Research On Situation Awareness Of Rail Traffic Security

ABSTRACT:Rail traffic has a key position in the transport system and has played an increasingly important role in national economic development. In the event of accidents, rail traffic system will result in a large number of casualties and property losses, so the protection of rail traffic safety is especially important. Security situation awareness is an emerging security technology that can analyze the system's security status and security trends. The thesis has studied issues of situation awareness of track traffic security and the main contents are as follows:(1) After research results at home and abroad in the field of rail traffic security and security situation awareness are reviewed, the necessity of applying security situation awareness to rail traffic security is analyzed. The definition and connotation of situation awareness of rail traffic security is cleared, and the technical route of adopting suitable methods of security situation awareness for different rail traffic system is established.(2) With the combination of cause-consequence diagram and risk analysis, situation awareness of rail traffic safety based on cause-consequence diagram for rail traffic system with a clear mechanism is proposed. First risk varying with time is determined by cause-consequence diagram, and then the risk curve is used to evaluate the rail traffic system's security status. At last, the motor overheating accident analyzed by the cause-consequence diagram demonstrates the security situation awareness of rail vehicles.(3) The hidden markov model has the function of security state training and identification so that it is used to the situation awareness of rail traffic system with an unclear mechanism and easy access to system output. The first step is to use Baum-Welch algorithm to train the HMM of various security state, and the second step is to use the forward algorithm based on dynamic programming to do probabilistic inference and security state identification. Through the use of hidden markov model to achieve security situation awareness of the vehicle bogie suspension system, the feasibility and practicality of situation awareness of rail traffic safety based on HMM is demonstrated.(4) Based on the characteristics of support vector machine classification and regression, the method of security situation awareness of rail traffic system with unclear mechanism and easy access to system input is put forward. Support vector machines classification is used to evaluate the security state, and support vector machine regression can predict the security parameters, so as to achieve the purpose of security situation awareness. In the case of analysis of rail curve passing-through using support vector machines, cross validation is used to optimize the penalty coefficient C and radial basis function parameterγ. In the end, validation data show clearly the accuracy of security situation awareness of rail curve passing-through based on support vector machine.