Multi-Source Information Fusion Model of Traffic Lifeline Based on Improved D-S Evidence Theory

Abstract

The status information collection of the post-disaster traffic lifeline has many characteristics, such as wide ranges of sources, different information formats, and severe changes in time and space. In order to improve the processing accuracy of multi-source heterogeneous information, an information fusion model based on improved D-S evidence theory was proposed, which relied on information source reliability. Through collecting different traffic-lifeline state information, the features could be extracted from the collected information, and the preliminary decision results about the lifeline state was obtained correspondingly. According to certain fusion criterion and the reliability of each information source, the final fusion result was obtained with an example applied to certify the specific application of the model.