摘要
针对现有的预测方法参数较多、精确度不高的问题,采用了时间序列挖掘的方法对合成旅未来一定时期内的装备维修保障能力进行预测。首先建立了指标体系,利用"装备云"平台相关数据对指标及装备维修保障能力随时间变化的序列进行计算;然后对多元时间序列进行线段化拟合、聚类、符号化表达、Apriori关联挖掘,通过差分整合移动平均自回归-支持向量回归组合模型及反向传播神经网络对合成旅装备维修保障能力进行预测,最后通过事例验证了本文所提出的方法。
In view of the problem that there are too many parameters in the existing prediction method and the accuracy is not high,the time series mining method is used to predict the equipment maintenance support capability of the synthetic brigade in a certain period of time in the future.Firstly,the index system is established and the data of the"equipment cloud"platform is used to calculate the sequence of indicators and equipment maintenance support ability.Then,the segmentation fitting,clustering,symbolic expression and Apriori association mining are applied to the multivariate time series.The autoregressive integrated moving average-support vector regression(ARIMA-SVR)model and back propagation(BP)neural network are used to predict the equipment maintenance support capability of the synthetic brigade.Finally,the proposed method is verified by an example.
作者
宋星
贾红丽
王谦
赵汝东
SONG Xing;JIA Hongli;WANG Qian;ZHAO Rudong(Department of Equipment Command and Management,Shijiazhuang Campus,Army Engineering University,Shijiazhuang 050003,China)
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2020年第4期878-886,共9页
Systems Engineering and Electronics
基金
军内科研计划项目（012016012600B12102）资助课题
关键词
时间序列挖掘
合成旅
装备维修保障
APRIORI算法
差分整合移动平均自回归
支持向量回归机
预测
time series mining
synthetic brigade
equipment maintenance support
Apriori algorithm
autoregressive integrated moving average(ARIMA)
support vector regression(SVR)
prediction
