Research On Seheduling Optimization Problems In Enterprises Railway Marshalling Yards

Railway is the main way of enterprise transportation, so railway occupies a dominantposition in the large enterprise production. At present, most of enterprises marshalling yardsare small and use artificial scheduling method. Marshalling yards often occur overload andtraffic jam phenomenon or even happen catastrophic accidents, which lead to lower efficiencyof railway transportation. With the expansion of enterprise production scale and transportvolume, traditional scheduling method can not content with the requirement of enterprisesproduction. So the research on enterprise railway marshalling yards dispatching optimizationproblem has an important practical significance. The dissertation takes marshalling yardsdispatching system as research object. On the detailed analysis of the operation and process inmarshalling yards. Then hybrid optimization strategy which combined optimization algorithmand expert system is put forword.Operation characteristics of enterprise marshalling yards is analysed in detail. The hybridoptimization strategy which combined optimization algorithm and expert system is putforword to optimize scheduling system. The new optimization method divides transportationscheduling model into six sub-modules: train grouping, appliation of departure line, trainformation, station route choice, locomotive supply and tumaround, operations for placing-inand taking-out wagons. Then the new method uses different optimization strategies to solveevery sub-module.The optimization problem of branch-shaped placing-in and taking-out wagons is studied.The type and operation characteric of the branch-shaped placing-in and taking-out wagons isanalysed in detail. On this basis, a mathematical model for several trains arriving at station inbatches is built; meanwhile, genetic ant colony algorithm (GACA) is put forward to resolvethis model. The principle is to make full use of traction and optimization goal is the least timeof vehicle total consumption. Compared with traditional model, this new model regards order,time and batch as a system and applies to a variety of delivery mode of operation. At the sametime, this new model considers the traction capacity of adjustable machine and vehicleholding ability of loading areas. The simulation result verifies rationality and feasibility ofthis model. And by contrasting with ant colony algorithm, the superiority of this algorithm isillustrated.The optimization problem of routing selection is studied. A network topological structurefigure is built according to the distribution characteristics of marshalling yards. Then amathematical model whose optimization goal is the least delay time and the least walkingtime is built, then GACA is put forward to resolve this problem. This model can describeaccurately the relation between space and time of selected route, and can arrange reasonably the routing order according to the priority of the job. GACA can find out the satisfactorysolution in a short time, and then can optimize the route selected problem.