Demand Forecast And Gradation For Demand Urgency Of Emergency Materials Based On Intelligence Algorithm

Earthquake disasters occurred frequently in recent years,due to the nature of it is difficult to forecast in advance,to respond to disasters of rapid becomes the most effective means of reduce casualties and losses.Reasonable emergency supplies is a basic work of emergency response,and demand forecast and gradation for demand urgency of emergency materials is the rationalization of emergency supplies one of the most main work.In emergency supplies forecast,due to the complexity of emergency data sources,statistical difference between the data is not accurate and less,the levels of data in case very greatly,and historical cases is old.Caused the traditional data processing method is difficult to find emergency data hidden information,supplies to predict when the subjective factors influencing the choice,and supplies forecast method is easily influenced by the old case,and a great deal of research about the demand urgency of emergency materials is limited by the urgency of place.According to the characteristics of the emergency work and emergency data,demand forecast and gradation for demand urgency of emergency materials were given in this paper.The first part was the important influence factors of the number of casualties be got in the emergency data mining.The second part was demand forecast of emergency materials by the method with two-step prediction combined with the number of casualties and safety stock theory.The third part gave the demand urgency level of emergency materials combined with the casualties of different scale.In view of the above,the research route of this paper will be carried out from the following aspects:(1)In the part of emergency data mining,firstly,in view of the fact that there are few researches of influencing,the influencing factors were expanded and discussed.And the influencing factors that can be easily obtained in the first time of earthquake are given.Secondly,according to the characteristics of emergency data,an improved method of discretization was proposed to discretize the training data and successfully separate the magnitude of the hidden data in the emergency data.Finally based on the of problem influencing factors subjective selection,this paper puts forward an improved algorithm of attribute dependency of rough set to construct data mining model,it was concluded that the importance of the influence factors of the earthquake casualties,to dig out the important influence factors.Compared with the traditional rough set attribute reduction algorithm was verified in this paper,proved that the method has certain rationality and superiority.(2)In the part of demand urgency of emergency materials,firstly,a method be proposed by to predict the scale and then calculate the casualty range in reverse based on the traditionaltwo-step method by consider that small data statistics and the large number of emergencies after the earthquake and the final demand will also change.Second use of selected factors as input layer,scale of casualties as the output layer,aiming at the effects of old case for training,a neural network based on the metabolism of the RBF neural network to predict the scale of casualties was built.And compared with the BP neural network and RBF neural network and support vector machine(SVM),found that the accuracy of the improved RBF neural network model was higher.Finally,combining with casualty range and the theory of safety inventory,the formula of material prediction was gave,and then predicted the common emergency materials and epidemic prevention emergency materials.(3)In the part of demand urgency of emergency materials.Firstly,the urgency of materials was classified based on the ambiguity synthetic judgment,and then materials are classified based on the modularization of materials.Finally,the demand urgency of emergency materials was calculated,and the classification results of material grouping and non-grouping under different casualty scales are analyzed,and the rationality is verified by comparing with some examples.To some extent,the research results of this paper can provide reference for the emergency materials research and scientific basis for improving the rescue capability.