Optimization Of Emergency Dispatch Considering Heterogeneous Behaviors

With the frequent occurrence of major natural disasters and public health emergencies,how to carry out emergency dispatch scientifically and efficiently is a crucial problem in emergency relief.Now,human behavior is a hot research topic.However,behavior differs from person to person,that is,the behavior is heterogeneous.Aiming at the heterogeneity of behaviors,the paper focuses on the learning behaviors of decision makers and discusses the influence of its heterogeneity on emergency deployment decisions.It uses a learning effect function that includes the initial decision time and learning rate to quantify the learning behavior.And the initial decision time mainly refers to the time of emergency response decision,namely,the time to decide whether to respond to the disaster area.Thus,the paper first studies emergency response decision-making and explores relevant factors that affect its heterogeneous behavior.An empirical study on the correlation between the factors and its decision and decision speed is conducted to determine an indicator system to quantify the initial decision time.Then,combined with the differentiated learning rate to comprehensively characterize the heterogeneous learning behavior,so as to study the location-allocation problem in the emergency dispatch considering the heterogeneous learning behavior.The details are as follows:First,taking the Jiuzhaigou earthquake in 2017 as the background,the paper uses casebased empirical research methods to study the behavioral heterogeneity of alternative city rescuers in emergency response decision-making.Based on the Baidu index,the paper innovatively proposes the “emergency attention”.Combined with the economic development level,infrastructure competitiveness,transportation accessibility,safety,logistics development degree and natural geographic location determined by domestic and foreign literature,seven variables are constructed.And it constructs two discrete choice models to explore the correlation between these seven variables and the city's emergency response decision-making and its speed.Using the cross-section data related to the Jiuzhaigou earthquake,the two models are estimated separately.It found that the combined impact of the seven variables on the emergency response decision-making and its speed of the candidate relief city is significant,and two new indicator systems are also determined.Then,the paper mainly studies the location-allocation problem in emergency dispatch decisions,and discusses the impact of heterogeneous behavior on it.Aiming at the learning behavior existing in the decision-making process,the decision-makers of supply points and distribution centers are selected as the main research subjects.The learning effect function is used to describe its learning behavior,and the differentiated initial decision time and learning rate are used to comprehensively characterize the heterogeneity of learning behavior.The new index system determined in the empirical research on emergency response decision speed is used to measure the initial decision time of each supply point and distribution center.Similarly,a comprehensive indicator system is constructed to measure the learning rate of each supply point and distribution center.And an emergency material location-allocation model with the objective of minimizing cost and decision time is established.The NSGA-? algorithm is designed and applied to the simulated model in the context of the 2017 Jiuzhaigou earthquake case.The solution results are compared with the location-allocation scheme without considering the heterogeneity of learning behavior,and the sensitivity analysis of key parameters is performed.Finally,we draw scientific research conclusions and put forward relevant management opinions and suggestions,which provides a useful reference for decision makers to make correct decisions quickly and effectively.