Research On Optimization Of Air-ground Cooperative Dispatching Based On Prediction Of Earthquake Disaster Emergency Resource Demand

Earthquake has become the most important disaster because of its destructive power,strong suddenness,high defense difficulty and far-reaching social impact.In the various links of post-earthquake emergency rescue,the prediction of emergency relief materials is the basic work of emergency logistics,and its accuracy directly affects the follow-up work,while the dispatching and distribution of materials is the core problem of the optimization of the whole rescue decision,and is the key link of emergency rescue.At the same time,aircraft has gradually joined the emergency rescue transportation system due to its advantages of flexibility,quick response and less geographical space restrictions,increasing the diversity and flexibility of rescue means.As the uncertain demand and insufficient supply capacity of emergency rescue materials in the post-earthquake emergency resource,this thesis studies the prediction of postearthquake emergency resource demand and the sky ground cooperative scheduling model,and designs the corresponding models and methods,which can effectively predict the demand of emergency materials and provide an efficient and flexible air ground cooperative dispatching scheme.The models proposed in this thesis improve the efficiency of the whole rescue operation,minimize the loss of earthquake and its secondary disasters.In view of the above problems,the main research work of this thesis is as follows:(1)By analyzing the influencing factors of the number of casualties after the earthquake,the corresponding index system is established,and the two-stage emergency resource demand prediction model of the number of casualties and material demand after the earthquake is constructed.The prediction value of the number of casualties is obtained by using the improved case-based reasoning method based on the idea of KNN algorithm and the hybrid algorithm of particle swarm optimization and neural network.Through the number of casualties,combined with the relevant knowledge of inventory management,the demand for various materials at each disaster point is indirectly predicted.The results and performance of the algorithm designed in this thesis are compared with different algorithms to verify the effectiveness of the prediction model and hybrid algorithm designed in this thesis.(2)By analyzing the influencing factors of vehicle scheduling and aircraft scheduling in air-ground cooperative emergency rescue,the constraints such as vehicle quantity constraint,demand constraint and vehicle performance are established.Taking the shortest emergency rescue time and minimum emergency rescue cost as the optimization objective,the air-ground cooperative scheduling model of multi material distribution points,multi transfer points and multi disaster points is established,and the design of genetic immune hybrid algorithm is provided for the solution of the model.The results and performance of the hybrid genetic immune algorithm and the single algorithm are compared to fully verify the effectiveness of the hybrid algorithm in solving the air-ground cooperative scheduling problem established in this thesis.