基于Transformer-Bi-LSTM模型的武器装备剩余寿命预测方法
Method for Predicting the Residual Life of Weaponry based on Transformer Bi-LSTM Modell
袁玉昕 1程跃兵 1熊敏艳 1高王升 1张昱彤1
作者信息
- 1. 上海机电工程研究所,上海 201109
- 折叠
摘要
武器装备担负保卫国土安全的重要使命,其保持稳定运行状态具有重大国防、政治意义;因其装备运行状态不便中断、故障定位过程复杂,使得传统维修方式效率较低;装备使用数据具有连续性、长期性、不平稳性,甚至一些深度学习模型无法处理其中的退化状态历史依赖与关联问题;通过构建元器件层级的剩余寿命预测架构,对特征工程、退化指标构建以及Trans-former-Bi-LSTM模型开展研究,采用距离编码技术,实现针对深度学习模型的技术创新,优化模型预测效果;基于某型武器装备主要器件正常试样数据,进行本方法分析验证,在器件已运行时间达到90％设计试验寿命长度时能够进行有效且准确的剩余寿命预测,所提方法满足武器装备器件寿命预警及更换提醒、保障装备战备完好性的应用需求.
Abstract
Weaponry is responsible for the important mission of safeguarding national security,and its stable operation is of great nation defense and political significance.Due to the inconvenient interruption of the equipment operation status and the complex fault location process,it results in lower efficiency of traditional maintenance methods.The equipment usage data has the characteristics of continuity,long-term,and instability,and some deep learning models cannot deal with the historical dependence and association of the degraded states.The remaining life prediction architecture at the component level is built to study the feature engineering,degradation index construction and Transformer bi-directional long short-term memory(Bi-LSTM)model.The distance coding is used to realize the technological innovation of the deep learning model and optimize the prediction effect of the model.Based on normal sample data of primary components of certain weapons and equipment,this method analyzes and validates the remaining life of the device,it can be effectively and accurately predicted in operation with 90％of the designed test life.The proposed method meets the application re-quirements of early warning and replacement for weaponry components,ensuring weaponry readiness.
关键词
武器装备/寿命预测/健康管理/Transformer/Bi-LSTM/退化指标/距离编码Key words
weaponry/residual life prediction/health management/Transformer/Bi-LSTM/degradation indicators/distance coding引用本文复制引用
出版年
2024
NETL
NSTL
万方数据