摘要
针对舰船装备保障资源需求确定既有结构化规律可以遵循,又受非结构化不确定因素影响的特点,建立了备品备件需求的半结构化预测模型。该模型利用使用时间规模的线性函数表征备件需求的趋势项;用AR模型补偿时间规模统计误差及使用环境差异带来的影响;用GARCH模型刻画非结构因素导致预测误差的条件异方差特性。在此基础上,给出了利用半结构化模型预测保障资源需求的具体步骤。最后,对该方法进行了验证对比,结果表明:该舰船装备保障资源需求半结构化预测方法是有效、可行的。
Because of the support resources demand of warship equipment are influenced not only by some structural rules, but also by some non-structural factors, a semi-structural method of spares part prediction is proposed. The method uses the linear function of the equipment service time to de- scribe the requirement's trend term, auto-regressive to compensate the influence of service time bias and using environment differences and generalized ARCH to portray the heteroskedasticity characte- ristic due to non-structural factors. On this basis, the steps of how to use the semi-structural method to predict the support resources demand are given. Finally, a group of real data are used to verify and evaluate these methods, and the results prove the effectiveness and feasibility of the proposed predic- tion method.
出处
《海军工程大学学报》
CAS
北大核心
2015年第4期54-59,共6页
Journal of Naval University of Engineering
关键词
舰船装备
预测
半结构化
保障资源
广义自回归条件异方差
warship equipment
prediction
semi-structural
support resource
generalized auto-regressive conditional heteroskedasticity