Research On The Decision Process Of The Distribution Of Emergency Supplies After Sudden Disaster Events

Since the sudden disaster events happened frequently in China and worldwide caused great damage to the socio-economic, the research on emergency management and the development of decision support systems have become an urgent need to deal with the increasingly severe sudden disaster events. Because the distribution of emergency supplies is the key problem for the relief to unexpected disaster events, emergency supplies distribution decision, as an important branch of the emergency management and emergency logistics management, has become a hot issue in academic research at home and abroad.Firstly, the paper proposed the research thought of emergency supplies optimization distribution decision based on the whole process of the decision in order to provide effective protection for the demand of the emergency supplies based on the analysis of the current research status on emergency decision-making and emergency supplies management in domestic and foreign. Secondly, the paper explained the connotation of emergency supplies distribution decision, analyzed its characteristics and the decision-making process. Lastly and importantly, the paper provided emergency supplies prediction model, emergency supplies distribution models and evaluation model for the decision scheme and their algorithms to the emergency decision-making bodies and staff. The main research content and innovation in this paper are as follows.(1)Analysis of the characteristics of the emergency supplies distribution decision. Through studying the connotation and the elements of emergency supplies distribution decision system, the characteristics of emergency supplies distribution decision was analyzed from the angle of system elements and the emergency supplies distribution decision-making process was designed. It was believed that emergency supplies distribution decision was a dynamic decision-making process, and the emergency supplies distribution decision-making system was a circulatory system composed of multiple stages. Every decision cycle should include complete four stages, namely, information collection, distribution decision, program evaluation, plan implementation.(2)Research on the emergency supplies demand forecast model. Based on the different points of supplies demand forecast methods between the emergency logistics situation and general logistics situation, a indirect prediction method was provided to forecast the emergency supplies demand in a gold rescue time after sudden disaster events occurred, namely to forecast the population casualties first, then the emergency supplies demand. Four steps was designed to predict the demand, including qualitative analysis and quantitative analysis of factors associated with injuries population, a BP neural network model to predict the casualty population and the prediction of the emergency supplies demand. According to the four steps, specific models were put forward and described as an example of the emergency supplies demand forecast in a large earthquake.(3)Research on the optimal emergency supplies distribution model including two stages. The first stage is the building of the model based on two-layer node networks and the game theory. A strategy of so-called incomplete put-out was proposed when relief supplies could not meet the emergency needs within a short time. A non-cooperative game model based on complete information was presented, in which the affected points corresponded to the players, and the distribution schemes to the strategies. A phased planning approach was used to reduce the number of the strategies caused by too many nodes and distribution supplies. The approach can realize the initial distribution of aiming at shorting the response time, and the second planning of establishing a game model for the conflicting nodes after the initial distribution. The Nash equilibrium of the model was found by using particle swarm optimization algorithm through constructing a fitness function. A numerical analysis was conducted to test the effectiveness of the model. The results show that the model does well in distributing the emergency supplies efficiently and fairly when imbalance between supply and demand occurs, and reaches a better rescue effect. The second stage is the building of the distribution model of emergency supplies with three-layer node networks existing in the practice of emergency management. The model aims at minimizing the system losses and meets the constraints of emergency response time and fairness constraint. A modified particle swarm optimization algorithm was proposed for the integer nonlinear programming, the spatial search abilities of the particles were improved by determining different learning objects for different dimensions of the particles. Finally a numerical example was conducted to test the effectiveness of the model and the algorithm. (4)Research on the evaluation model of emergency supplies distribution scheme. Firstly, the principle on the emergency supplies distribution scheme evaluation should follow was analyzed; it is believed that the principle of fairness should be emphasized when evaluating the emergency supplies distribution scheme after the analysis of the dialectical relationship between impartiality and efficiency; Secondly, the meaning of the fair and the equitable distribution, the connotation of the emergency supplies fair distribution was analyzed; Finally, a simple, easy to understand fair measure model was proposed which took into account the different requirements and urgency degree for supplies. The model was applied to the evaluation of the distribution scheme in the example with three-layer node networks.The above research can provide scientific basis for emergency supplies distribution decision-making, thereby enhance the operational efficiency of the emergency logistics and emergency supplies support system, and reduce casualties and property losses caused by disasters.