport lifeline  proposed multi-source information fusion framework；against uncertain information  fusion.

Finally, the  results of this study are summarized, pointed out the inadequacies of the  future research directions are discussed.

Key words：highway lifeline,  information needs, information collection, remote sensing Images,  multi-source information fusion, D-S evidence theory