Research On Repair Crew Scheduling For The Damaged Road Network With Enormous Demand Points

Today,the construction time of the Chinese emergency management system is still short,and the development of basic support technology of emergency management is still relatively weak.It is still facing great challenges to prevent and resolve the major risks and effectively deal with extraordinarily serious natural disasters.With the rapid development of the new generation of information technology characterized by "big,smart,cloud",it provides new opportunities and challenges for the informatization,intellectualization,and scientization of emergency management.Consequently,in the“strategic planning framework for the development of emergency management informatization(2018-2022)”,the Chinese emergency management department clearly proposes to use big data,artificial intelligence,machine learning,and other new generation information technologies to improve the ability of risk perception,monitoring and early warning,collaborative rescue and emergency response of extraordinarily serious natural disasters,so as to provide scientific and technological support for promoting the modernization of emergency management system,and finally realize the smart emergency of the big China.In the post-disaster response,repairing the damaged road network is a basic premise for emergency response and rescue after the occurrence of severe natural disasters.This dissertation mainly studies how to reasonably schedule the road repair crew to quickly restore the smooth road network to ensure that the rescue crews and emergency resources are timely transported from the rescue point to each demand point,and provide guidance for the rapid implementation of emergency rescue,which is of great practical significance to minimize the disaster losses.However,the existing studies rarely consider the problem of continuously damaged road sections,especially for the road network with enormous demand nodes and serious damage,and have difficulty in giving an effective solution.Additionally,most of the existing studies focus on the decision-making of a single repair crew and rarely consider the overall scheduling and cooperative operations of multiple repair crews.To this end,this dissertation studies the repair crew scheduling for the damaged road network with enormous demand nodes.The main research work of this dissertation is summarized as below.(1)The research significance and objectives of the repair crew scheduling for the damaged road network is expounded.The existing work is reviewed,analyzed,and summarized.The existing problems and possible solutions are discussed.(2)To solve the scheduling of a single repair crew,the key factors of repairing the damaged road network are first analyzed according to the road network model and Markov decision-making process,based on which a double-feedback reward function is designed.Then,the deep Q-learning is utilized to solve the optimal scheduling strategy of the repair crew.Finally,the comparative experimental results demonstrate that under enormous demand nodes,the proposed approach has good stability and reliability,considers both the repair and transportation efficiencies of the damaged road network,can make all the demand nodes accessible with less repair cost,and provides a useful attempt to repair the damaged road network in complex emergency scenarios of post-disaster.(3)To solve the concurrent scheduling of multiple repair crews,a road network partition method is first proposed to make different repair teams responsible for different subnets.Next,the learning experience of multiple repair crews is put into the same experience pool of the deep Q-learning to share experience and achieve hybrid learning.Finally,a concurrent scheduling method of multiple repair crews is designed on the basis of the deep Q-learning and the multithreading technology.The experimental results demonstrate that,compared with the traditional scheduling method of the single repair crew,the proposed scheduling method of multiple repair crews significantly improves the repair efficiency and transportation efficiency,which provides a certain technical support for the overall scheduling and cooperative operation of multiple rescue forces in the post-disaster response.