Optimization Of Tamping Maintenance Decisions For Railway Tracks

Heavy-machine tamping is the most important and effective maintenance operation adopted by railway permanent way departments to improve the track geometry of ballasted track,and it is also one of the most expensive track maintenance operations.At present,periodic tamping maintenance is mainly performed on Chinese railways.With the continuous expansion of railway network scale,the development of high-speed and heavy-load trains,the traditional tamping maintenance mode has been difficult to meet the requirements for safety,stability and reliability of track state in railway transportation,as well as the urgent needs of managers in maintenance cost control.Therefore,the railway tamping maintenance mode in Chinese railway is gradually changing from periodic maintenance to preventive condition-based maintenance.In order to realize this transformation,it is necessary to study and solve the key problems such as the accurate prediction of track geometry deterioration rules and tamping maintenance cycle,the reasonable optimization of tamping maintenance planning and intervention scheduling.Focusing on the preventive tamping maintenance decision-making of railway track,this paper studies three aspects of tamping maintenance cycle prediction,optimization of long-term tamping maintenance planning and optimization of short-term tamping intervention schedule.The individualized prediction model for tamping maintenance cycle of railway track segment,bi-objective optimization model for tamping maintenance five-year planning and bi-objective optimization model for monthly tamping work scheduling are developed.The main contents of the dissertation are as follows.(1)An individualized prediction model of tamping maintenance cycle for track segments based on Time-Transformed Wiener process(TTWP)is proposed.Considering the heterogeneity and uncertainty of track geometry deterioration process,the linear,continuous and long railway track is divided into several 200-m-long track segments.Taking each track segment as the modeling object,the TTWP method and the track geometry inspection data and tamping maintenance management data of each 200 m segment is creatively used to individually model the deterioration rule of longitudinal level standard deviation between two adjacent tamping interventions.On this basis,the tamping maintenance cycle of each 200 m track segment is predicted according to the tamping maintenance threshold of longitudinal level standard deviation.Taking 2187200 m track segments(mileage range from K548 + 000 to K985 + 400)of Lanzhou Xinjiang railway as the case study objects,the validity of the proposed model is verified using the longitudinal level standard deviation inspection data and tamping maintenance record data from April 2015 to November 2018.The results show that the proposed model can assist the managers accurately analyze the tamping maintenance demand of railway track,which can provide decision support for reasonable arrangement of preventive tamping maintenance.(2)A bi-objective optimization model for railway track tamping planning considering reliability-maintenance cost trade-off(BORTTP model)is proposed.Aiming at the five-year planning for tamping maintenance of a railway line,the model takes each200 m track segment as the decision-making unit and takes a quarter as the decisionmaking time step.On the basis of predicting the track geometry condition and tamping maintenance demand of each segment of the studied line using the model in Chapter 3,the BORTTP model takes the maximal mean reliability of the track and minimal maintenance costs during the planning period as the two objectives,and considers the constraints of the latest tamping time based on tamping maintenance cycle and the earliest tamping maintenance time based on residual life.On this basis,the five-year planning scheme for tamping maintenance of a line,that is,when to perform tamping operations on which track segments in the planning period,is determined using the optimization algorithm based on Nondominated Sorting Genetic Algorithm II(NSGA-II).The model considers the opportunity maintenance strategy by considering the track possession cost of tamping maintenance in the objective function of maintenance cost,and considers the life loss by considering the constraint of the earliest tamping maintenance time.In order to verify the validity of the model,the author takes 2187 200 m track segments(mileage range from K548 + 000 to K985 + 400)of Lanzhou Xinjiang railway as the case study object,and uses the BORTTP model and the designed algorithm to solve the five-year tamping maintenance planning scheme of the studied line,and compares the solution results with the actual management data and other models(the model without considering the opportunity maintenance strategy and the model without considering the life loss).The results show that:(1)the model can provide managers with several alternative tamping maintenance planning schemes,and can help managers to intuitively analyze the impact of tamping maintenance costs on track reliability;(2)the average annual tamping maintenance workload corresponding to the solutions of BORTTP model is generally lower than the actual values,at most 21.4% lower than actual;(3)compared with the model without considering the opportunity maintenance strategy,the track occupation cost corresponding to the BORTTP model is lower and more consistent with the actual tamping maintenance management needs;(4)compared with the model without considering the life loss,the optimization result of BORTTP model can avoid the occurrence of unscientific tamping maintenance interventions such as over-tamping and pre-tamping.(3)A bi-objective optimization model for monthly tamping intervention scheduling considering possession time-track geometry trade-off(BOMTIS model)is developed.Aiming at the monthly tamping intervention scheduling of lines in a permanent way depot,the BOMTIS model takes the minimal track possession time and the optimal track geometry as the two objectives,and considers the constraints of tamping maintenance planning scheme,maintenance resources,movable range of tamping vehicle in a maintenance window and minimum intervention workload in maintenance window.On this basis,the monthly tamping schedule including intervention date,intervention time and intervention mileage location is determined using the optimization algorithm based on NSGA-II.Taking the tamping intervention schedule for Lanzhou Xinjiang railway line within jurisdiction of Jiayuguan permanent way depot in April 2016 as the case,the author uses the BOMTIS model proposed in this paper to solve the intervention date,intervention time and intervention mileage location of the tamping maintenance of the studied line in April 2016,and compares the solution results with the actual tamping maintenance intervention schedule data to verify the effectiveness of the model.The results show that the model can provide managers with several alternative schemes of monthly tamping intervention schedule,and compared with the actual tamping intervention schedule,it can achieve acceptable track geometry condition level with less track possession time.