Key Optimization Problems In Emergency Blood Supply Under Unconventional Emergencies

The blood supply is an important section to save lives and protect health. The past emergency management practice have put forward the urgent request to build and perfect emergency blood supply system of China. So, to study problems in emergency blood supply system optimization during unconventional emergencies has practical significance.Four kinds of key optimization problems in unconventional emergencies were studied in this research:the location-inventory problem of national blood strategic reserves, dynamic decision problem of emergency Blood collection for unconventional emergencies, optimization of emergency Blood collection and storage problem for unconventional emergencies considering substitution of blood type and supply constraint, reliability location problem of new blood banks in post-disaster reconstruction areas. These four kinds of problems were studied aiming to improve the emergency blood supply level at different stages based on different scenarios.Main research contents in the paper are as follows:(1) Based on the the characteristics of Large scale disasters involving blood supply,taking the highest timeliness of blood supply under unconventional emergencies, and the total system cost under conventional condition as goals, a joint multi-objective LIP model with lead-time was built, considering multi-cenarios,multi-stages,multi-blood types, stochastic demands, facility capacity constraints and coordinate location. The goal is to minimize system cost and maximize system timeliness. A discrete nonlinear mixed integer programming model with 2 goals is built to describe the problem. An improved NSGAII is worked out to solve the model. Numerical example indicated that the pateto front solution set can be obtained. Optimal decision schemes can be selected from a cluster of pateto solutions according to the preferences and actual needs of decision makers.(2) According to the characteristics of emergency blood inventory control problem in unconventional emergencies, a dynamic decision model for emergency blood collection was developed. And the transfer equations of inventory state were derived. The expressions of outdating, shortage and collection capacity constraint of blood were formulated. The forecasting method of blood demand was also introduced. A numerical example was presented to illustrate the validity of the model. And the sensitive analysis of model parameters was made. The results show that the quantity of scrapped blood can be reduced evidently by using the proposed model, and the setting of safety stock level and target stock level, prediction error, and demand fluctuating range all have a significant impact on inventory control effect.(3) In reaction to the situation of blood supply constraint but demand identical to conventional scene for some unconventional emergencies, an emergency blood collection and storage problem considering substitution of blood type was studied. At first,the substitution effect on inventory control was analyzed. Then,an expected cost minimization model stochastic demands, service level, the inventory relationship between the blood center and hospitals, A discrete nonlinear mixed integer programming model was built to describe the problem. According to the characteristics of model, an accurate algorithm was desired.The calculation results showed that: the introduction of blood substitution strategy can reduce the blood shortage and expected system cost, and can help to ease the blood shortage,improve emergency blood support ability.(4) Locating relief supplies reserve bases is a strategic decision problem. It is necessary to consider the risk of facility failure in design stage because reserve bases may fail as a result of public emergencies. Taking the timeliness and reliability of material support as goals, a reliability P-median location model for relief supplies reserve bases was developed by considering different probability of facility failure in different regions. According to the characteristics of the model, a linearization technique was applied to convert the model and a Lagrangian relaxation(LR) algorithm was proposed to solve the problem. Finally, two groups of numerical examples were given to test and verify the model and algorithm. The performance of LR algorithm was compared with CPLEX. The results show that CPLEX is more efficient for small or middle-scale problems. And for large-scale problems, LR algorithm is more efficient.