Research On Demand Forecasting And Distribution Of Emergency Materials

Natural disasters are often extremely destructive and continuous,and once they occur,they can cause a large number of casualties and bring huge economic losses.After a disaster occurs,in order to alleviate the damage in the affected area and minimize the loss and impact caused by the disaster,rescue should be carried out to the affected area at the first time.In the actual rescue operation,due to the suddenness and destructive nature of the disaster,the information on the demand for emergency materials in the affected area cannot be obtained in a timely and accurate manner,which may lead to uneven distribution of emergency materials and other situations.In order to ensure the quality of rescue and reduce the loss of people’s lives and properties,the demand for disaster relief materials in each disaster area should be scientifically predicted as well as timely and reasonable distribution.How to develop a reasonable and effective distribution plan is crucial for the whole emergency logistics management field.For the problem of predicting the demand of emergency supplies in each disaster area after the disaster,different output layer indicators can be set according to the types of emergency supplies when building the prediction model.For example,the demand prediction of emergency medical supplies can be translated into the prediction of the number of injured people in the disaster area,and a BP neural network prediction model is established with the six factors with greater influence weight as the input layer and the total number of injured people as the output layer,based on which a prediction model of the demand of emergency supplies considering the number of injured people,seasonal differences and regional differences is established to estimate the demand of emergency supplies at the disaster site.The demand of emergency supplies at the affected sites is estimated.Regarding the distribution of emergency relief materials,when the supply of emergency materials is insufficient at each disaster site in the early post-disaster period,relief materials should be delivered and distributed to the disaster sites from the surrounding non-earthquake regions in a timely manner.On the basis of the predicted demand for emergency supplies,the dual constraint objectives are the shortest total delivery time of supplies and the greatest satisfaction rate of the victims’ demand at each demand point,the degree of road damage has an impact on the total transportation time,and the road network repair time factor is added to the model to establish an emergency supplies distribution model in which multiple supplies are delivered and distributed to multiple disaster areas from multiple supply stockpiles,using a quadratic tournament The non-dominated ranking genetic algorithm(NSGA-II)with quadratic tournament strategy is used to solve the distribution model in order to arrive at an emergency relief material distribution scheme that takes into account the timeliness and demand satisfaction rate.Finally,the Pareto-optimal solution of the distribution scheme is obtained by solving the distribution model with the 2013 Ya’an earthquake as the background and the validity of the model is verified by case analysis.