摘要
为解决单一模型预测装备故障率预测误差大、精度低的问题,提出了一种基于ARMA-BP组合模型的装备故障率预测方法。在建立ARMA模型和BP神经网络模型的基础上,采用加法集成法建立ARMA-BP组合预测模型,并利用方差倒数法确定ARMA模型和BP神经网络模型的权重系数。以某型装甲装备故障率数据为研究对象,对比ARMA模型、BP神经网络模型和ARMA-BP组合模型故障率预测结果,表明:相比于单一预测模型,ARMA-BP组合模型的装备故障率预测结果精度更高。
In order to solve the problem of large errors and low prediction accuracy existing in the equipment failure rate prediction based on single model,an equipment failure rate prediction method based on ARMA-BP combined model is proposed.Based on the establishment of ARMA model and BP neural network model,the ARMA-BP combined prediction model is established by the additive integration method,and the weight coefficient of ARMA model and BP neural network model is determined by the inverse method of variance.The failure rate data of a certain type of armored equipment is regarded as the research object,the failure rate prediction results of ARMA model,BP neural network model and ARMA-BP combined model are compared,the result shows that compared with the single prediction model,the equipment failure rate prediction result of ARMA-BP combined model is more accurate.
作者
徐达
周诚
关矗
王小闯
XU Da;ZHOU Cheng;GUAN Chu;WANG Xiao-chuang(Department of Arms and Control,Academy of Army Armored Force,Beijing 100072,China)
出处
《火力与指挥控制》
CSCD
北大核心
2021年第1期83-87,共5页
Fire Control & Command Control
基金
国防装备预研基金资助项目(41404010202)。
