文献类型：期刊文献

中文题名：基于AE-LSTM网络模型的机场周界入侵报警及分类算法

英文题名：Airport Perimeter Intrusion Alarm and Classification Algorithm Based on AE-LSTM Network Model

作者：许奕杰[1];王嵘[1];万永菁[1];孙静[2]

机构：[1]华东理工大学信息科学与工程学院,上海200237;[2]上海卓希智能科技有限公司研发部,上海201611

年份：2021

卷号：47

期号：3

起止页码：323

中文期刊名：华东理工大学学报（自然科学版）

外文期刊名：Journal of East China University of Science and Technology

收录：CSTPCD;;Scopus;北大核心:【北大核心2020】；CSCD:【CSCD_E2021_2022】；

基金：国家自然科学基金(61872143)。

语种：中文

中文关键词：机场周界;入侵报警系统;AE-LSTM;隐含编码特征

外文关键词：airport perimeter;intrusion alarm system;AE-LSTM;hidden encoder feature

摘要：针对传统的机场周界入侵报警系统存在的恶劣气象条件下误报率高、不能区分入侵类别等问题,提出了一种自编码长短时记忆(AE-LSTM)网络模型;提取输入信号的隐含编码特征,构建融合时序信息的特征向量矩阵,降低网络模型的复杂度。网络模型的性能评价结果表明,该模型的误报率低,振动状态分类准确率高,且复杂度低,具有很好的实际应用前景。
The airport perimeter intrusion alarm system is the first line of defense in the airport's flight zone.Traditional airport perimeter intrusion alarm system has higher false alarm rate and cannot classify different intrusion categories under the influence of severe weather.In order to solve the problem,a self-encoder long-short term memory network model is proposed in this paper.This model extracts the hidden encoder feature from input signals and constructs the feature vector matrix fused with timing information such that its complexity can be reduced.It is shown via the results of network model performance evaluation that the proposed network model has a low false alarm rate and a high accuracy of vibration states classification.Besides,this model has a low complexity that can guarantee a good practical application prospect.