Research On Model Of Disruption Management In Railway Emergency Resources Scheduling

Railway serves as one of the most important means of transportation with a low possibility of emergency occurrence. However, once railway emergencies occurred the accident would bring about variety of unpredictable negative effect. As emergency resource turns out to be a major guarantee for emergency relief, rescue commanders have to formulate a scientific and reasonable solution to schedule emergency resource quickly based on disaster situation. In another word, the commanders should make the most of limited emergency resources and transportation capability to distribute large amount of emergency resources, then do the rescue and relief work at full speed thus for minimize losses. In addition, the uncertainty characteristic of the emergency may induce further unpredictable secondary accidents, which would lead to a phenomenon that the demands in emergency resources in accident spot change known as disruption incident. Disruption incidents would interfere with the implementation of original optimal resource scheduling solution. At this point, losses from this emergency would increase if resource scheduling solution were not adjusted immediately, beset by unmet demand in part of accident spots. As a result, it has great social significance to study how to effectively dealing with the disruption incidents and realize the minimizing impact on the whole emergency resource scheduling solution.Disruption management is a method to cope with disruption incidents in real time, which aims to minimize the deviation of original scheduling solution by adjusting it. On the basis of situation after disruption incidents, full consideration should be given to original scheduling solution and current new constraints, thereby estimating possible adverse impacts caused by disruption incidents and finally making a new scheduling plan that meets current demand with the smallest adjustment of unallocated resources.This thesis focused on studying emergency resource scheduling when railway accidents occur and addressing infeasibility of initial resource scheduling solution caused by disruption incident with the guidance of disruption management method. First of all, given the restriction the reserve, transport capacity, distance or time in the emergency resource and upgraded Technique for Order Preference by Similarity to an Ideal Solution(TOPSIS) method would be utilized to determine the supply priority of emergency resources and then establish an initial dynamic multi-stage scheduling model based on supply priority as a key factor and aiming to achieve the minimal losses in the whole system. Secondly, analysis should be made to evaluate the possible occurred secondly incidents whether makes the scheduling initial solution inaccessible. Disruption incidents fall into three categories by the nature of railway accidents and its impact on scheduling solution. The disturbing degree would then be measured from the perspectives of accident spot, resource scheduling network and railway authority. Moreover, when disruption incidents occurred, minimal relative losses to the initial scheduling solution and minimum disruption disturbing degree should be sought for by developing a multi-target optimized disruption management model. Considering the complexity of developed models and the urgency of rescue in this thesis designed an improved bat algorithm that can be used to solve problem quickly. Finally, through a simulated a railway accident example validates the effectiveness of the initial dynamic multi-stage scheduling model and the disruption management model in the context of MATLAB, while uses the improved algorithm to calculate this example in order to validate the reasonability of the improved bat algorithm.