摘要
针对动调陀螺仪传统维修方式存在维修不足和维修过剩等问题,设计了一种离线预测与健康管理系统;在分析动调陀螺仪状态影响参数的基础上,构建了状态监测平台,并采用时间序列方法对随机信号进行特征提取;利用采集的数据进行信息处理,采用神经网络进行健康状态评估,采用灰色支持向量机进行剩余寿命预测;最后,在对各种信息进行管理、利用的基础上,制定维修决策及进行保障资源管理;实际应用表明,本系统能有效提高动调陀螺仪的使用可用度,降低保障费用。
According to the problems of the insufficient or superfluous maintenance which existed by the traditional maintenance method for dynamically tuned gyroscope, designed a line prognostic and health management system. Based on analyzing the state parameters of dy- namically tuned gyroscope, constructed the condition monitoring platform, and used time series method to extract the feature of random sig- nal; Processed information based on the collected data, used neural network to assess the health status and utilized grey support vector ma- chine to prognostic the remaining life; At last, based on the management of various kinds of information, made the maintenance decision and managed the support resource. The actual application shows that the system can effectively improve the availability of dynamically tuned gy- roscope and reduce support costs.
出处
《计算机测量与控制》
北大核心
2014年第3期670-673,共4页
Computer Measurement &Control
基金
总装武器装备预研基金项目(9140A27020212JB14311)
关键词
动调陀螺仪
预测与健康管理
健康状态评估
剩余寿命预测
维修决策
trending structure time series data particle swarm optimization
immune genetic fuzzy--C means
clustering
