The more the scope of the earthquake disaster and the complexity of the emergency response become demanded, the more perfection of the risk identification, the precision of demand forecasting, and the effectiveness of the resources distribution in emergency material reserves and dispatch become demanded. However, existing methods still have limitations in areas such as insufficient collaborative modeling of multi-source data, imprecise characterization of demand changes driven by disasters, and weak linkage between reserve and dispatch decisions, making it difficult to meet the needs of rapid and precise support in high-risk areas. To address these issues, this paper proposes a tiered reserve and efficient dispatch method for earthquake emergency materials based on big data intelligent modeling algorithms. By integrating historical earthquake events, population and infrastructure distribution, and emergency material inventory and transportation data, a collaborative model system is constructed for disaster risk prediction, material demand forecasting, tiered reserve optimization, and dynamic dispatch. Machine learning models are used to predict disaster risks in gridded areas at different levels. Based on this, a time series model is introduced to characterize the changing characteristics of material demand in areas with different risk levels. Simultaneously, a multi-objective optimization model coordinates reserve costs, demand coverage, and response timeliness, and heuristic algorithms are used to dynamically optimize material allocation routes. Experimental results show that in disaster risk prediction, the identification results of high-risk and medium-risk areas are in good agreement with the actual disaster distribution. Regarding material demand prediction, the deviation between predicted and actual demand in typical high-risk areas is controlled within 2–3 tons, with a relatively small overall prediction error. These findings indicate that the suggested approach can play a major role in enhancing the accuracy of demand forecasting, optimization of reserve system, and dispatch optimization, which is an effective intelligent decision support implementation of earthquake emergency material management.
Liang et al. (Thu,) studied this question.