Research On Reliability Evaluation Method Of High-speed Railway Dispatching System

Dispatching system is one of the cores of railway transportation organization and an important barrier to ensure safe,punctual and efficient operation of trains.In recent years,China's high-speed railway has developed rapidly.By the end of 2019,China's high-speed railway has reached 35,000 kilometers,ranking first in the world.The characteristics of high speed,high density and large traffic volume of high-speed railway bring severe challenges and higher requirements to the dispatching system.Strengthening the reliability and safety of high-speed railway dispatching system is very important to the safe operation of high-speed railway.High-speed railway traffic dispatching system is a highly coupling system with "man-machine" interaction.Its structure is huge,its elements are numerous and its functions are complex.All kinds of equipment in the system are not only independent,but also have certain relevance.In addition,the randomness of "people" in the system makes it very difficult to study the reliability of high-speed railway traffic dispatching system.Therefore,at present,most of the researches on the reliability of high-speed railway traffic dispatching system remain on a qualitative level,without in-depth research.Aiming at the problem that the research on system reliability is not deep enough at present,this paper starts with the key equipment and nodes of the system,and deconstructs and analyzes the reliability in depth around the four dimensions of hardware,software,human factors and system hierarchy.The main contents of this paper are as follows:(1)The organization structure,post setting,functional tasks and information interaction of the system are combed in detail.According to the information transmission mechanism and information attributes of the system,the topological network structure diagram of the high-speed railway traffic dispatching system is constructed based on the complex network theory.Through the analysis of the importance of system nodes and edges,the core position and function of the high-speed railway traffic dispatching system in the whole dispatching system are verified.At the same time,the conclusion that the train dispatcher is the central key node and the autonomous machine is the station key node is also drawn.(2)There will be a large number of fault records in the operation or maintenance of the highspeed railway traffic dispatching system.Due to the lack of effective processing technology,the fault records cannot be effectively utilized.To solve such problems,this paper constructs an algorithm model based on TF-IDF and Text-Rank,and extracts the keywords of system faults through this model.On this basis,the key features of system faults are extracted by using Dirichlet model.By analyzing the key words and subject characteristics of the system,it is found that the station system is the faultprone location of the high-speed railway traffic dispatching system,and the autonomous machine is the fault-prone equipment of the station subsystem.(3)To solve the problem that the state equation of autonomous computer equipment in the system is difficult to solve in reliability analysis,a formula method based on Markov process is proposed.The formula makes the state reliability analysis no longer be solved by complicated Laplace transform or C-K(Chapman-Kolmogorov)equation.The calculation results prove that the method is consistent with the results obtained by Laplace transform method.(4)Aiming at the problem of neglecting the fault of switching unit and not considering maintenance workers in the reliability analysis of autonomous computer equipment,an extended Markov process method is proposed.By introducing the supplementary variable method,the Markov process can still carry out reliability modeling analysis on autonomous machine systems with maintenance workers' vacation time and maintenance time following the general distribution.Through analysis,it is found that the repairman's vacation time has a great influence on the reliability,so the repairman's vacation time should be reasonably arranged under certain cost.(5)Aiming at the problem of irregular detection rate of software faults found during software testing,an improved Non-Homogeneous Poisson Process(NHPP)software reliability analysis model is proposed by introducing irregular model parameters,and the model is applied to the software testing of autonomous machines.According to the test fault data,the reliability requirements under the specified conditions can be calculated when the test days are 45 days.(6)As a "man-machine" interactive system,reliability modeling and analysis for dispatchers are indispensable.Considering that CREAM Basic Law is easy to operate and analyze,an improved method is proposed based on CREAM Basic Law.This method makes the calculation of human error probability by CREAM method no longer an interval value.This method is used to calculate human error probability for dispatchers,and the practical application analysis is carried out with "CTC control mode transformation" and "train control temporary speed limit" as examples.(7)To solve the problem of lack of comprehensive reliability evaluation at system level,an evaluation method based on group decision and interval binary semantics is proposed.Group decision-making method reduces the subjectivity brought by expert subjective evaluation system reliability.Using interval binary semantics as the language of system evaluation reduces the problem of information loss in the process of system reliability evaluation and improves the reliability of evaluation results.