Research On Several Key Technologies Of Emergency Assistance Smart Watch For The Elderly

As the increasing problems of an aging society, health problems in the elderly has attracted much attention. Therefore, monitoring daily exercise posture and other physiological signals such as heart rate of the elderly in real-time, assessing the health status of the elderly, has important social significance. Wearable device is small and easy to carry, and it can dynamically collect physiological signals at low physical and psychological pressure, it has great practical value of daily health care for the elderly, patients with rehabilitation and other aspects of. The combination of wearable device and the medical care, the health, will become important device about health guardianship in the new medical model.With the rapid development of wearable devices, several key technical problems need to be solved. The major research results summarized as follows:1.The improvement research of heart rate detection algorithm. At present, there are so many devices with the function of heart rate detection, but most of them are cumbersome, expensive and some even require professional operation, resulting in unable to monitor heart rate in a real-time. Smart watch, which is small in size, easy to carry, integrated multiple sensors, can monitor a variety of physiological signals such as heart rate and state of motion. When heart rate value exceeds the threshold, it will make an alarm initiative. Therefore, it is particularly necessary to research a heart rate detection algorithm with high accuracy. However, the research found that the heart rate detection algorithm mastered by several heart rate sensor suppliers, such as Philips, ADI and so on, lead to the costs of other wearable device industry increased. Based on the study of heart rate detection theory, research and improvement of pulse oximetry monitor’s heart rate detection algorithm, this paper is trying to find a heart rate detection with high accuracy and portability, which can be used on smart watch, in order to decrease the cost of the emergence assistant smart watch for the elderly. This paper presents a heart rate detection based on PPG signal. When detect trough, this paper use an algorithm which is similar with K-means algorithm, resulting in increasing of trough detection accuracy, and ensure the accuracy of heart rate detection algorithm. At last, by comparing experiment, we can prove that the algorithm with high accuracy and stability.2.Research on improvement of fall detection algorithm. The research shows that most current fall detection is based on an acceleration sensor fixed on the waist. The reason is that waist action is not obvious. Instead, wrist action is rich and complexity, there are few of fall detection on the wrist, even there are a few of smart watch with fall detection function on the market, its accuracy is low, and easy to miss. This paper summarized the previous fall detection technology and combined with wrist movement characteristics, using accelerometer gyroscopes and altimeters to detect fall. This paper presents a fall detection algorithm with SMVA, SMVW, and height of the threshold, in order to increase accuracy and decrease miss. To ensure the accuracy, we should judge the state of body who fell.3.Research on user classification methods. Smart watch can collect heart rate information, motion information, track information and height information, these information contains a huge commercial value. Analysis and processing information collected by smart watch, mining user’s features and interest from historical data and classify users, so as to provide a personalized service. This paper focuses on the user group classification preliminary research, in order to find a suitable method used for information which is collected by smart watch and provides a theoretical basis for user classification and providing personalized service.