The Research On Traffic Network Failure Degree And Safety Level Evaluation Method

The traffic network management takes an important area in the transportion reseach.Which means the same to the traffic network failure degree evaluation in the traffic network management.It can not only supplies an objective evaluation to network operation situation and reliability but also provides an effective decision surport by quantificational evaluation of the traffic network failure degree. The network safety level reflects the probability that the network can operate normally.It is one of the most pivotal target of network management.The safety level can be evaluating by the failure data of the network under different types of attack.The bayesian network have some superiorities compare to other methods in modeling such as graphing and incomplete data theorization.The grey system theory can quantify the network safety level with incomplete information.The paper does main research on the application of bayesian network on the traffic network failure degree evaluation and the application of gray fixed-weight clustering on the road safety level, and has such content as:(1)The paper gives the method of qualify the traffic network failure degree with the connectivity reliability to make the failure degree's defination more clearly.(2)The paper does some research on three methods of evaluating the traffic network failure degree based on bayesian network. There are the method that based on fault tree and event tree,the method based on knowledge and the method based on case and rule.The paper uses these methods to build bayesian network to the sample road network.And then use the software MSBNX to qualify the failure degree under different types of attack on these road networks. (3)The paper uses the traffic network failure degree as the qualification standard of traffic network safety evaluation. Applying the gray fixed-weight clustering method to evaluate the safety level of the sample road networks which was given in the chapter 3.Finally,the research work involved in the thesis is summarized and the future developments in bayesian network learning in traffic network are forecasted.