A real-time monitoring system for falling status and anti-lost position information

Abstract

The present invention relates to a real-time monitoring system for falling status and anti-lost position information, which at least includes a mobile terminal and a processing unit. A first sensor, a second sensor and a third sensor arranged on the mobile terminal and the surrounding ground respectively detect the user's movement information, physiological information and pressure information. The mobile terminal divides the user's current state into at least a static state, a moving state, a pre-fall state and a post-fall state according to the warning level obtained from the processing unit and based on the processing of the user's movement information, physiological information and pressure information, and changes the warning level accordingly. The processing unit analyzes the user's falling trend according to the user's various state information and sends the trend information to the mobile terminal, and the mobile terminal determines at least three warning types based on the user's falling trend.

Description

Technical Field

[0001] The present invention relates to the technical field of fall monitoring, and in particular to a real-time monitoring system for fall status and anti-lost position information. Background Art

[0002] Falls are the most common household accidents among the elderly in modern times, which often cause physical damage to the elderly and even threaten their lives, such as cerebral hemorrhage. It is reported that in the millennium, fatal falls and non-fatal falls in the United States caused losses of US$200 million and US$19 billion respectively. In addition, the elderly often suffer from post-traumatic stress disorder after falling. After falling, the elderly often lock themselves in their homes and gradually isolate themselves from society because they are afraid of falling again. But in fact, this isolation causes more falls in the elderly. Given the important role played by the elderly in today's society, the socioeconomic impact of this self-imposed isolation is becoming more and more important. Therefore, there is an urgent need for a real-time monitoring system for fall status and anti-lost location information to ensure the normal life of this group of people.

[0003] Chinese patent CN103976739B discloses a wearable dynamic real-time fall detection method and device, the wearable detection device includes a specific hardware structure composed of feet, waist, wrist, and back; the dynamic real-time fall detection method can complete the real-time judgment of human body falls in any state, and its core algorithm includes a threshold opening logic structure caused by a sudden change in gait acceleration; a human posture dynamic stability judgment algorithm based on zero moment point (ZMP); a fall pattern recognition algorithm based on a human motion database. A complete dynamic real-time detection system is formed, which achieves the purpose of wide application field, low dependence on application environment and real-time detection, and the detection method and device have high accuracy, low system cost, and the advantages of small system volume and easy to wear. However, there are too many detection parts, which cannot be used flexibly in normal use and life, and it is troublesome to wear, and it is difficult to solve the problem of how to set the detection parameters of different parts of the body. When falling, the acceleration changes of various parts of the human body are different, and when the falling posture is different, the acceleration changes of various parts of the human body are also different. The judgment method of this patent has low accuracy and is prone to multiple false alarms. According to the summary data of clinical care in some existing hospitals, one of the places where the elderly are more likely to fall is at the bedside. This is because the elderly can perform many actions at the bedside, such as lying, sitting, standing, bending over, leaning against the bed, etc., so the probability of falling is higher. Therefore, it is necessary to configure fall prevention equipment next to the bed.

[0004] In addition, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the inventor studied a large number of documents and patents when making the present invention, but due to space limitations, not all details and contents are listed in detail. However, this does not mean that the present invention does not have the characteristics of these prior arts. On the contrary, the present invention already has all the characteristics of the prior art, and the applicant reserves the right to add relevant prior art to the background technology. Summary of the invention

[0005] The most widely used solution for falls in the prior art is to provide early warning after a fall. The device is equipped with a button for triggering an alarm and a communication device for sending the alarm signal. However, since a person may be in a state where it is difficult to press a button and / or trigger an alarm after a fall (coma, shock and / or broken arm, etc.), the device cannot respond in real time or even cannot respond. In order to solve the above problems, the prior art uses a portable intelligent fall detection device that can automatically alarm without any intervention from the person who falls. However, how to set up the device to detect various parts of the body is a problem that is difficult to solve in the prior art. The triggering of the alarm of the device follows the quantitative principle, which makes it easy for a variety of false alarms to occur, such as when the user is going to the toilet, getting dressed, or sleeping.

[0006] In view of the shortcomings of the prior art, the technical solution of the present invention is to provide a real-time monitoring system for falling status and anti-lost position information, which includes at least a mobile terminal and a processing unit, wherein the processing unit is configured to store data about user status information sent by at least one of the mobile terminals, and derive the user warning level based on the data of the user status information. The first sensor, the second sensor and the third sensor arranged on the surrounding ground respectively detect the user's movement information, physiological information and pressure information. The mobile terminal divides the user's current state into at least a static state, a motion state, a pre-fall state and a post-fall state according to the warning level obtained from the processing unit and based on the processing of the user's movement information, physiological information and pressure information, and changes the warning level accordingly. The processing unit analyzes the user's falling trend according to the user's various state information and sends the trend information to the mobile terminal, and the mobile terminal determines at least three warning types based on the falling trend, wherein one warning type includes at least a reminder level for movement information, wherein another warning type includes at least a reminder level for physiological information, wherein the third type includes at least a reminder level determined according to ground pressure information. Through the above settings, on the basis of detecting and analyzing the user's current state, a comprehensive analysis and judgment of the user's static state, motion state, pre-fall state and post-fall state are conducted to implement targeted warnings of different degrees. This can not only further enhance the anti-fall effect, but also judge the user's different behavioral states, and take corresponding measures before, during and after the user falls.

[0007] According to a preferred embodiment, the processing unit performs similarity measurement on the movement information and compares it with a representative fall sample, and analyzes the comparison result to issue a high probability fall signal. The high probability fall signal is also improved in reliability by at least adding physiological information, wherein the addition of physiological information includes: the processing unit compares the user's physiological parameters with preset human parameter characteristics obtained at a given exercise level by measuring the difference of the physiological information, and issues a fall warning signal based on the comparison result.

[0008] According to a preferred embodiment, when the first sensor in the mobile terminal is triggered by a high probability fall signal, the mobile terminal activates the second sensor to detect the user's physiological information. Wherein, when the user's physiological information meets the pre-fall state, the processing unit and / or the mobile terminal warns the user based on the reminder level for the physiological information, and when the user is warned, the processing unit stores it as a fall event to record the occurrence time and duration of the fall event. Wherein, if the user continues to confirm that he is in the post-fall state, the processing unit and / or the mobile terminal will raise the level of reminder based on the reminder level for the physiological information.

[0009] According to a preferred embodiment, when the first sensor in the mobile terminal is triggered by a high probability fall signal and the physiological information of the user meets the preset parameter characteristics of the person obtained at a given exercise level, the mobile terminal turns off the first sensor and records information through the second sensor and / or the third sensor.

[0010] According to a preferred embodiment, when it is determined that at least one sensor in the mobile terminal is triggered by a high probability fall signal, the processing unit controls the activation of the third sensor and detects the pressure information based on whether the movement information and / or physiological information obtained by the mobile terminal meets the pre-fall state, thereby judging the current state of the user according to the pressure information to issue an early warning. After the processing unit and / or mobile terminal issues an early warning to the user based on the movement information and / or physiological information, if there is no change trend in the pressure information, or if there is a change trend in the pressure information, the processing unit and / or mobile terminal can modify the early warning level according to whether the pressure signal has a change trend.

[0011] According to a preferred embodiment, when the mobile terminal determines that the user is in a post-fall state based on the state information, the mobile terminal issues an alarm based on the fall trend for the movement information, physiological information and pressure information of the user after the fall, thereby detecting the user's fall state. The alarm level includes at least a mild alarm level, a moderate alarm level and a severe alarm level formulated according to the time when the mobile terminal detects that the user is in a post-fall state.

[0012] According to a preferred embodiment, the first sensor estimates the kinetic energy of the human body by obtaining the quadratic norm of the multidimensional signal, and sends the measured kinetic energy of the human body as movement information to the processing unit, and the processing unit extracts instantaneous features and reference features from at least the movement information detected by the first sensor. The processing unit qualitatively compares the instantaneous features and the reference features through similarity measurement to determine the current state of the user and issue a high probability fall signal.

[0013] According to a preferred embodiment, the processing unit extracts heart rate data and / or blood pressure data from at least the physiological information detected by the second sensor. The frequency of extracting heart rate data and / or blood pressure data by the processing unit based on the physiological information sent by the second sensor is adjusted by at least the change of the movement information.

[0014] According to a preferred embodiment, the movement information includes the acceleration of the mobile terminal, the physiological information includes the heart rate and / or blood pressure, and the pressure information includes the pressure applied to the third sensor.

[0015] According to a preferred embodiment, the processing unit is also used to determine whether the mobile terminal is correctly worn on the user. If the mobile terminal is not correctly worn, the processing unit can issue a prompt and ignore the physiological information detected by the second sensor to correctly determine the user's current state.

[0016] Beneficial technical effects of the present invention:

[0017] The present invention solves the problem of false alarms in the prior art by dividing the user's current state into a static state, a moving state, a pre-fall state, and a post-fall state, and detecting the parameters of the user's different states through a first sensor, a second sensor, and a third sensor. The judgment result of the processing unit and the warning level issued are corrected through multi-parameter measurement to make them consistent with the user's current state. The present invention qualitatively compares the characteristics of each parameter through similarity measurement, and its accuracy is high. Compared with the quantitative principle of the prior art, the present invention has a low probability of false alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1It is a topological diagram of a preferred embodiment of a real-time monitoring system for falling status and anti-lost position information of the present invention;

[0019] Figure 2 It is a module flow chart of a preferred embodiment of a real-time monitoring system for falling status and anti-lost position information of the present invention;

[0020] Figure 3 It is a flow chart of a preferred embodiment of a real-time monitoring system for falling status and anti-lost position information of the present invention.

[0021] Reference numerals list

[0022] 1: processing unit; 2: mobile terminal; 3: first sensor; 4: second sensor; 5: third sensor; 6: warning type. DETAILED DESCRIPTION

[0023] The following is a detailed description with reference to the accompanying drawings.

[0024] Example 1

[0025] The present application relates to a real-time monitoring system for falling status and anti-lost position information, which includes at least a mobile terminal 2 and a processing unit 1. The processing unit 1 is configured to store data about user status information sent by at least one mobile terminal 2, and derive the user warning level based on the data of the user status information. The first sensor 3, the second sensor 4 and the third sensor 5 arranged on the surrounding ground of the mobile terminal 2 respectively detect the user's movement information, physiological information and pressure information. The mobile terminal 2 divides the user's current state into at least a static state, a motion state, a pre-fall state and a post-fall state according to the warning level obtained from the processing unit 1 and based on the processing of the user's movement information, physiological information and pressure information, and thus changes the warning level. Preferably, the processing unit 1 can be a processor chip, such as FPGA, CPU, etc., or a processor of ARM architecture, such as ThunderX2 chip, Cortex-A76 chip, or a cloud-based intelligent chip, such as a cloud-based intelligent chip of model MLU100. Preferably, the mobile terminal 2 can be an intelligent terminal such as a mobile phone, a notebook, etc. Preferably, the mobile terminal 2 is an intelligent terminal that can at least measure the heart rate, temperature, pulse, blood pressure, acceleration, blood oxygen and other data of the human body. Preferably, the mobile terminal 2 may be a wearable device, such as a smart bracelet, a smart insole, a smart watch, etc.

[0026] Preferably, the mobile terminal 2 divides the current state of the user into at least a static state, a motion state, a pre-fall state and a post-fall state according to the warning level obtained from the processing unit 1 and based on the processing of the user's movement information, physiological information and pressure information. Preferably, the relevant information can be collected in a preset manner and the preset information of the static state, the motion state, the pre-fall state and the post-fall state can be obtained by comprehensive analysis. For example, the basic parameters, i.e., the preset information, are obtained by collecting the simulated static, simulated movement and simulated fall data of the user for multiple times. When the warning level changes, the mobile terminal 2 reminds the user and / or the medical center at least by flashing different colors and / or broadcasting languages. Preferably, the processing unit 1 analyzes the user's fall trend according to the user's multiple state information and sends the trend information to the mobile terminal 2, and the mobile terminal 2 determines at least three warning types 6 based on the fall trend, wherein one warning type 6 includes at least a reminder level for the movement information, wherein another warning type 6 includes at least a reminder level for the physiological information, wherein the third type includes at least a reminder level determined according to the ground pressure information. Through the above settings, on the basis of detecting and analyzing the user's current state, a comprehensive analysis and judgment of the user's static state, motion state, pre-fall state and post-fall state are conducted to implement targeted warnings of different degrees. This can not only further enhance the anti-fall effect, but also judge the user's different behavioral states, and take corresponding measures before, during and after the user falls.

[0027] Because falling is difficult to prevent and avoid in advance, the most widely used solution for falling in the prior art is to provide an early warning after falling. The early warning is issued by a sensor for fall detection. The sensor can be integrated by a portable device, which is equipped with a button for triggering an alarm and a communication device for issuing the alarm signal. However, since a person may be in a state where it is difficult to press a button and / or trigger an alarm after falling (coma, shock and / or broken arm, etc.), the device cannot respond in real time or even cannot respond. In order to solve the above problems, the prior art uses an intelligent fall detection device that can automatically alarm without any intervention by the person who falls. The principle of its automatic alarm is mainly based on the detection of a person's appearance or the environment, and it uses a camera and / or a vibration sensor placed in the user's surrounding environment. The advantage of using the above device is that it does not need to be worn with the body. In contrast, the above device can only be used in the installed position and cannot be moved with the person's change of position. In addition, the surrounding environment is always in a changing state, and it is difficult for the device to be in an appropriate detection position, resulting in a more complicated detection of a person's appearance or the environment. When using a camera for appearance detection, it will affect the user's privacy and bring inconvenience to the user's life. For the wearable device, it directly detects the occurrence of a person falling and issues an alarm. The detection method of the device is to detect the sudden change of motion parameters when the fall occurs through a gyroscope and / or an accelerometer. This type of device has a compact structure, low cost, easy to use, and can be worn on various parts of the body (chest, wrist, knee, waist, etc.). However, how to set up the device for detection in various parts of the body is a problem that is difficult to solve in the prior art. For example, a device worn near the center of gravity of the human body can achieve the best detection effect. The detection method is to detect sudden movement along the axis of the human body. If the movement is detected, the inclination of the human body is measured to determine whether the human body falls. The detection method takes the detection of sudden movement as the primary condition, and then performs subsequent detection, thereby reducing power consumption. However, the triggering of the alarm follows the quantitative principle, that is, the mobile acceleration and / or inclination value of the sudden movement is higher or lower than the predetermined threshold, then the alarm is triggered. The defect of this principle is that it is easy to have a variety of false alarms, such as when the user is going to the toilet, dressing or sleeping. In addition, such devices in the prior art often use bracelets for detection, but do not take into account the position changes and acceleration changes that the user's wrist experiences during daily movements, which increases the chance of false alarms. The fall detection data is almost the same as the detection data of the user's hand hitting the table, resulting in that this detection method cannot be used for bracelet-shaped devices, especially for measuring the inclination of the forearm to measure falls.

[0028] The present invention aims to provide a real-time monitoring system for falling status and anti-lost position information, which can be worn by users at will, especially on the wrist, and has reliability in detecting falls.

[0029] According to a preferred embodiment, at least a first sensor 3 for detecting acceleration is provided in the mobile terminal 2 to detect movement information. The first sensor 3 detects the kinetic energy of the human body in three-dimensional space. The first sensor 3 can adopt an accelerometer or a gyroscope. The first sensor 3 estimates the kinetic energy of the human body by obtaining the quadratic norm of the multidimensional signal. In addition, the first sensor 3 can include a memory and a microprocessor, and the microprocessor is programmed to perform the above process, or the above process is performed by the processing unit 1. Specifically, if the first sensor 3 is an accelerometer. The first sensor 3 detects the acceleration at the wrist of the person wearing the device at a certain frequency (for example, at a frequency of 100 to 900 Hz). The processing unit 1 (or microprocessor) extracts the instantaneous feature St at time t from the first sensor 3 to represent the kinetic energy distribution of the person in the time window Δt (including time t). The distribution is represented by the quadratic norm in this application as: e(t)=a x 2 (t)+a y 2 (t)+a z 2 (t). where a x 、a y and a z Represents the acceleration in all directions. It can be found from the above description that the characterization of human kinetic energy can obtain special and reliable features in the distribution in the time window Δt, and the feature can be used to judge whether a person falls. Therefore, if the reference feature recorded in the simulated fall process is similar to the above feature, it indicates that the user has a high probability of falling. The detection of falls in the present invention can be carried out by a qualitative method, that is, the similarity of the analysis feature and the human kinetic energy in the time window Δt is detected, rather than the quantitative measurement of the prior art. Its essence lies in the comparison of the instantaneous kinetic energy of the human body (instantaneous feature) and the predetermined threshold (reference feature). Therefore, the present invention mainly compares the feature St for detecting the current state of the user through mobile information, and compares the mean and variance of the feature with the predetermined threshold. The processing unit 1 (or memory) records the characteristics of the human body falling in all directions in space, including the distribution of the kinetic energy of the human body in the time window Δt.

[0030] The acquisition of reference features is to obtain reference features and record them by statistical information through the mean and variance of movement information. Its statistical method is to identify the specific value of each movement information to align the data statistics, such as the maximum value, minimum value, zero point, maximum slope, etc. The specific value is used to statistically align each movement information to obtain a set of specific features. The reference feature is obtained by averaging the specific features at each moment based on various aligned specific values. Then, the average value of the acceleration caused by gravity is subtracted from the reference feature to eliminate the influence of gravity, and the reference feature is normalized. At each moment, the variance of the value of the specific feature can also be determined. The variance value is set on the diagonal of the matrix, and the instantaneous feature St is compared and analyzed by the reference feature and the matrix. The instantaneous feature St is extracted by the processing unit 1 at time t. The instantaneous feature St characterizes the kinetic energy distribution of the person measured by the first sensor 3 in the time window Δt. The time window Δt is usually between 2 and 3s. The instantaneous feature St is normalized by the second-order norm for comparison and analysis with the reference feature. The instantaneous feature S is also identified by its own specific value, such as the maximum value, minimum value, zero point, maximum slope, etc. The instantaneous feature St is moved so as to move the specific value to the middle of the time window Δt. It should be noted that the specific value of the instantaneous feature St is the same as the specific value recorded for statistical alignment when determining the reference feature. By processing the instantaneous feature St and the reference feature in the above manner, the instantaneous feature St and the reference feature at time t can be obtained, and the time t is arbitrary, the specific value changes over time, and the obtained instantaneous feature St and the reference feature also change over time. The comparison frequency of the instantaneous feature and the reference feature is between 1s / time and 10s / time.

[0031] According to a preferred embodiment, the processing unit 1 qualitatively compares the instantaneous feature and the reference feature by similarity measurement. The comparison is performed by comparing the samples forming each feature. Among the large number of samples contained in the reference feature, some samples have the problem of low reliability due to the estimated variance of the reference feature being too large. In order to eliminate this problem, the present invention uses the Mahalanobis distance to determine it. The above describes how to obtain the reference feature and the matrix. The variance reduces the influence of the sample with the lowest reliability on the final calculation result. The above Mahalanobis distance is normalized, that is, the value of the distance is between 0 and 1, and thus directly reflects the possibility of falling. The preset threshold of the present invention can be defined by those skilled in the art according to the required sensitivity when the user simulates falling. When the detection value is higher than the preset threshold, the probability of the user sending a fall is calculated by the above Mahalanobis distance, and an alarm is issued. Preferably, the comparison frequency between the reference feature and the instantaneous feature is performed according to a preset time interval. Preferably, the selection of the comparison frequency is determined by the amplitude of the time window Δt. For example, the kinetic energy change experienced by the wrist of a falling user is about one second. Therefore, considering the comparison frequency performed by the processing unit 1, the maximum interval between two samples during the comparison process is one second. It should be noted that the comparison is only performed when the user may be in danger of falling (in motion or pre-fall state), for example, when the movement information exceeds a preset threshold.

[0032] According to a preferred embodiment, the processing unit 1 performs similarity measurement on the movement information and compares it with a representative fall sample, and analyzes the comparison result to issue a high probability fall signal. The high probability fall signal is also improved in reliability by at least adding physiological information, wherein the addition of physiological information includes: the processing unit 1 compares the user's physiological parameters with the preset parameter characteristics of the person obtained at a given exercise level by measuring the difference of the physiological information, and issues a fall warning signal based on the comparison result. Preferably, in order to improve the reliability of fall detection, at least the human physiological information detected by the second sensor 4 is used to correct the high probability fall signal. When a human body falls, the physiological parameters of the human body will change to a certain extent, especially the cardiovascular parameters of the human body have a direct relationship with the human body falling. The cardiovascular disorder of the user may be the direct cause of the user's fall or the direct result of the fall event. For example, the loss of consciousness during the fall is caused by a drop in blood pressure. In multiple fall events, the heart rate of the human body will also drop rapidly. In addition, although the fall causes changes in the physiological information of the human body, there is also the possibility that the physiological information changes due to changes in the user's posture. For example, between the standing position and the lying position, the decrease in the user's heart rate is normal due to the reduction in upright pressure. Preferably, the processing unit 1 also corrects the high probability fall signal based on the emotional stress change caused by the user's fall. The user's fall will cause a change in emotional stress, and the user's supine position after falling will cause a double increase in heart rate and blood pressure. This increase will be aggravated by the injury suffered by the user after the fall. Therefore, if the detection abnormality of physiological information (physiological information generated under emotional stress) is accompanied by the detection abnormality of user movement information, it indicates that the probability of falling is greatly increased, making the early warning signal more accurate. Preferably, the second sensor 4 can adopt a photoelectric volumetric gauging sensor. The second sensor 4 is used to provide physiological information, and on this basis, estimate the user's heart rate and related items with blood pressure. When the device is set in the form of a bracelet, the second sensor 4 can be set on the back of the device. The second sensor 4 can also be set on the user's finger in the form of a ring. The second sensor 4 is more accurate in detecting the heart rate, but it is difficult to detect the blood pressure value directly. It is necessary to measure the blood pressure value indirectly by measuring the change in the user's blood volume related to the blood pressure value. Even if the relevant items of the blood pressure value are normalized, the accuracy of the estimation is affected by the sample reference value. In this regard, the present invention processes the physiological information in the manner of processing the mobile information as described above, that is, increases the accuracy of the blood pressure value measurement by variance and weighting. The second sensor 4 can include a memory and a microprocessor, which is programmed to perform the above process, or perform the above process through the processing unit 1. The processing unit 1 (or microprocessor) separately extracts the heart rate data or the blood pressure data from the physiological information detected by the second sensor 4.Preferably, the frequency of extracting heart rate data and / or blood pressure data by the processing unit 1 based on the physiological information sent by the second sensor 4 is at least driven by the change of the movement information to be adjusted. The extraction frequency refers to the frequency at which the processing unit 1 extracts heart rate data and / or blood pressure data based on the physiological information sent by the second sensor 4. The processing unit 1 can divide the current state of the user into at least a static state, a motion state, a pre-fall state and a post-fall state through the movement information. If the processing unit 1 determines that the user is in a normal state (static state) at this time, that is, there is no significant and / or abnormal change in the movement signal, the heart rate data and / or blood pressure data are periodically extracted at a frequency of 4Hz at each heartbeat and stored. Each extracted physiological information determines a mean and a variance, and the above-determined values ​​are in a state of continuous refreshing. Thereafter, when the movement information indicates that the probability of the user falling is high (motion state or pre-fall state), the stored continuously refreshed physiological information is combined with the latest physiological information for normalization processing, and the physiological information obtained at this time is compared with the preset threshold value through the difference measurement. The comparison is determined by the distance between the point and the distribution, such as the Mahalanobis distance. When the user is in motion (aerobic exercise), the instantaneous feature St includes one, two or more physiological information. The preset threshold includes the average value of one, two or more physiological information changing over time. For the physiological information detected by the second sensor 4, the comparison distance with the preset threshold is between 0 and 1. A distance of 0 corresponds to normal physiological information, while a distance of 1 corresponds to physiological information that is theoretically impossible to meet the current state of the user. The closer to the distance 1, the greater the possibility of the user falling. Preferably, the processing unit 1 can sum and weight the physiological information. The comparison distance of the physiological information obtained by the processing unit 1 intersects with the acceleration value of the movement information, and the weighting coefficient between the two is determined according to the reliability of the corresponding distance measurement. Preferably, the processing unit 1 determines the probability of the user falling based on at least the following method. If the movement information shows that the user is in a stationary state, the user is in a sitting, standing or lying position. In this case, if the physiological information shows that the user's blood pressure and heart rate are too large, the user may be in an emergency. If the physiological information is normal, the probability of the user falling is reduced. If the movement information shows that the user is in motion or in a state of pre-fall, the user may be in an emergency. If the physiological information shows that the user's blood pressure and heart rate are too high, but compared with the movement information, it is a normal increase, then the probability of the user falling is significantly reduced, and it is determined that the user is in motion, and there is no need to issue an alarm. If the movement information and physiological information show similar changes before and after the abnormal event, the probability of the user falling is significantly reduced.

[0033] It should be noted that the four states in the present invention (static state, motion state, pre-fall state and post-fall state) characterize the posture characteristics of the user in different situations. The static state refers to the user being in a non-moving state, such as sleeping, sitting still, etc. The motion state refers to the user being in a moving state, such as walking, running, etc. The pre-fall state refers to the situation where the part of the user wearing the device has a sudden acceleration change and / or the user's body has a sudden change in physiological information (heart rate, blood pressure, etc.), causing the processing unit to determine that the user may be about to fall. The post-fall state refers to the user's own physiological state after he has fallen to the ground.

[0034] According to a preferred embodiment, the processing unit 1 can determine whether the device is correctly worn by the user through physiological information and / or received movement information. If the user does not wear it correctly, the processing unit 1 issues an alarm to the user or to the medical center. On this basis, the processing unit 1 also determines whether the wearing position of the device is correct through the detection of physiological information.

[0035] According to a preferred embodiment, when the first sensor 3 in the mobile terminal 2 is triggered by a high probability fall signal, the mobile terminal 2 starts the second sensor 4 to detect the physiological information of the user. Wherein, when the physiological information of the user meets the pre-fall state, the processing unit 1 and / or the mobile terminal 2 warns the user based on the reminder level for the physiological information, and when the user is warned, the processing unit 1 stores it as a fall event to record the occurrence time and duration of the fall event. Wherein, if the user continues to confirm that it is in the post-fall state, the processing unit 1 and / or the mobile terminal 2 will increase the level of reminder based on the reminder level for the physiological information. Preferably, the high probability fall signal is a signal issued by the processing unit 1 when an abnormal situation is detected in the user's mobile information. Preferably, the high probability fall signal can also be a signal issued by the processing unit 1 when it is detected that the user's mobile information exceeds a preset threshold. Preferably, the mobile terminal 2 will only start the second sensor 4 when the high probability fall signal is issued. Through this setting, the power consumption of the mobile terminal 2 can be reduced. Preferably, when the processing unit 1 issues a high probability fall signal, the second sensor 4 records the user's physiological information, and the third sensor 5 records the surrounding ground pressure information. Preferably, the high probability fall signal is issued when the processing unit 1 determines that the user is in a motion state or a pre-fall state.

[0036] According to a preferred embodiment, when the first sensor 3 in the mobile terminal 2 is triggered by a high probability fall signal and the user's physiological information meets the preset parameter characteristics of a person obtained at a given exercise level, the mobile terminal 2 turns off the first sensor 3 and records information through the second sensor 4 and / or the third sensor 5. Preferably, when it is determined that the user is in a pre-fall state, the detection work of the first sensor 3 can be turned off for a short time to reduce the power consumption of the mobile terminal 2. When the user is in a pre-fall state, the first sensor 3 cannot record effective information related to the user's fall, and even if it is recorded, it cannot bring more accurate detection results, so the first sensor 3 is turned off for a short time to save power consumption.

[0037] According to a preferred embodiment, when it is determined that at least one sensor in the mobile terminal 2 is triggered by a high probability fall signal, the processing unit 1 controls the opening of the third sensor 5 and detects the pressure information based on whether the mobile terminal 2 obtains the movement information and / or physiological information to meet the pre-fall state, so as to judge the current state of the user according to the pressure information for early warning. Wherein, after the processing unit 1 and / or the mobile terminal 2 warns the user for the movement information and / or physiological information, when there is no change trend in the pressure information, or when there is a change trend in the pressure information, the processing unit 1 and / or the mobile terminal 2 can correct the early warning level according to whether there is a change trend in the pressure signal. Preferably, the third sensor 5 is arranged on the surrounding ground and used for detecting the pressure. For users who need to maintain a lying posture for a long time, the third sensor 5 is arranged on the surrounding ground beside the bed can greatly increase the reliability of fall detection. The third sensor 5 can be a pressure sensor. The third sensor 5 can be integrated in the ground beside the patient's bed. For example, integrated in the carpet. Preferably, when the processing unit 1 determines that the user is in a pre-fall state, the processing unit 1 can control the third sensor 5 to be turned on to detect the pressure of the surrounding ground. Preferably, the processing unit 1 determines the current state of the patient based at least on the short-term pressure change of the third sensor 5. When the pressure detected by the third sensor 5 exceeds the preset threshold, the processing unit 1 sends a warning signal. When the user falls to the ground, due to the effect of gravity acceleration, the short-term pressure will be greater than the user's weight. When the pressure detected by the third sensor 5 is abnormal or sudden, the user may fall. Preferably, multiple third sensors 5 are set on the ground around the user to detect the range of pressure changes. The processing unit 1 determines the current state of the patient based on several contact points of the patient detected by multiple sensors 5. When the user is standing or sitting on the bed or jumping out of bed, although the third sensor 5 detects pressure changes, the range of pressure changes is small. When the user falls, there are usually pressure changes at multiple contact points, such as support formed by the legs, hands and buttocks or large-area support of the back, rather than only supported by the feet (two-point support) when the user is standing (or other postures). Preferably, the processing unit 1 determines the current state of the user by the coverage and pressure points of the pressure change of the third sensor 5. Preferably, the processing unit 1 performs zoning processing on at least the carpet (or other component) integrated with the third sensor 5 based on the distribution range of the plurality of third sensors 5. The processing unit 1 divides the carpet into at least a near-bed area, a far-bed area, a buffer area or other required areas to detect each area separately. The detection results of different areas are combined with the joint analysis of the first sensor 3 and / or the second sensor 4 to obtain the possible posture and fall risk level of the user in the area, and adjust the warning level and the detection level of the first sensor 3, the second sensor 4 and / or the third sensor 5 accordingly. Especially the detection level of the third sensor 5.Preferably, in this solution, the accuracy of judging the possibility of a user's fall is very important for the fall of the user, especially for the fall of the elderly with limited mobility. Therefore, the following steps are given: when it is detected that the user is in the near-bed area, determine whether the user's movement information and / or physiological information has a sudden change and determine the pressure size and the number of support points of the user in the near-bed area detected by the third sensor 5. If the user's movement information and / or physiological information has not changed, and the pressure size of the third sensor 5 remains unchanged, there is no support point, indicating that the user is lying on the bed to rest or sleep. At this time, the first sensor 3, the second sensor 4 and / or the third sensor 5 are in the first detection level. If the user's movement information and / or physiological information has not changed, and the pressure size of the third sensor 5 remains unchanged, there are two support points, indicating that the user is standing or sitting by the bed. At this time, the first sensor 3, the second sensor 4 and / or the third sensor 5 are also in the first detection level. If the user's movement information and / or physiological information has a sudden change, and the pressure size of the third sensor 5 remains unchanged, there is no support point, indicating that the user is exercising or moving on the bed. At this time, the first sensor 3, the second sensor 4 and / or the third sensor 5 are in the first detection level. If the user's movement information and / or physiological information suddenly changes, and the pressure of the third sensor 5 remains unchanged, and there are two support points, it means that the user is moving at the bedside or wants to leave the bedside. At this time, the first sensor 3 and the second sensor 4 are at the first detection level, and the third sensor 5 is at the second detection level. More specifically, when it is detected that the user is between the near-bed area and the buffer zone, the third sensor 5 in the near-bed area and the buffer zone detects the existence of pressure and has four support points, the processing unit 1 determines that the user is kneeling at the bedside, and the timer integrated in the processing unit 1 and / or the mobile terminal 2 counts. When the counted time exceeds a preset value (for example, one minute), a warning is issued to prompt the user to stand up slowly to prevent fainting due to ischemia in the head. Furthermore, when it is detected that the user is in the far-bed area or the buffer zone, the user is far away from the bed at this time, and may not be able to hold the edge of the bed or fall onto the bed when falling. At this time, the first sensor 3, the second sensor 4 and / or the third sensor 5 are at the second detection level to prevent the user from falling away from the bedside. Preferably, the processing unit 1 also determines the user's fall risk level based on the user's step movement degree and pressure degree in each area. The area setting of the present invention is also to judge the user's step movement degree and pressure degree to reflect the user's current physical state. The above-mentioned step movement degree means: when the support point detected by the third sensor 5 changes, the previous support point and the current support point are measured to obtain the user's step distance and step frequency, and the user's next support point position is predicted based on this.When the deviation between the actual next support point position of the user and the predicted next support point position exceeds a preset threshold (the preset threshold can be a distance deviation and a frequency deviation exceeding 50%), the detection level of the first sensor 3, the second sensor 4 and / or the third sensor 5 is increased. When the user's support point position deviates multiple times (three times or more), the processing unit 1 issues an alarm and reminds the user to move slowly to prevent the user from falling due to unstable center of gravity due to multiple changes in direction. The above-mentioned pressure level refers to the pressure of the user on the ground when moving in each area. By detecting the pressure level, it can be determined whether the user has slipped or sprained his ankle. Preferably, when the user is in a pre-fall state, the first sensor 3, the second sensor 4 and / or the third sensor 5 are at the third detection level for real-time monitoring.

[0038] In the prior art, the fall detection for users is usually performed after the user falls or even after a long time, and the user falls, and an alarm is issued. It can only be used for the user's fall detection but not for early warning prompts based on the user's behavior characteristics, especially for the fall of the elderly. In many cases, the best prompt time has been missed when the fall has already occurred. For example, the elderly may have problems with head congestion and dizziness caused by standing up suddenly after sitting or kneeling for a long time. In addition, various activities of users may cause misjudgment of such devices in the prior art. If they cannot be repeatedly judged and identified by multiple sensors, too many prompts and warnings will cause users to have a rebellious mentality, which is undoubtedly a very energy-consuming problem, and ultimately leads to users' unwillingness or even resistance to such devices, which then causes the problem of no one monitoring when users fall.

[0039] The first detection level, the second detection level and the third detection level reflect the detection degree of each sensor. The third detection level has the highest detection degree, and each sensor detects continuously to ensure the current state of the user and warns of possible falls. The first detection level has the lowest detection degree, and each sensor detects at intervals to save the power consumption of the sensor. In this case, the user is in a relatively safe position or state, and there is no need for long-term detection.

[0040] Furthermore, for the living space where the elderly live, especially for the space next to the bed where the elderly live, whether in the case of a hospital bed or a nursing bed in a nursing facility, or next to the bed where the elderly sleep at home, there are a large number of other people walking around the elderly's bed, which makes some existing solutions for detecting the state of the elderly to prevent them from falling very difficult, because it is impossible to distinguish the posture of the elderly from the posture data of other unrelated people, resulting in information confusion and difficulty in accurately and quickly detecting the corresponding data. Based on this, the present solution further provides that the third sensor (sensor for detecting pressure) described in the above embodiment is configured to record and / or upload data only when contact with a user equipped with a mobile terminal is detected. One feasible way of the above scheme is that the mobile terminal is equipped with a positioning unit, and the location where the third sensor unit is located is also equipped with a positioning unit or has a fixed position range when its position is fixed. Only when the positioning information obtained by the positioning unit of the mobile terminal is consistent with the positioning information or the fixed position range obtained by the positioning unit at the location of the third sensor, the third sensor collects and / or uploads data; or another feasible way is to use the bioelectricity conducted by the human body to establish a loop with the human body through the location where the third sensor is located (such as the carpet next to the bed mentioned above) to send a signal of a special frequency to the mobile terminal worn by the human body (such as a smart watch) to determine whether a specific user (such as the target person of this scheme, i.e., the elderly) has contacted the third sensor. The device can effectively screen specific users and completely eliminate the interference of irrelevant personnel, so that the relevant sensors can accurately collect data of specific users and the relevant processing equipment can quickly process and obtain the posture of the elderly, so as to seize the initiative in preventing the elderly from falling. This solution has significant advantages especially for some nursing scenarios that use nursing robots. Based on the above effects, the processing equipment can only quickly process the data collected from the elderly, and can issue instructions to the nursing robot in advance based on this to manipulate the robot to implement means to prevent the elderly from falling. The significant time advantage obtained by this solution can enable the nursing robot to move directly behind the elderly before the elderly squat and stand up, and operate the mechanical arm to hug or lift up to prevent the elderly from falling after standing up. This solution can significantly reduce false alarms and significantly reduce the amount of data processing, so that the cost of the equipment is significantly reduced, and the battery life can be significantly improved.

[0041] According to a preferred embodiment, the third sensor can also determine whether the elderly person has an imbalance in left and right feet based on the sequence and / or pressure difference of pressures generated by the elderly person's two feet on the third sensor when the elderly person gets out of bed and / or stands up. For example, when the pressure difference between the left and right feet is greater than a preset value, it is determined that the elderly person has an imbalance in left and right feet. Such elderly people are more likely to fall down due to unstable standing when getting out of bed. Based on this, when it is detected that the elderly person has an imbalance in left and right feet, the first or second detector, when detecting that the elderly person has changed from a sleeping state to a clear state, sends a prompt message to remind the elderly person to get out of bed slowly, and at the same time controls the nursing robot to enter an anti-fall preparation state, such as causing the robot to go to the bedside to perform a guiding action.

[0042] According to a preferred embodiment, when the mobile terminal 2 determines that the user is in a post-fall state based on the state information, the mobile terminal 2 issues an alarm based on the fall trend for the user's movement information, physiological information and pressure information after the fall, thereby detecting the user's fall state. Among them, the alarm level includes at least a mild alarm level, a moderate alarm level and a severe alarm level formulated according to the time when the mobile terminal 2 detects that the user is in a post-fall state. Preferably, the mobile terminal 2 is also integrated with a GPS positioning unit to display the user's current location in real time. Preferably, the housing of the mobile terminal 2 is provided with an LED light or a reflective strip to facilitate observation of the user's location at night.

[0043] According to a preferred embodiment, the movement information includes the acceleration of the mobile terminal 2. The physiological information includes the heart rate and / or blood pressure. The pressure information includes the pressure on the third sensor 5.

[0044] According to a preferred embodiment, the processing unit 1 is also used to determine whether the mobile terminal 2 is correctly worn on the user. If the mobile terminal 2 is not correctly worn, the processing unit 1 can issue a prompt and ignore the physiological information detected by the second sensor 4 to correctly determine the user's current state.

[0045] Example 2

[0046] This embodiment is a further supplement to the above embodiment, and the repeated contents will not be repeated here.

[0047] According to a preferred embodiment, the processing unit 1 performs the following steps: Figure 2 Steps shown:

[0048] S1: Detect the user's movement information through the first sensor 3 set in the mobile terminal 2, and detect the user's physiological information through the second sensor 4. Preferably, at least a first sensor 3 for detecting acceleration is provided in the mobile terminal 2 to detect movement information. The first sensor 3 detects the kinetic energy of the human body in three-dimensional space. The first sensor 3 can adopt an accelerometer or a gyroscope. The first sensor 3 estimates the kinetic energy of the human body by obtaining the quadratic norm of the multidimensional signal. In addition, the first sensor 3 can include a memory and a microprocessor, which is programmed to perform the above process, or the above process is performed by the processing unit 1. Preferably, in order to improve the reliability of fall detection, at least the human physiological information detected by the second sensor 4 is used to correct the warning signal. Preferably, the second sensor 4 can adopt a photoelectric volumetric recording sensor. The second sensor 4 is used to provide physiological information, and on this basis, estimate the user's heart rate and related items with blood pressure. When the device is set in the form of a bracelet, the second sensor 4 can be set on the back of the device. The second sensor 4 can also be set on the user's finger in the form of a ring.

[0049] S2: Determine whether the user has fallen or the current state of the user based on whether the movement information and / or physiological information meet the preset threshold. Preferably, the processing unit 1 determines the probability of the user falling based on at least the following method. If the movement information shows that the user is in a stationary state, the user is in a sitting, standing or lying position. In this case, if the physiological information shows that the user's blood pressure and heart rate are too high, the user may be in an emergency. If the physiological information is normal, the probability of the user falling decreases. If the movement information shows that the user is in a state of motion or a pre-fall state, the user may be in an emergency. If the physiological information shows that the user's blood pressure and heart rate are too high, but it is a normal increase compared to the movement information, the probability of the user falling is significantly reduced, and it is determined that the user is in a state of motion, and there is no need to issue an alarm. If the movement information and physiological information show similar changes before and after the abnormal event, the probability of the user falling is significantly reduced.

[0050] Throughout the text, the features referred to as “preferably” are merely optional and should not be understood as having to be set. Therefore, the applicant reserves the right to abandon or delete the relevant preferred features at any time.

[0051] It should be noted that the above specific embodiments are exemplary, and those skilled in the art can come up with various solutions inspired by the disclosure of the present invention, and these solutions also belong to the disclosure scope of the present invention and fall within the protection scope of the present invention. Those skilled in the art should understand that the present invention description and its drawings are illustrative and do not constitute a limitation of the claims. The protection scope of the present invention is defined by the claims and their equivalents.

Claims

1. A real-time monitoring system for falling status and anti-lost position information, comprising at least a mobile terminal (2) and a processing unit (1), wherein the processing unit (1) is configured to store data on user status information sent by at least one of the mobile terminals (2), and to derive a user warning level based on the data on the user status information, characterized in that: A first sensor (3), a second sensor (4) arranged on the mobile terminal (2), and a third sensor (5) arranged on the surrounding ground respectively detect movement information, physiological information, and pressure information of the user. The mobile terminal (2) divides the current state of the user into at least a static state, a moving state, a pre-fall state and a post-fall state according to the warning level obtained from the processing unit (1) and based on the processing of the user's movement information, physiological information and pressure information, and changes the warning level accordingly. The processing unit (1) analyzes the user's falling tendency according to the user's multiple status information and sends the tendency information to the mobile terminal (2), and the mobile terminal (2) determines at least three warning types based on the user's falling tendency, wherein one warning type at least includes a reminder level for movement information, wherein another warning type at least includes a reminder level for physiological information, wherein a third type at least includes a reminder level determined according to ground pressure information, The processing unit (1) performs similarity measurement on the movement information and compares it with a representative fall sample, and analyzes the comparison result to issue a high probability fall signal. The high probability fall signal is further improved in reliability by at least adding physiological information, wherein the adding of physiological information includes: The processing unit (1) compares the user's physiological parameters with preset parameter characteristics of a person obtained at a given exercise level by measuring the difference in physiological information, and issues a fall warning signal based on the comparison result. When it is determined that at least one sensor in the mobile terminal (2) is triggered by a high probability fall signal, the processing unit (1) controls the activation of the third sensor (5) and detects pressure information based on whether the movement information and / or physiological information obtained by the mobile terminal (2) conforms to the pre-fall state, thereby determining the current state of the user based on the pressure information to issue an early warning, wherein: After the processing unit (1) and / or the mobile terminal (2) warns the user regarding movement information and / or physiological information, if the pressure information has no changing trend, or if the pressure information has a changing trend, the processing unit (1) and / or the mobile terminal (2) can modify the warning level according to whether the pressure signal has a changing trend.

2. The real-time monitoring system for falling status and anti-lost position information according to claim 1, characterized in that: When the first sensor (3) in the mobile terminal (2) is triggered by a high probability fall signal, the mobile terminal (2) activates the second sensor (4) to detect physiological information of the user, wherein: When the physiological information of the user meets the pre-fall state, the processing unit (1) and / or the mobile terminal (2) warns the user based on the reminder level of the physiological information, and In the case of warning the user, the processing unit (1) stores the fall event to record the time of occurrence and duration of the fall event, wherein: If the user continues to confirm that he is in a post-fall state, the processing unit (1) and / or the mobile terminal (2) issues a reminder with an increased level based on the reminder level of the physiological information.

3. The real-time monitoring system for falling status and anti-lost position information according to claim 2, characterized in that: When a first sensor (3) in the mobile terminal (2) is triggered by a high probability fall signal and the physiological information of the user meets preset parameter characteristics of a person obtained at a given exercise level, the mobile terminal (2) turns off the first sensor (3) and records information via a second sensor (4) and / or a third sensor (5).

4. The real-time monitoring system for falling status and anti-lost position information according to claim 3, characterized in that: When the mobile terminal (2) determines that the user is in a post-fall state based on the state information, The mobile terminal (2) issues an alarm based on the falling trend for the movement information, physiological information and pressure information of the user after the fall, thereby detecting the falling state of the user, wherein: The alarm levels at least include a mild alarm level, a medium alarm level and a severe alarm level, which are determined according to the time at which the mobile terminal (2) detects that the user is in a post-fall state.

5. The real-time monitoring system for falling status and anti-lost position information according to claim 4, characterized in that: The first sensor (3) estimates the kinetic energy of the human body by obtaining the quadratic norm of the multidimensional signal, and sends the measured kinetic energy of the human body as movement information to the processing unit (1), and the processing unit (1) extracts at least instantaneous features and reference features from the movement information detected by the first sensor (3), wherein: The processing unit (1) qualitatively compares the instantaneous feature with the reference feature through similarity measurement to determine the current state of the user and to issue a high probability fall signal.

6. The real-time monitoring system for falling status and anti-lost position information according to claim 5, characterized in that: The processing unit (1) extracts heart rate data and / or blood pressure data from at least the physiological information detected by the second sensor (4), wherein: The processing unit (1) extracts heart rate data and / or blood pressure data based on the physiological information sent by the second sensor (4), and the frequency of extracting the heart rate data and / or the blood pressure data is adjusted at least by the drive of the change of the movement information.

7. The real-time monitoring system for falling status and anti-lost position information according to claim 6, characterized in that: The movement information includes the acceleration of the mobile terminal (2), The physiological information includes heart rate and / or blood pressure, The pressure information includes the pressure experienced by the third sensor (5).

8. The real-time monitoring system for falling status and anti-lost position information according to claim 7, characterized in that: The processing unit (1) is also used to determine whether the mobile terminal (2) is correctly worn on the user; if the mobile terminal (2) is not correctly worn, the processing unit (1) can issue a prompt and ignore the physiological information detected by the second sensor (4) to correctly determine the current state of the user.

Images