即時室內定位及跌倒偵測系統設計與實現

Design and Implementation of Real-time Indoor Positioning & Fall Detection System

陳泰翔(Tai-Siang Chen)

指導教授 ： 闕志達

國立臺灣大學/電機資訊學院/電子工程學研究所/碩士(2016年)

https://doi.org/10.6342/NTU201700158

摘要

台灣逐漸進入高齡化社會，老人居家照護一直都是被探討的議題，根據衛福部國民健康署，跌倒是老人事故傷害或死亡的主要因素之一，不只影響長輩的身心健康，也使照護者的負擔更加沉重。

本論文依據參考文獻所提出的居家照護系統(Home-Care System)進行改良，此系統可分為兩大部分:跌倒偵測及室內定位，當使用者穿戴的智慧型手錶偵測到跌倒行為時，系統會自動發出警報並顯示使用者所在位置。在室內定位部分，本論文使用接收機接收的訊號強度(Received Signal Strength Indicator)，通過感測器融合(Sensor Fusion)結合原有演算法可增加定位精準度。跌倒偵測部分則利用自動編碼器構成深層類神經網路來增加精準度，且利用多數決來防止單一模型失效。

關鍵字

類神經網路 ； 跌倒偵測 ； 智慧手錶 ； 室內定位 ； 軟體定義無線電 ； 感測器融合 ； 實時系統

並列摘要

Taiwan has gradually become an aging society. Home care for the elderly has always been the subject of discussion. According to Health Promotion Administration, Ministry of Health and Welfare, falling is the main reason for the elderly accident. It not only affects the physical and mental health of the elders, but also greatly increases the burden of caregivers.

In this thesis, we improve the previously developed system, which consists of two functions: indoor positioning and fall detection. When the user falls, the smart watch worn by the user will send a short message to the caregiver and show the user’s location. Moreover, we enhanced the indoor positioning system by using received signal strength indicator (RSSI) as well as the time difference of arrival (TDOA) of the RF signals received by the sensors to locate the user. By combining these sensor information by sensor fusion via extended Kalman filter, we obtained better positioning accuracy. For fall detection, we applied the auto encoder technique to build deep neural network (DNN) models and used majority decision to achieved better accuracy.

並列關鍵字

Neural Network ； Fall detection ； Smart watch ； Indoor Positioning ； Soft Defined Radio ； Sensor fusion ； Real-time System

參考文獻

[2] B.-C. Huang, “Development of an Indoor Positioning & Fall Detection System Based on Wearable Devices” Master Thesis, National Taiwan University, Jan. 2016.

[3] Z. Farid, R. Nordin, and M. Ismail, “Recent Advances in Wireless Indoor Localization Techniques and System,” Journal of Computer Networks and Communications, vol. 2013.

[4] X. Yu, "Approaches and principles of fall detection for elderly and patient," in e-health Networking, Applications and Services, 2008.

[5] J. Tsui, “Fundamentals of Global Positioning System Receivers - A Software Approach” 2nd, John Wiley & Sons, Inc., Nov. 2004.

[6] M.-Y. Tsai, “Design and Implementation of an Indoor Positioning System on SDR Platform” Master Thesis, National Taiwan University, Oct. 2014.