Theory And Methods For Optimization Of Electric Multiple Units High-level Maintenance Plan

As a significant part of railway transportation,high-speed railway plays an important role in completing passenger transport tasks efficiently and improving the service quality of passenger transportation.After more than ten years of leap-forward development,remarkable achievements have been made in the construction of the China's high-speed railway.With the opening of high-speed railway lines,the number of EMUs is also dramatically increasing since EMU trains are the only carriers of high-speed railway.Therefore,how to operate EMUs on such a scale efficiently and reasonably is a problem faced by railway operators.High-level maintenance plan contains the third-level,the fourth-level and the fifth level maintenance.There are various factors,broad planning range and high costs involved in the high-level maintenance planning process,of which the optimization has become a vital issue.Generally speaking,the high-level maintenance plan is the core document for guiding the operation and maintenance work of EMUs,and plays an important role in ensuring the safe operation,meeting passenger traffic peak demand,and using maintenance resources rationally.It is also the precondition of the EMU train utilization,daily maintenance and the work organization of the EMU train maintenance base.Currently,the EMU high-level maintenance plan is made basically based on personal experience and Excel worksheets,which is already difficult to find a local feasible solution,not to mention a global optimization plan.In practical operations,the manual high-level maintenance plan often requires constant adjustments in order to barely complete maintenance tasks.The operations of EMU high-level maintenance are generally excessive based on the manual plan under the cross-constraints of the limited maintenance capacity and the great passenger demand during peak period,which has seriously affected the utilization efficiency of EMUs.Therefore,it is of great theoretical and practical significance to study the scientific and rational optimization approaches of the high-level maintenance plan,which can improve the utilization efficiency of EMUs,meet the passenger demand during the peak period,reduce maintenance costs and make full use of maintenance resources.In addition,as the increasing number of EMUs,the optimization of high-level maintenance plan can reduce maintenance costs,increase passenger transportation revenue and bring huge economic benefits on one hand.On the other hand,it can also satisfy the passenger demand during the peak period as much as possible,and then bring inestimable social benefitsAt present,experts and scholars around the world focus on the EMU train utilization,daily maintenance and the maintenance process optimization in the EMU maintenance bases,while there are barely literatures devoted to the optimization problem of EMU high-level maintenance plan according to the author's knowledge.This thesis concentrates on the high-level maintenance plan optimization,the focuses of this thesis can be summarized as follows:(1)The current management situation and future development trend are explicated and the related research status is summarized based on the practice of EMU management.The logical relations among sub-problems of EMU management are clarified.The practical constraints that affect the design of the high-level maintenance plan are emphatically analyzed.The objectives and the input and output data are defined clearly and the mathematical background are extracted,which lays a solid foundation for studying the comprehensive optimization method of EMU high-level maintenance plan from the quantitative perspective hereafter.(2)The optimization approach of the EMU high-level maintenance plan with fixed maintenance spots based on the state function is analyzed.For this problem,through analyzing the hold and shift of EMUs' states in planning horizon,a 0-1 step function is proposed to demonstrate the states of the EMUs based on Kronecker delta function.Furthermore,a non-linear 0-1 integer programming model is established considering a series of practical constraints comprehensively.For the sake of adopting commercial solvers,non-linear programming model is transferred into linear ones.Several heuristic algorithms for solving mathematical model are developed based on simulated annealing strategy,genetic evolution strategy and particle swarm optimization strategy respectively.And the results of these algorithms are also compared with accurate solutions.In addition,the coding design of 0-1 integer programming model is completed by real number coding approach.(3)The optimization approach of the two-stage EMU high-level maintenance plan problem based on the time-space network technology is studied.Firstly,the EMU high-level maintenance plan optimization approach of fixed maintenance spots is analyzed.The plan optimization problem is transferring to path selection problem through constructing discrete time-state network which can reveal current and shifting process of the EMUs'states in the planning horizon.Based on this,a multi-commodity model is constructed.The result comparison of two models based on state function and time-state network of the same case shows that,the multi-commodity model based on time-space network technology has a higher computational efficiency and a more concise expression.Secondly,the dimension of space is added into the time-state network to study the optimization approach of EMU high-level maintenance plan considering compatible and multiple maintenances.The corresponding multi-commodity model is established and solved by commercial solvers.Besides,a linear 0-1 integer programming model is constructed directly based on 0-1 integer modeling approach.(4)The optimization approach of the EMU high-level maintenance outline plan problem based on multi-objective programming model is investigated.For the multi-objective EMU high-level maintenance outline plan which takes both maximum utilization efficiency of EMUs and maximum revenue of EMU depots into account,the granularity in time is fractionized into "month",and a multi-objective programming model is constructed considering corresponding practical constraints.Besides,to make the model more concise,the number of EMUs in maintenance at workshops under the Euler coordinates and the current states of EMUs under the Lagrangian coordinates are coupling.At last,a corresponding solving approach is designed based on ?-constraint approach to obtain an approximate Pareto-optimal solution in order to provide several optimized plans with various "preferences" to the management apartment.(5)Real-world case studies are carried out based on the Shanghai Railway Bureau supported by the projects "Analysis of EMU train maintenance technologies-Study on the EMU high-level planning optimization".The rationality and validity of the proposed method are verified by applying above optimization approaches based on the real application background and operation data.