基于PSO-SVM-SST模型的地震应急物资需求预测研究

Earthquake Emergency Eupplies Demand Forecasting Based on PSO-SVM-SST Model

ES评分 0

DOI 10.14075/j.jgg.2025.01.011
刊名
Journal of Geodesy and Geodynamics
年,卷(期) 2026, 46(1)
作者
作者单位

1 防灾科技学院应急管理学院, 三河, 065000;
2 河北省资源环境灾变机理及风险监控重点实验室, 三河, 065000;
3 中国地震应急搜救中心, 北京, 100049

摘要
建立基于粒子群算法(PSO)优化的支持向量机(SVM)震后受灾人口预测模型,依据安全库存理论建立SST地震应急物资需求预测模型。选取地震危险性、破坏程度等9项指标参数,经降维和去冗处理后作为基于PSO优化的SVM模型输入变量,并开展受灾人数预测,根据受灾人口与应急物资间的内在关联,应用SST模型对九寨沟地震震后初期所需的典型物资数量进行间接估算。结果表明,通过采用误差对比分析方法对模型进行有效性验证,PSO-SVM模型较SVM模型的预测误差降低14.27%, 预测精度显著提高。估算得到九寨沟地震震后典型物资需求量,预测结果具有一定的参考价值,表明PSO-SVM-SST预测模型在理论和实践层面均具有一定的合理性和实用性。
Abstract
A post-earthquake affected population prediction model based on support vector machines(SVM) optimized by particle swarm optimization(PSO) is established, and the SST(safety stock theory) earthquake emergency supply demand prediction model is constructed. Nine indicator parameters, including seismic hazard and damage severity, are selected and processed through dimensionality reduction and redundancy removal as input variables for the PSO-optimized SVM model to predict the affected population. Based on the relationship between the affected population and emergency supplies in disaster areas, the SST model is applied to indirectly estimate the quantities of typical supplies required in the immediate aftermath of the Jiuzhaigou earthquake. The experimental results are as follows: By employing an error comparison analysis method to validate the model's effectiveness, the PSO-SVM model demonstrates a 14.27% reduction in prediction error compared to the SVM model, with a significant improvement in prediction accuracy. The estimated demand for typical supplies in the aftermath of the Jiuzhaigou earthquake provides a certain degree of reference, indicating that the PSO-SVM-SST prediction model possesses both theoretical and practical rationality and utility.
关键词
地震应急物资; 需求预测; 支持向量机; 安全库存理论
KeyWord
earthquake emergency supplies; demand forecasting; support vector machine(SVM); safety stock theory
基金项目
页码 86-93
  • 参考文献
  • 相关文献
  • 引用本文

唐彦东, 程梅, 刘军, 于汐, 林浩. 基于PSO-SVM-SST模型的地震应急物资需求预测研究 [J]. 大地测量与地球动力学. 2026; 46; (1). 86 - 93.

  • 文献评论

相关学者

相关机构