Study On Prediction Method For Railway Track Irregularity

Security is always the subject of rail transport.To guarantee the railway transportation's security,the research on the change state of railway track must be thoroughly put in practice.Study on the change state of rail track quality is regarded as the study object,by starting from the analysis of the track geometry inspection data. Based on the analysis of track irregularity prediction method abroad,the author established the prediction model of track irregularity.And it is of great practical significance to accurately master the transformation trend of rail track quality and scientifically arrange track maintenance,especially to the rail track maintenance of Passenger Transportation Special Line.In this paper,based on the analysis of track irregularity prediction method abroad, the author made the following main aspects of the work.1) From the aspect of collecting inspecting data,papers analysis data format and quality problems of different vehicles.Using Data Warehouse to store and manage dynamic and static track inspection data.The Data integration processes and methods described.2) From the aspect of analysis inspecting data,by comparing two different analysis way to get their advantage and disadvantages.Through on-site research,by analysis of data at the scene of the actual work and pointed out the problems.3) Based on the analysis of track irregularity prediction method abroad,combining the actual situation of Kunming,the author use one variable linear regression and grey forecast model to forecast the growth of track irregularity.Considered the randomness of data,the author use the gray-markov to predict the growth of track irregularity for the first time,and the result is good.4) With the theory study,according to the present situation of Kunming Railway Administration,Kunming Railway Engineering Production Management System is constructed,using the rail track irregularity predict model discussed in this thesis.And it is already taken into practice.