基于CNN-LSTM的独居老人室内行动轨迹监测模型研究  

Research on Indoor Activity Trajectory Monitoring Model for Elderly Living Alone Based on CNN-LSTM

作　　者：周磊 Zhou Lei(Zhejiang Business Technology Institute,Ningbo 315012,China)

机构地区：[1]浙江工商职业技术学院,宁波315012

出　　处：《办公自动化》2025年第7期103-105,共3页Office Informatization

基　　金：浙江工商职业技术学院2024年校级科研资助项目“基于CNN-LSTM神经网络的独居老人室内行动轨迹跟踪与监测模型研究”(项目编号:KYND202416);浙江工商职业技术学院电子信息分院2024年基本科研资助项目“基于BERT模型的中学生职业规划策略探究”。

摘　　要：提出一种新型的神经网络模型,结合卷积神经网络(CNN)和长短期记忆网络(LSTM),旨在长期跟踪与预测独居老人的健康状态。模型构建中,CNN负责从每日行动轨迹图中提取行为习惯和健康相关的时空特征,LSTM则基于这些特征序列学习老人的正常行为模式,处理时间序列数据的长期依赖。该模型不仅能识别即时异常,还能预判潜在的健康问题,如阿尔茨海默症。最后,通过定义异常行为标签并编码为监督学习目标,模型实现对老人异常状态的有效预测,为独居老人的健康监护提供更为全面和前瞻性的解决方案。This paper presents an innovative neural network model that integrates Convolutional Neural Networks(CNN)and Long Short-Term Memory networks(LSTM)to monitor and predict the health status of elderly individuals living alone over time.In the model architecture,CNN extracts spatiotemporal features related to behavioral patterns and health from daily activity trajectory maps,while LSTM uses these feature sequences to learn the normal behavioral patterns of the elderly,addressing long-term dependencies in time series data.This dual approach not only identifies immediate anomalies but also forecasts potential health issues,such as Alzheimer's disease.By defining abnormal behavior tags and encoding them as supervised learning objectives,the model achieves effective prediction of abnormal states among the elderly,providing a more comprehensive and proactive solution for their healthcare monitoring.

关 键 词：CNN-LSTM模型 轨迹图分析 异常状态识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术]