基于反馈补偿K-means的救援物资需求预测

摘要

针对低空应急救援过程中救援物资需求的多样性与不确定性，提出基于反馈补偿K-means相似搜索的物资需求预测算法。通过比较非平稳数据处理方法，采用差分自回归平均移动模型（ ARIMA）对历史灾情数据进行平滑处理；将预处理后的灾情数据运用基于反馈补偿的K-means方法进行聚类分析；再对比夹角余弦，杰卡德相似系数以及相关系数这3种方法，搜索出相似度最大历史灾情案例，并线性求解当前低空应急救援所需物资量。实验结果表明，在基于反馈补偿K-means相似搜索的物资需求预测过程中，运用相关系数搜索的误差是最小的，方法不仅提高了大数据处理能力，而且在一定程度上提高了预测精度。%Aiming at sloving the diversity and uncertainty of the supplies demand in the process of aviation rescue,a prediction algorithm based on feedback compensation K -means similar search is proposed . First,smoothing the historical disaster data through ARIMA model;then,making a cluster analysis on the pretreated disaster data through feedback compensation K -means algorithm; finally, searching out the most similar case in historical disaster data by using correlation coefficient ,and obtaining the current sup-plies demand for aviation rescue through linear contrast .A mass of real seismic data was conducted to ver-ify the validity of the method , and the experimental results show that the forecasting method based on feedback compensation K-means can not only improves the ability of data processing ,but also improves the accuracy of prediction .