摘要
随着智能家居环境下传感数据的不断增长,服务器采用集中式分析存储无法满足数据处理需求,此外,数据传输过程中网络负载量的增加,会导致较长的网络延迟。因此,在Zigbee节点、智能网关、远程服务器具有一定计算存储能力的基础上,提出一种面向智能家居老人看护系统的设计方案。采用Zigbee技术、异常值处理、特征提取实时地获取传感序列,基于Zigbee开源协议栈Z-Stack 2007和智能网关实现传感数据的采集和异构网络的互联互通;利用固定时间滑动窗口法和隐式马尔可夫模型预测老人基本日常行为ADLs,并提供历史记录查询和远程监控功能。测试结果表明,该方案能有效地减少通信开销,降低传输时延,实时地监控预测老人行为动作且平均准确率达92%,实现了智能家居环境下老人的看护。
With the growth of sensor data in smart home,the server using centralized analysis and storage cannot meet the needs of data processing. In addition,the increase of network load during data transmission will lead to longer network latency. Therefore,considering that Zigbee node,intelligent gateway and remote server have storage and computing capacity,we propose a scheme for elderly care system in smart home,using Zigbee technology,exception handling,feature extraction to obtain real time sensor sequences. Based on Z-Stack 2007 and intelligent gateway the sensor data is collected and the heterogeneous network interconnection is achieved. The fixed-time sliding window method and hidden Markov model are used to predict the activities of daily living of the elderly, and the historical query and remote monitoring is provided. The test shows that the proposed scheme can effectively reduce communication overhead,reduce transmission delay,monitor and predict the activity of elderly people in real time. The average accuracy rate of activity recognition is 92%,which can realize the elderly care in the smart home.
作者
毛驾燕
沈苏彬
MAO Jia-yan;SHEN Su-bin(School of Computer,Nanjing University of Posts and Telecommunications,Nanjing 210003,China)
出处
《计算机技术与发展》
2018年第8期164-169,共6页
Computer Technology and Development
基金
国家自然科学基金(61502246)
南京邮电大学科研启动基金项目(NY215019)
未来网络前瞻性研究项目(BY2013095-1-08)
