Research On Emergency Logistics Planning And Scheduling

China had been suffering most serious natural disasters in the world wide, which is featured by large variety and high frequency. The occurrence of flood disaster in 1998, SARS in 2003,snowstorm in 2008, Wenchuan earthquake in 2008, and Yushu earthquake in 2010 have remind us of the key to emergency relief, that is, a competent emergency logistics system. It is worthy of note that during 20-year emergency relief process emergency logistics have represented low efficiency and effectiveness. Owing to its unsatisfactory performance, relevant researches aimed at improving emergency logistics operation have increasingly emerged.Accordingly, this thesis mainly deals with issues concerning integration of emergency and business logistics, construction of general logistics information system, unification of emergency and business logistics scheduling, selection of emergency logistics centers, and planning of vehicles used in emergency logistics operation. The issues in question are studied via theoretical models and cases simulation.Firstly, a general logistics information system(GLIS) is designed based on a comparative study between emergency logistics and business logistics, which may be commonly used for largesized logistics enterprise. The GLIS works through an effective coordination of controlling plane and operative plane. Besides, a unified set of standards are needed in managing inter-company data and knowledge. As a result, emergency logistics and business logistics are integrated through business identification and path computation.Secondly, two approaches are proposed to meet the demand for profit margin as much as possible, of which one is called prediction-based scheme and the other is priority-directed scheme.The former may predict the amounts of disaster-relief materials and commercial business and then accept the business that will generate maximum profits; the latter may firstly group commercial and emergency business according to priority grades and then schedules both types of business jointly and simultaneously by arriving at the maximum priority in total. Compared with the prediction-based scheme, the other has the edge of more marginal profit for a more reasonable vehicles scheduling.Thirdly, based on both virtual network method and robust node p-center model method, the problem of selecting emergency logistics centers is discussed on uncertain transportation condition.To start with, a hierarchy network is established by dividing traffic network into backbone transport plane network and the secondary network in consideration of complexity of disaster-affected areas.Furthermore, in accordance with uncertainty of transportation time and unpredictability of occurrence of disasters, this thesis defines the supplies distribution center location as a robust node p-center problem that satisfies time constraints or minimum journey constraint conditions, followed by the simulated annealing method with lower risk of poor solution.Fourthly, the issues related with emergency vehicles planning and scheduling are explored through the multi-dimensional time-space network, improved VRP-TW algorithm and dynamic VRP method. In details, first of all, a random emergency supplies distribution model is proposed based on multi-dimensional space-time network method and then is improved accordingly, which takes into account the concepts of “average” and “buffers” used for random-factor adjustment.Then, the VRP-TW algorithm is improved by using soft- time- window limit constraint with the purpose of optimizing minimum penalty function as the goal, followed by a plug-in algorithm,ultimately achieving the minimum travel time. Next, considering uncertainty of disaster occurrence, this thesis probes into dynamic VRP based on real-time information aimed at minimum route among all penalty function through Tabu searching method commonly used to solve the problem of dynamic update of vehicles, finally realizing minimum total transportation cost. On the conditions where both share the same amount of emergency vehicles and variance, the approach based on soft time window may result in delay and greater punishment, while the dynamic VRP based on real information will leads to minimum punishment with the assistance of GPS and communications facilities.At last, the research focus in near future is presented in accordance with cases study and its result.