Fall detection device based on mobile communication

Abstract

The invention discloses a fall detection device based on mobile communication, which comprises a body-worn sensor and a cloud server, and is characterized in that the body-worn sensor comprises a battery power supply module, an acceleration sensor, a low power consumption processor, an NB-IoT communication circuit and an alarm button; the body-worn sensor sends falling information to the cloud server through the NB-IoT communication circuit, and the cloud server informs a corresponding guardian through a 4G LTE communication mode. The wearable sensor can be carried to send tumble information to the cloud server in an NB-IoT communication mode, and finally, a guardian is informed of an abnormal condition in time, so that the accidental tumble condition of the old people is found in time, alarming and positioning are performed, a guarantee is provided for timely treatment of an accident happening person, and the safety of the old people is ensured. An effective technical guarantee is provided for the safety problem of old people living alone at present, use is easy, and cost is low.

Description

Technical Field

[0001] The present invention belongs to the technical field of communication and computer application, and in particular relates to a fall detection device based on mobile communication. Background Art

[0002] With the acceleration of the aging of society, accident monitoring and emergency rescue of the elderly are becoming more and more important. Falling is an important cause of disability in the elderly, and it is also a non-specific manifestation of the decline of body function and various acute and chronic diseases in the elderly. Timely rescue of the elderly who fall will greatly reduce the disability rate and mortality rate.

[0003] Falling is when a person loses balance while in an active state and falls unconsciously to the ground or some lower plane, and the person who falls has no way to react in time. Due to different postures when falling, the ways of falling vary greatly. At present, in the research field of human behavior recognition, the more common fall detection technologies include visual fall detection technology, audio vibration fall detection technology, and inertial sensor fall detection technology. These three technical methods have their own advantages and disadvantages. Although visual fall detection technology is easy to configure and has high judgment accuracy, it is limited by ambient light, spatial area, etc., and involves privacy issues and is costly. Audio vibration fall detection technology has poor practicality, many interference sources, low accuracy, high cost, and cannot be used outdoors. Although inertial sensor fall detection technology has low cost, wide coverage, and is easy to use, its accuracy is low and the energy consumption and charging frequency are high. Summary of the invention

[0004] In order to make up for the shortcomings of the prior art, the present invention aims to provide a fall detection device based on mobile communication, which monitors whether the elderly user carrying the body-worn sensor falls in real time, and sends the fall information to the cloud server through the NB-IoT communication method, and finally informs the guardian of the abnormal situation in time, so as to achieve timely rescue of the elderly who fall. To achieve this purpose, the specific technical scheme of the present invention is as follows:

[0005] The mobile communication-based fall detection device includes a body-worn sensor and a cloud server, wherein the body-worn sensor includes:

[0006] A battery powered module for providing electrical energy;

[0007] The acceleration sensor uses the MPU6050 acceleration sensor to collect three-axis acceleration and three-axis angular velocity in real time;

[0008] Low-power processor, using the STM32L151 chip;

[0009] NB-IoT communication circuit, using LTE BC20 NB-IoT / GNSS module for narrowband IoT communication and GNSS positioning functions, while uploading GPS location information of fall events;

[0010] The alarm button is used to actively initiate a call for help. When an abnormal situation occurs, you can also actively call for help;

[0011] When a fall event is detected, the body-worn sensor sends the fall information to the cloud server through the NB-IoT communication circuit, and the cloud server informs the corresponding guardian through 4G LTE communication.

[0012] Furthermore, the body-worn sensor uses the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a fall occurs. Through the changes in the SVM value, variation index value CV, pitch angle pitch and roll angle roll during the fall process, a pre-alarm is first issued when abnormal data occurs, and continuous sampling is performed in the following period of time. If it meets the movement characteristics of a person when falling, an alarm operation is performed.

[0013] Furthermore, the body-worn sensor can read the three-axis acceleration values ​​aacx, aacy, aacz and the three-axis gyroscope values ​​gyrox, gyroy, gyroz, and can calculate the Euler angles pitch, roll, and yaw through its internally integrated DMP. The acceleration signal vector SVM is used to represent the human body motion state, and its expression is:

[0014]

[0015] To reduce the error, the SVM value is calculated using the mean of five consecutive sampling acceleration values.

[0016] Furthermore, according to the change of SVM value, the human body motion state is analyzed:

[0017] Stable upright state: SVM varies between 0.9g and 1.5g, and pitch and roll vary between -25° and 25°;

[0018] Shaking unstable state: SVM keeps changing suddenly, and pitch or roll changes greatly in a short time;

[0019] Free fall in the air: The time is about 0.3s to 0.4s, the SVM decreases significantly, and the pitch or roll changes continuously;

[0020] At the moment of touchdown: SVM increases sharply and then quickly returns to normal value. Pitch and roll stop changing at this time.

[0021] Stable falling state: SVM, pitch and roll all stop changing.

[0022] Furthermore, the variation index CV is related to the angular velocity ω, where CV is defined by the following formula:

[0023]

[0024] In the formula, x i is the value of SVM, and the variation index CV is the ratio of the mean value of SVM to the standard deviation, which reflects the degree of dispersion of SVM. When the human body is in normal activity, the CV value is small; when the person is exercising vigorously or falling, the CV value will increase; and the angular velocity ω is defined by the following formula:

[0025]

[0026] The above formula represents the speed of change of pitch angle and roll angle during human movement. The sampling frequency is 20 Hz. When both SVM and CV exceed the threshold and ω continues to change, it is judged that a fall has occurred.

[0027] Furthermore, the body-worn sensor uses the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a fall occurs, and the specific contents are as follows:

[0028] Normal working state: sampling frequency is 20Hz, and CV value is updated every 1 second;

[0029] Pre-alarm stage: when SVM is less than 0.9g or ω exceeds 40, a pre-alarm is issued;

[0030] Detection phase: In the next 0.35s, the sampling frequency is 100Hz, and the SVM value is obtained by five average values. There are 7 SVM values ​​in total. If the following four conditions are met at the same time, and the SVM will surge to 3g in a very short time and then return to normal, and ω stops changing, an alarm will be triggered, which meets the movement characteristics of a fall. The four conditions are: 1) All SVM sampling points are less than 0.9g; 2) The SVM mean is less than 0.7g; 3) ω always exceeds the threshold; 4) The CV value is greater than 0.2 and the CV value in the previous second is less than 0.2.

[0031] Compared with the prior art, the present invention has the following advantages:

[0032] The portable body-worn sensor of the present invention sends the fall information to the cloud server through the NB-IoT communication mode, and finally informs the guardian of the abnormal situation in time, aiming to timely detect the accidental fall of the elderly and alarm and locate them, so as to provide guarantee for timely treatment of the accidental fall. This device provides an effective technical guarantee for the safety of the elderly living alone at present, and it is simple to use, low cost, and has certain social significance and practical value. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a module diagram of a fall detection device of the present invention;

[0034] Figure 2 A human body coordinate diagram involved in the fall detection algorithm of the present invention;

[0035] Figure 3 The figure is a flow chart of the fall detection method of the body-worn sensor of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0037] In the social context of an aging population and a rapid increase in the number of empty-nest elderly people, this patent proposes an accidental fall detection device based on wireless communication and positioning to timely detect the accidental fall of the elderly and provide alarm and positioning, so as to provide timely treatment for the accident victims.

[0038] like Figure 1 As shown, a fall detection device based on mobile communication includes a body-worn sensor and a cloud server. The body-worn sensor includes a battery-powered module, an acceleration sensor, a low-power processor, an NB-IoT communication circuit, and an alarm button. When a fall event is detected, the body-worn sensor sends the fall information to the cloud server through the NB-IoT communication circuit, and the cloud server informs the corresponding guardian through the 4G LTE communication method. The battery-powered module is used to provide power; the acceleration sensor adopts the MPU6050 acceleration sensor to collect three-axis acceleration and three-axis angular velocity in real time; the low-power processor adopts the STM32L151 chip, which has strong processing power and low power consumption; the NB-IoT communication circuit adopts the LTE BC20 NB-IoT / GNSS module for narrowband Internet of Things communication and GNSS positioning functions, and can simultaneously upload the GPS location information of the fall event; the alarm button is used to actively initiate a call for help, and in the event of an abnormal situation, it can also actively send a call for help.

[0039] Specifically, the fall detection method of body-worn sensors mainly analyzes the movement state of the human body throughout the fall process, and uses the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a fall has occurred. Through the changes in the SVM value, variation index value CV, pitch angle pitch and roll angle roll during the fall process, a pre-alarm is first issued when abnormal data occurs, and continuous sampling is carried out in the following period of time. If it meets the movement characteristics of a person when falling, an alarm operation is performed.

[0040] like Figure 2 As shown in the figure, the body-worn sensor can read the three-axis acceleration values ​​aacx, aacy, aacz and the three-axis gyroscope values ​​gyrox, gyroy, gyroz. The Euler angles pitch, roll, and yaw can be calculated through its internal integrated DMP. The acceleration signal vector SVM is used to represent the human body motion state, and its expression is:

[0041]

[0042] To reduce the error, the SVM value is calculated using the mean of five consecutive sampling acceleration values.

[0043] The process of falling usually involves falling forward, backward or sideways, so the pitch angle or roll angle will change greatly. At the same time, due to the change in the state of motion, the total acceleration SVM of the human body will also change. The entire fall is divided into five states, and the change of SVM, pitch and roll in these five states is analyzed to determine whether a fall has occurred. Since the yaw angle will not change much during the entire fall process, and the yaw angle may also change greatly in some daily activities, this patent does not consider the influence of the yaw angle.

[0044] According to the changes in SVM, pitch and roll, the human body motion state (i.e. the five states of the fall process) is analyzed:

[0045] (1) Stable upright state: SVM varies between 0.9g and 1.5g, and pitch and roll vary between -25° and 25°.

[0046] (2) Shaking unstable state: SVM keeps changing suddenly, and the pitch or roll changes greatly in a short period of time.

[0047] (3) Free fall in the air: The time is about 0.3s to 0.4s, the SVM decreases significantly, and the pitch or roll continues to change.

[0048] (4) Touchdown moment: SVM increases sharply and then quickly returns to normal value. Pitch and roll stop changing at this time.

[0049] (5) Stable falling state: SVM, pitch and roll all stop changing.

[0050] Among them, the variation index CV is related to the angular velocity ω, where CV is defined by the following formula:

[0051]

[0052] In the formula, xi is the value of SVM, and the variation index CV is the ratio of the mean value of SVM to the standard deviation, which reflects the degree of dispersion of SVM. When the human body is in normal activity, the CV value is small; when the person is exercising vigorously or falling, the CV value will increase; and the angular velocity ω is defined by the following formula:

[0053]

[0054] The above formula represents the speed of change of pitch angle and roll angle during human movement. Since the frequency of daily human activities does not exceed 20 Hz, the sampling frequency is taken as 20 Hz. When both SVM and CV exceed the threshold and ω continues to change, it is judged that a fall has occurred.

[0055] like Figure 3 As shown, based on the analysis of the falling process, the body-worn sensor uses the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a fall occurs. The specific contents are as follows:

[0056] When the body-worn sensor is in normal working state: the sampling frequency is 20Hz, and the CV value is updated every 1 second;

[0057] When the body-worn sensor is in the pre-alarm stage: when the SVM is less than 0.9g or ω exceeds 40, a pre-alarm is issued;

[0058] When the body-worn sensor is in the detection stage: in the next 0.35s, the sampling frequency is 100Hz, and the SVM value is obtained by five average values. There are 7 SVM values ​​in total. If the following four conditions are met at the same time, and the SVM will surge to 3g in a very short time and then return to normal, and ω stops changing, an alarm will be triggered, which meets the movement characteristics of a fall. The four conditions are: 1) All SVM sampling points are less than 0.9g; 2) The SVM mean is less than 0.7g; 3) ω always exceeds the threshold; 4) The CV value is greater than 0.2 and the CV value in the previous second is less than 0.2.

[0059] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A fall detection device based on mobile communication, characterized in that: The invention comprises a body-worn sensor and a cloud server, wherein the body-worn sensor comprises: A battery powered module for providing electrical energy; The acceleration sensor uses the MPU6050 acceleration sensor to collect three-axis acceleration and three-axis angular velocity in real time; Low-power processor, using the STM32L151 chip; NB-IoT communication circuit, using LTE BC20 NB-IoT / GNSS module for narrowband IoT communication and GNSS positioning functions, while uploading GPS location information of fall events; The alarm button is used to actively initiate a call for help. When an abnormal situation occurs, you can also actively call for help; When a fall event is detected, the body-worn sensor sends the fall information to the cloud server through the NB-IoT communication circuit, and the cloud server informs the corresponding guardian through 4G LTE communication.

2. A fall detection device based on mobile communication according to claim 1, characterized in that: The body-worn sensor uses the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a fall occurs. Through the changes in the SVM value, variation index value CV, pitch angle pitch and roll angle roll during the fall process, a pre-alarm is first issued when abnormal data appears, and continuous sampling is performed in the following period of time. If it meets the movement characteristics of a person when falling, an alarm operation is performed.

3. A fall detection device based on mobile communication according to claim 2, characterized in that: The body-worn sensor can read the three-axis acceleration values ​​aacx, aacy, aacz and the three-axis gyroscope values ​​gyrox, gyroy, gyroz, and can calculate the Euler angles pitch, roll, and yaw through its internal integrated DMP. The acceleration signal vector SVM is used to represent the human body motion state, and its expression is: To reduce the error, the SVM value is calculated using the mean of five consecutive sampling acceleration values.

4. A fall detection device based on mobile communication according to claim 3, characterized in that: According to the changes in the SVM value, the human body movement state is analyzed: Stable upright state: SVM varies between 0.9g and 1.5g, and pitch and roll vary between -25° and 25°; Shaking unstable state: SVM keeps changing suddenly, and pitch or roll changes greatly in a short time; Free fall in the air: The time is about 0.3s to 0.4s, the SVM decreases significantly, and the pitch or roll changes continuously; At the moment of touchdown: SVM increases sharply and then quickly returns to normal value. Pitch and roll stop changing at this time. Stable falling state: SVM, pitch and roll all stop changing.

5. A fall detection device based on mobile communication according to claim 4, characterized in that: The relationship between the variation index CV and the angular velocity ω, where CV is defined by the following formula: In the formula, x i is the value of SVM, and the variation index CV is the ratio of the mean value of SVM to the standard deviation, which reflects the degree of dispersion of SVM. When the human body is in normal activity, the CV value is small; when the person is exercising vigorously or falling, the CV value will increase; and the angular velocity ω is defined by the following formula: The above formula represents the speed of change of pitch angle and roll angle during human movement. The sampling frequency is 20 Hz. When both SVM and CV exceed the threshold and ω continues to change, it is judged that a fall has occurred.

6. A mobile communication-based fall detection device according to claim 5, characterized in that: The specific contents of the body-worn sensor using the acceleration signal vector SVM threshold method and the Euler angle auxiliary judgment method to detect whether a person falls are as follows: Normal working state: sampling frequency is 20Hz, and CV value is updated every 1 second; Pre-alarm stage: when SVM is less than 0.9g or ω exceeds 40, a pre-alarm is issued; Detection phase: In the next 0.35s, the sampling frequency is 100Hz, and the SVM value is obtained by five average values. There are 7 SVM values ​​in total. If the following four conditions are met at the same time, and the SVM will surge to 3g in a very short time and then return to normal, and ω stops changing, an alarm will be triggered, which meets the movement characteristics of a fall. The four conditions are: 1) All SVM sampling points are less than 0.9g; 2) The SVM mean is less than 0.7g; 3) ω always exceeds the threshold; 4) The CV value is greater than 0.2 and the CV value in the previous second is less than 0.2.

Images