CN112998475B

CN112998475B - Signal induction system for triggering early warning

Info

Publication number: CN112998475B
Application number: CN202110187085.9A
Authority: CN
Inventors: 周清峰; 毛世鑫
Original assignee: Guangdong Sanshui Institute Of Hefei University Of Technology; Ai Gan Technology Guangdong Co ltd
Current assignee: Guangdong Sanshui Institute Of Hefei University Of Technology; Ai Gan Technology Guangdong Co ltd
Priority date: 2018-12-27
Filing date: 2018-12-27
Publication date: 2022-10-11
Anticipated expiration: 2038-12-27
Also published as: CN112998475A; CN112932171A; CN112932171B; CN109480532A; CN109480532B

Abstract

The invention relates to a signal sensing system for triggering early warning, which comprises a mattress body (1), a sensing unit (2) and a processing unit (3), wherein the sensing unit (2) responds to external excitation received by the mattress body (1) and acquires power information of the mattress body (1), and the processing unit (3) identifies the power information based on screening conditions, so that the processing unit (3) can identify knocking signals generated by brainstem/myocardial stem people from the power information comprising a plurality of disturbing signals; when the limbs of the cerebral infarction/myocardial infarction crowd can generate the external excitation, the induction unit (2) corresponding to the limbs is started, so that the induction unit (2) can effectively receive the knocking signal and prevent excessive receiving of the disturbing signal when the cerebral infarction/myocardial infarction crowd has the knocking behavior, and the processing unit (3) can accurately identify the knocking signal.

Description

Signal induction system for triggering early warning

The invention relates to a divisional application of an early warning system based on a knocking signal, which has the application number of 201811606238.3, the application date of 2018, 12 and 27, and the application type of the invention.

Technical Field

The invention belongs to the technical field of intelligent home furnishing, and relates to a signal induction system for triggering early warning.

Background

With the continuous improvement of the material level of people, more and more people begin to pay attention to the quality of life of the people, and young people are provided with multiple pressures formed by combining working pressure, old-age pressure and child-care pressure. Nowadays, many solitary old people and empty nesters appear in the society. Therefore, the phenomena of sudden death due to work, incapability of timely rescue of the old in the independent living emergencies and the like occur in the society. These emergencies are of great social concern as to how society can effectively help the solitary population from the emergencies. With the development of intelligent technologies, more and more intelligent homes or intelligent systems can be used for monitoring the physiological state of a user, but the technical problem of how to timely help the user in an emergency cannot be solved.

For example, chinese patent publication No. CN106406182A discloses an intelligent nursing system based on an intelligent mattress, which includes an environment monitoring module, a sleep quality monitoring module, a data processing module, a control module, a network connection module, an intelligent terminal, an early warning module, a temperature control system, a battery module connected to the control module, and a storage module. The environment monitoring module and the sleep quality monitoring module are respectively connected with the data processing module in a wired or wireless mode; the data processing module is sequentially connected with the control module, the early warning module and the temperature regulation and control system; the control module is connected with the intelligent terminal through the network connection module, and the storage module is connected with the data processing module. The invention can record, store and compare the sleep quality information at any time; the remote home health care can be realized; the operation is simple and the reliability is high.

For example, chinese patent publication No. CN106846735A discloses an intelligent mattress alarm system. The system comprises an acquisition device, an abnormality monitoring module, a cloud server, a mobile terminal and an early warning module; the abnormality monitoring module monitors the sleep data acquired by the acquisition module and judges individual types, a direct early warning mode or an analysis early warning mode is judged based on abnormal changes of the sleep data relative to the individual types, the cloud server unifies physiological information of at least one individual based on the sleep data acquired by the acquisition device and identifies the sleep mode, the physiological information, the individual types, the sleep mode and/or the abnormal data are interactively related to analyze the abnormal state grade of a user, and the cloud server sends corresponding early warning request information to the early warning module and/or the mobile terminal according to the abnormal state grade. The early warning module sends out corresponding early warning information based on the early warning request information, and the mobile terminal sends out rescue information to preset rescue personnel and/or rescue mechanisms based on the early warning request information. The invention carries out early warning based on individual categories, and is accurate and quick.

For example, chinese patent publication No. CN107065719A discloses a data analysis system for an intelligent mattress, which includes a collection device, a channel selection module, a cloud server, an intelligent terminal, and an alarm transmission module, wherein the collection device includes a pressure collection device composed of a plurality of ceramic piezoelectric sensors connected to at least one signal channel, the channel selection module selects at least one signal channel for receiving and transmitting pressure data based on a data source threshold and a data source number of qualified data that is derived from the ceramic piezoelectric sensors and satisfies a data selection condition in the signal channel within a limited time, the cloud server counts first physiological information data and identifies a sleep mode based on the pressure data transmitted by the signal channel, and determines at least one user, analyzes and feeds back sleep quality information, abnormal state information, and/or medical advice of the user to the intelligent terminal in a manner of interactively correlating the first physiological information, the stored second physiological information, and/or the sleep mode. The invention ensures the integrity of the collected data and the accurate analysis result.

For example, chinese patent publication No. CN106618526A discloses a sleep monitoring method and system. The method comprises the following steps: the monitoring system judges the current state of the monitored person and extracts a judgment threshold value of a physiological parameter according to the state, wherein the physiological parameter comprises heart rate, blood pressure and respiratory rate; and reading the heart rate, the blood pressure and the respiratory rate of the monitored person, comparing the heart rate, the blood pressure and the respiratory rate with the judgment threshold of the corresponding parameter, and if the heart rate is higher than the alarm upper limit of the threshold or lower than the alarm lower limit of the threshold, sending an alarm to the appointed family or nursing staff by the monitoring system. In the invention, the judgment threshold value of the physiological parameter is stored on the cloud server, and the personal terminal calls the adaptive judgment threshold value from the server to judge according to the parameters of the monitored person, such as the physique, the age, the sex, the current season and the like. Once any one of the heart rate, the blood pressure or the respiratory rate is abnormal, an alarm can be given in time to prompt family members or medical staff to check.

For example, chinese patent publication No. CN 108245918 a discloses a non-contact monitoring mattress and monitoring system based on multi-sensor fusion, which aims to solve the technical problem of how to monitor patients with physiological disorders through a mattress, including: the mattress comprises a mattress body, a signal processing and analyzing device and an upper computer monitoring device; the mattress body is provided with an infrared thermometer, a piezoelectric film sensor and a conductive fabric; the signal processing and analyzing device comprises a piezoelectric signal processing module, an electrocardiosignal processing module, an AD conversion module and an ARM processing unit; the ARM processing unit processes the converted digital signals to obtain heart rate, respiration rate, temperature and electrocardiosignals. The upper computer monitoring device analyzes the heart rate, the respiration rate, the temperature and the electrocardiosignals to obtain the body action, the sleep condition and the health condition of the patient. The monitoring mattress and the monitoring system can continuously monitor patients for 24 hours, give an alarm prompt to doctors or family members when the conditions are abnormal, timely treat the state of an illness or find accidents of leaving the bed, and are convenient to make rapid remedial measures. However, this patent requires that the patient be in a passive monitoring state and requires full-time monitoring by an intelligent mattress.

For another example, a multifunctional intelligent mattress system disclosed in chinese patent publication No. CN 105534150A. The multifunctional intelligent mattress system comprises a sleep quality monitoring system, an environment monitoring system and an intelligent mobile terminal; the sleep quality monitoring system comprises a respiratory frequency sensor, a heart rate monitoring sensor and an induction unit, the environment monitoring system comprises a temperature sensor, a humidity sensor and a brightness sensor, the intelligent mobile terminal is in data connection with the sleep quality monitoring system and the environment monitoring system, information monitored by the sleep quality monitoring system and the environment monitoring system is transmitted to the intelligent mobile terminal and displayed in a display screen of the intelligent mobile terminal, and a corresponding visual sleep quality report is generated. The intelligent mattress system of this patent fills up the body inside including any one or more in the middle of magnetic asbestos, far-outside red cotton and the anion cotton, can effectual reinforcing health function, cooperatees this kind of health function and sleep quality monitored control system, environment monitored control system and intelligent mobile terminal simultaneously, just can be fine with the effect of keeping healthy through the demonstration of modes image such as data, charts for the user. This patent has at least the following disadvantages: (1) Huge data needs to be collected, so that data processing is difficult, and the intelligent mattress is difficult to operate; (2) insufficient monitoring power for abnormality monitoring of the user.

It is well known that patients with cerebral/myocardial infarction are often accompanied by loss of speech function and loss of limb movement or function. At the moment, the patient can not save oneself through terminals such as the 120 terminal, and if the patient knocks certain objects under the conscious condition, the patient sends out a self-rescue signal, so that a lot of time is provided for self rescue undoubtedly. None of the above prior art techniques or combinations thereof can be used to solve this technical problem.

Disclosure of Invention

Aiming at the defects of the prior art, the invention provides a knocking signal-based early warning system which comprises a mattress body, a sensing unit, a processing unit, an early warning unit, a decision-making unit and a monitoring unit, wherein the sensing unit responds to external excitation received by the mattress body and acquires power information of the mattress body, and the processing unit identifies the power information based on screening conditions, so that the processing unit can identify knocking signals generated by brainstem/myocardial stem crowds from the power information comprising a plurality of disturbing signals; the processing unit marks the power information as a knocking signal and obtains at least one characteristic value of the knocking signal; the decision unit responds to the successful matching of the characteristic value and the corresponding trigger condition, determines a monitoring scheme based on the characteristic value and the matching relation of the characteristic value and the corresponding trigger condition, and starts the monitoring unit based on the monitoring scheme, and the processing unit generates early warning information and transmits the early warning information to the early warning unit based on the physiological data information and/or environmental data information acquired by the monitoring unit under the condition that the physiological data information and/or the environmental data information are abnormal; the early warning unit transmits the matching relation, the characteristic value, the physiological data information and/or the environmental data information to a service terminal, wherein the mattress body is provided with the sensing unit according to a knocking range in which the cerebral infarction/myocardial infarction people can generate the external stimulus, so that the sensing unit can effectively receive the knocking signal and prevent excessive reception of the disturbing signal when the cerebral infarction/myocardial infarction people have knocking behaviors, and the processing unit can accurately identify the knocking signal.

According to a preferred embodiment, the screening conditions include a first screening condition and a second screening condition, wherein, in the case that at least one of the first screening condition or the second screening condition is satisfied, the processing unit marks the dynamic information as the tapping signal and obtains at least one characteristic value of the tapping signal; wherein the first screening condition is configured to: in the sampling time, the ratio of the maximum value MAX of the long-time window to M is greater than or equal to a preset long-time window threshold value; the second screening condition is configured to: in the sampling time, the ratio of MIN to M is greater than or equal to a preset short-time window threshold value, wherein MIN is the maximum value of the short-time window; wherein M is obtained by taking the hysteresis point area as background noise and taking an absolute value; the sampling time is obtained by multiplying the maximum pulse frequency by the maximum time interval between adjacent pulses and adding a lag point number, wherein the lag point number is a data point number reserved before the pulse is triggered for the first time in the sampling time.

According to a preferred embodiment, the trigger condition includes a plurality of consecutive data sub-intervals, and each of the data sub-intervals corresponds to a starting scheme stored in a scheme library; under the condition that the characteristic value falls into a certain data subinterval range, the decision unit can match a starting scheme corresponding to the data subinterval range from a scheme library and start the monitoring unit according to the starting scheme; the starting scheme at least comprises the monitoring device type, the monitoring frequency and the monitoring duration of the monitoring unit.

According to a preferred embodiment, the data subintervals of the trigger condition are configured in at least one of the following ways: a user logs in the system through a manual input unit and inputs basic information of a monitored person to the system and physiological information of the monitored person according to a certain time interval; the deep learning unit can generate a training model based on basic information and physiological information of the monitored personnel, the deep learning unit generates suggestion information for modifying the data subinterval based on the training model and pushes the suggestion information to the service terminal, and the user can confirm whether the data subinterval is modified/updated according to the suggestion information through the manual input unit based on the suggestion information; the deep learning module can also access information similar to or the same as the monitored personnel and revise the training model in a mode of connecting a cloud server through a network at certain time intervals.

According to a preferred embodiment, the decision unit is capable of activating at least one of a pulse acquisition unit, a respiration acquisition unit, a heartbeat acquisition unit, a voice acquisition unit, a video acquisition unit and an image acquisition unit; so that pulse acquisition unit breathe acquisition unit heartbeat acquisition unit voice collection unit video acquisition unit with at least one in the image acquisition unit can with service terminal establishes data connection, thereby service terminal's operating personnel can be based on pulse information that pulse acquisition unit gathered, the respiratory information that breath acquisition unit gathered, the heartbeat information that heartbeat acquisition unit gathered, the pronunciation that voice acquisition unit gathered, the video that video acquisition unit gathered and one in the image that image acquisition unit gathered establish intervention measure in advance or real time monitoring user's state.

According to a preferred embodiment, the characteristic values comprise at least a vibration amplitude, an energy ratio and an energy value mean; under the condition that the vibration amplitude is successfully matched with a first trigger condition, the decision unit starts the monitoring unit according to levels based on the vibration amplitude and the matching relation between the vibration amplitude and the corresponding first trigger condition; under the condition that the vibration amplitude is unsuccessfully matched with the first trigger condition, the processing unit matches the energy ratio with a second trigger condition, and under the condition that the energy ratio is successfully matched with the second trigger condition, the decision unit starts the monitoring unit according to the level based on the energy ratio and the matching relation between the energy ratio and the corresponding second trigger condition; and under the condition that the energy ratio is unsuccessfully matched with the second trigger condition, the processing unit matches the energy value mean value with a third trigger condition, and under the condition that the energy value mean value is successfully matched with the third trigger condition, the decision unit starts the monitoring unit according to levels based on the vibration amplitude and the matching relation between the vibration amplitude and the corresponding third trigger condition.

According to a preferred embodiment, the processing unit comprises a stage IV processing unit, responsive to a failed matching of the mean value of the energy values and the third trigger condition, for establishing a data connection with the stage I processing unit to obtain a first difference of the vibration amplitude and the first trigger condition, establishing a data connection with the stage II processing unit to obtain a second difference of the energy ratio and the second trigger condition, and establishing a data connection with the stage III processing unit to obtain a third difference of the mean value of the energy values and the third trigger condition, respectively; the IV level processing unit respectively matches the first variance, the second variance, and the third variance with a fourth trigger condition, and generates a suspected warning signal if the first variance, the second variance, and the third variance satisfy the fourth trigger condition.

According to a preferred embodiment, the mattress body is a mattress, the sensing unit comprises at least a flexible smart fabric sensor comprising a pressure sensitive layer, an electrode layer and a protective layer, wherein the electrode layer is led out from the surface of the pressure sensitive layer by at least one lead; the protective layer is arranged on the opposite surface of the surface, wherein the mattress body is divided into an upper mattress body and a lower mattress body by the flexible intelligent fabric sensor; at least one layer of insulating layer is arranged between the upper mattress body and/or the lower mattress body and the flexible intelligent fabric sensor.

According to a preferred real-time mode, the invention also discloses a knocking signal early warning method, which comprises the following steps: the induction unit is embedded in the inner part/outer edge of the mattress body and can be in data connection with the processing unit in a wired/wireless mode, wherein the induction unit responds to external excitation received by the mattress body and collects power information of the mattress body, and the processing unit identifies the power information based on screening conditions, so that the processing unit can identify knocking signals generated by cerebral infarction/myocardial infarction crowds from the power information comprising a plurality of disturbing signals; under the condition that the screening conditions of the preset sample library are met, the processing unit marks the power information as a knocking signal and obtains at least one characteristic value of the knocking signal; the decision unit responds to the successful matching of the characteristic value and the corresponding trigger condition and starts the monitoring unit according to the grade based on the matching relation between the characteristic value and the corresponding trigger condition, and the processing unit generates early warning information and transmits the early warning information to the early warning unit based on the physiological data information and/or environmental data information acquired by the monitoring unit under the condition that the physiological data information and/or the environmental data information are abnormal; the early warning unit transmits the matching relationship, the characteristic value, the physiological data information and/or the environmental data information to a service terminal; the mattress body can generate the knocking range of the external excitation according to the cerebral infarction/myocardial infarction crowd, so that the induction unit can effectively receive the knocking signal and prevent excessive reception of the disturbing signal when the cerebral infarction/myocardial infarction crowd has the knocking behavior, and the processing unit can accurately identify the knocking signal.

According to a preferred embodiment, the triggering condition includes a plurality of continuous data subintervals, each subinterval corresponds to a scheme for starting the monitoring unit, and the scheme at least includes a type of a monitoring device of the monitoring unit, a monitoring frequency and a monitoring duration; under the condition that the characteristic value falls into a certain data subinterval range, the decision unit can match a scheme corresponding to the data subinterval range from a scheme library and start the monitoring unit based on the scheme; wherein the data subinterval is configured in at least one of the following ways: a user logs in the system through a manual input unit and inputs basic information of a monitored person to the system and physiological information of the monitored person according to a certain time interval; the deep learning unit can generate a training model based on basic information and physiological information of the monitored personnel, the deep learning unit generates suggestion information for modifying the data subinterval based on the training model and pushes the suggestion information to the service terminal, and the user can confirm whether the data subinterval is modified/updated according to the suggestion information through the manual input unit based on the suggestion information; the deep learning unit can also access information similar to or the same as the monitored personnel and revise the training model in a mode of connecting a cloud server through a network at certain time intervals.

The invention provides a knocking signal-based early warning system, which at least has the following advantages:

(1) The mattress body is provided with a plurality of induction units within the range of knocking limbs according to the external excitation generated by the cerebral infarction/myocardial infarction crowd, and when the limbs of the cerebral infarction/myocardial infarction crowd can generate the external excitation, the induction units corresponding to the limbs are started; the invention can sense the active behavior of the patient along with the rehabilitation of the cerebral infarction/myocardial infarction patient, and can stimulate the rehabilitation of the patient; in addition, the induction unit can effectively receive the knocking signal and prevent receiving excessive disturbing signals when the cerebral infarction/myocardial infarction crowd has the knocking action, so that the processing unit can accurately identify the knocking signal.

(2) The knocking signal and the non-knocking signal can be distinguished; the knocking signal is a signal for sending out help in case of an accident, so that the processing unit must screen the signal collected by the sensing unit to obtain a correct knocking signal instead of a signal generated by normal work for early warning. Through the mode of screening for the first time, will strike the signal and discern and judge from multiple signal, can promote the exactness that this system sent the early warning on the one hand, on the other hand reduces the interference to service terminal in order to the utilization social resource that can be reasonable.

(3) The system can start the monitoring system after identifying the knocking signal, and monitor the state and environment of the user sending the knocking signal, so that the service terminal can take intervention measures, prepare to provide a coping scheme in advance, and obtain valuable clues in a most rapid, most convenient and most accurate mode.

(4) According to the flexible intelligent fabric sensor, the nano-scale sensitive functional material penetrates through the fabric material through an ultrasonic fusion technology, so that the linear range of the sensitive functional material to the deformation of the external pressure is widened, and the acquisition of knocking signals is better completed. For example, for people suffering from cerebral apoplexy, the mattress body cannot be knocked by strong strength, the generated knocking signal is weak, and the flexible intelligent fabric sensor adopts a nano sensitive functional material, so that the acquisition range is widened, and the knocking signal can be acquired.

(5) Since the values of the first trigger condition, the second trigger condition and the third trigger condition are obtained through manual statistics or experience, certain errors exist. In order to prevent the loss of the knocking signal, the invention is also provided with a fourth trigger condition, and the fourth trigger condition aims to overcome the statistical error, improve the precision of the early warning system, and improve the precision of identifying the knocking signal and the non-knocking signal, so that the rescue information can be sent out in time and the non-rescue information can be prevented from being sent out.

Drawings

FIG. 1 is a schematic diagram of logic blocks of a knock signal based early warning system provided by the present invention; and

fig. 2 is a schematic structural diagram of a mattress body of the pre-warning system based on knocking signals, which is provided by the invention.

List of reference numerals

1: a mattress body 10: deep learning unit

2: the sensing unit 11: scheme library

3: the processing unit 12: condition library

4: the early warning unit 1a: upper mattress body

5: the service terminal 1b: lower mattress body

6: sample pool 2a: pressure sensitive layer

7: the decision unit 2b: electrode layer

8: the monitoring unit 2c: protective layer

9: manual input unit 2d: insulating layer

Detailed Description

This is described in detail below with reference to figures 1 and 2.

In the description of the present invention, the terms "first", "second", "third" and "third" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implying any number of technical features indicated. Thus, features defined as "first," "second," "third," and so forth may explicitly or implicitly include one or more of such features. In the description of the present invention, "a plurality" means two or more unless specifically defined otherwise.

In the description of the present invention, it is to be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", "clockwise", "counterclockwise", "inner", etc., indicate orientations or positional relationships based on those shown in the drawings, and are used only for convenience of description and for simplicity of description, but do not indicate or imply that the device or element referred to must have a particular orientation, be constructed and operated in a particular orientation, and thus, are not to be construed as limiting the present invention.

In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "secured," and the like are to be construed broadly and can, for example, be fixedly connected, detachably connected, or integrally connected; can be mechanically or electrically connected; they may be connected directly or indirectly through intervening media, or they may be interconnected between two elements. The specific meanings of the above terms in the present invention can be understood according to specific situations by those of ordinary skill in the art.

In the present invention, the term "detachably" is one of an adhesive, a key connection, a screw connection, a pin connection, a snap connection, a hinge connection, a clearance fit, or a transition fit. The specific meanings of the above terms in the present invention can be understood by those skilled in the art according to specific situations.

Example 1

The invention is particularly suitable for users suffering from acute diseases and elderly users. As is known, when an acute disease patient, such as a stroke patient, basically loses the speech function and then loses the action function, the patient cannot save himself through terminals such as 120, and if the patient knocks a mattress consciously, a self-rescue signal is sent, so that much time is provided for self-rescue undoubtedly. However, since the mattress receives too much external load, in this case, the patient may not be identified by the system and the self-rescue fails even if the patient sends a knocking signal to the mattress body. Therefore, the invention provides an early warning system based on a knocking signal, which is particularly suitable for cerebral infarction/myocardial infarction crowds.

An early warning system based on knocking signals comprises a mattress body 1, a sensing unit 2, a processing unit 3, an early warning unit 4, a decision unit 7 and a monitoring unit 8. The sensing unit 2 is embedded in the inner/outer edge of the mattress body 1 and can establish data connection with the processing unit 3 in a wired/wireless manner. For example, the sensing unit 2 may preferably establish a data connection with the processing unit 3 through a Wifi module, an EnOcean module, or an optical fiber. Preferably, the EnOcean module is an ultra-low power consumption short-distance wireless communication technology based on energy collection, is applied to indoor energy collection, and also has application in smart home, industry, traffic and logistics. The module based on the EnOcean technology has the characteristics of high-quality wireless communication, energy collection and conversion and ultra-low power consumption. In an emergency, the EnOcean module can timely transmit data among the modules of the invention and timely and effectively send out early warning due to the characteristics of high-quality wireless communication, energy collection and conversion and ultralow power consumption.

Preferably, the sensing unit 2 is responsive to external stimuli received by the mattress body 1 and collects dynamic information of the mattress body 1 based on the stress and/or strain generated by the mattress body 1. Wherein, mattress body 1 is provided with a plurality of induction element 2 according to the scope that the cerebral infarction/myocardial infarction crowd can produce the striking limb of external excitation, and when cerebral infarction/myocardial infarction crowd's limb can produce external excitation, induction element 2 that this limb corresponds starts, thereby induction element 2 can effectively receive the signal of striking and prevent to receive too much disturbing signal when cerebral infarction/myocardial infarction crowd has the action of striking, so that processing unit 3 can accurately discern the signal of striking. The mattress body 1 can receive the most external stimuli of users, such as knocking, impacting, breathing, turning over, trampling and the like. Preferably, the power information comprises a plurality of disturbing signals such as a bump signal, a respiration signal, a turn-over signal, a step signal and the like, and a knock signal. The impact signal, the respiration signal, the turn-over signal, the treading signal and the knocking signal mainly comprise frequency, acceleration, speed, energy and the like. The dynamic information is the deformation and stress of the mattress caused by the impact, breathing, turning over, treading and knocking of the user. The knocking signal is a signal for sending out a distress call when a user has an accident. Therefore, the processing unit 3 must screen the signals collected by the sensing unit 2 to obtain the correct tapping signal rather than the signals generated by the normal operation for warning. Patients suffering from a cerebral or cardiac infarction are often accompanied by physical movement disorders such as hemiparalysis or hemiplegia, and therefore it is desirable to arrange sensing units in conjunction with the limbs which they are able to move. However, according to the rehabilitation process of a large number of patients, the muscle strength and the limb movement of the patient suffering from the cerebral infarction or the myocardial infarction are gradually recovered, and the finger movement is recovered firstly, so that the sensing unit 2 can be preferably opened and arranged in the middle of the mattress body 1 to meet the requirement that the patient suffering from the cerebral infarction or the myocardial infarction can be knocked by the fingers. During the treatment, the muscle strength and the joint movement are gradually restored, and the foot and the head start to be restored, so that the sensing units 2 corresponding to the foot and the head can be gradually activated.

The processing unit 3 acquires power information based on the sensing unit 2, reads the screening conditions of the sample database 6, and judges sample data based on the screening conditions. The processing unit 3 identifies the power information based on a screening condition, so that the processing unit 3 can identify a tapping signal generated by a cerebral infarction/myocardial infarction population from the power information including a plurality of disturbing signals. Specifically, the sample library 6 stores therein the screening conditions for the tapping signal. Through the mode of first screening, will strike the signal and discern and judge from multiple signal, can promote the exactness that this mattress body sent the early warning on the one hand, on the other hand has reduced the computational cost that this mattress body sent the early warning. Preferably, the screening conditions comprise a first screening condition and a second screening condition. Preferably, in case the sample data at least meets one of the first or second screening conditions, the processing unit 3 marks the power information as a tapping signal and gets at least one characteristic value of the tapping signal. I.e. the first and second screening conditions are in the logical or relationship, i.e. as long as the power information satisfies one of the screening conditions, the processing unit 3 marks the power information as a tap signal and obtains at least one characteristic value. Preferably, the first screening condition is configured to: and in the sampling time, the ratio of the maximum value MAX of the long-time window to M is greater than or equal to a preset long-time window threshold value. The second screening condition is configured to: within the sampling time, the ratio of MIN to M is greater than or equal to a preset short time window threshold, where MIN is the maximum value of the short time window. Preferably, M is obtained by taking the lag point region as the absolute value of the background noise. Preferably, the sampling time is obtained by multiplying the maximum pulse frequency by the maximum time interval between adjacent pulses plus the number of delay points, where the number of delay points is the number of data points reserved before the first trigger pulse within one sampling time. In the present invention, specific numbers of the length of the sampling time, the length of the long time window, the length of the short time window, the threshold value of the long time window, and the threshold value of the short time window may be obtained by those skilled in the art from experimental data and experience.

Example 2

The embodiment discloses a knocking signal-based early warning system, and the whole and/or part of the contents of the preferred implementation modes of other embodiments can be used as a supplement to the embodiment under the condition of not causing conflict or contradiction. Preferably, the system may be implemented by the method of the present invention and/or other alternative modules.

Preferably, the processing unit 3 is further capable of identifying the tapping signal by using an STA/LTA algorithm, and obtaining the pulse frequency and the time interval of the tapping signal. When the ratio of the average value STA in the long time window to the average value LTA in the short time window is greater than a preset ratio threshold, the processing unit 3 starts the early warning unit 4 at least once. And/or the processing unit 3 records a knocking pulse and the time of the knocking pulse so as to confirm the pulse frequency and the time interval of the knocking signal and store the pulse frequency and the time interval in the abnormal statistic module. In particular, the ratio of STA and LTA reflects the signal level or energy.

Preferably, at least an amplifier, an a/D converter and a filter are connected in series between the sensing unit 2 and the processing unit 3, so that the tapping signal is transmitted to the processing unit 3 in a wired and/or wireless manner after being respectively subjected to an amplification process by the amplifier, a digital-to-analog conversion process by the a/D converter and a dehumidification process by the filter. The knocking force of a user in a dangerous state is usually very small, and the generated wave signal is weak, so that the signal-to-noise ratio of the signal needs to be improved so as to be identified by the sensing unit.

In order to enable the service terminal 5 to quickly monitor the user status and/or the environment of the tapping signal by the monitoring unit 8. Thereby, the service terminal 5 can make intervention measures/rescue plans in advance. And under the condition that the sample data meets the screening conditions of the preset sample library, the processing unit 3 marks the power information as a knocking signal and obtains at least one characteristic value of the knocking signal. Preferably, the decision unit 7 responds to the successful matching of the characteristic value and the corresponding trigger condition, determines the monitoring scheme based on the matching relationship between the characteristic value and the corresponding trigger condition, and starts the monitoring unit 8 based on the monitoring scheme. The processing unit 3 is based on the physiological data information and/or the environmental data information acquired by the monitoring unit 8. Preferably, in the case that the physiological data information and/or the environmental data information has an abnormality, the processing unit 3 generates the warning information and transmits the warning information to the warning unit 4. Preferably, the early warning unit 4 transmits the matching relationship, the characteristic value, the physiological data information and/or the environmental data information to the service terminal 5. The service terminals 5 may be WeChat APP, 120 terminals and 110 terminals. Preferably, the decision unit 7 may employ at least one of a classification regression tree algorithm, an iterative binary tree algorithm, a random forest algorithm, or a conditional decision tree algorithm. Preferably, the system may determine the trigger condition according to the user's experience, e.g. the trigger condition is derived based on multiple simulations or after multiple taps by the user.

Preferably, the trigger condition includes several consecutive data sub-intervals, each sub-interval corresponding to a respective starting scheme stored in the scheme library 11. In the case that the characteristic value falls into a certain data subinterval range, the decision unit 7 can match a starting scheme corresponding to the data subinterval range from the scheme library 11 and start the monitoring unit 8 according to the starting scheme. For example, if the characteristic value is an amplitude value, the trigger condition is that the data subintervals are [10, 20), [20, 30), [30, 40). And the schemes corresponding to the schemes of (10, 20), (20, 30) and (30, 40) are respectively A (starting a video acquisition device), B (starting the video acquisition device and a temperature acquisition device) and C (starting the video acquisition device, the temperature acquisition device and a voice acquisition device). If the characteristic value of the tapping signal is 18, the decision unit 7 initiates the a scheme. Preferably, the decision unit 7 is capable of activating at least one of a pulse acquisition unit, a respiration acquisition unit, a heartbeat acquisition unit, a voice acquisition unit, a video acquisition unit and an image acquisition unit. So that at least one of the pulse acquisition unit, the respiration acquisition unit, the heartbeat acquisition unit, the voice acquisition unit, the video acquisition unit and the image acquisition unit can establish data connection with the service terminal. Therefore, an operator of the service terminal can establish an early intervention measure or monitor the state of the user in real time based on at least one of the pulse information acquired by the pulse acquisition unit, the respiratory information acquired by the respiratory acquisition unit, the heartbeat information acquired by the heartbeat acquisition unit, the voice acquired by the voice acquisition unit, the video acquired by the video acquisition unit and the image acquired by the image acquisition unit. Preferably, the start-up scheme includes at least the kind of monitoring means of the monitoring unit 8, the frequency of monitoring and the duration of monitoring. For example, the monitoring frequency may be applied to the detection of temperature, such as ambient temperature, temperature of the user. For example, the monitoring duration may be determined according to the distance between the location of the tapping signal and the location of the service terminal.

The user can correct and adjust the data subinterval of the trigger condition according to the actual situation. Preferably, the data subinterval of the trigger condition is configured in at least one of the following ways: the user logs in the system through the manual input unit 9 and inputs basic information of the monitored person to the system and physiological information of the monitored person at certain time intervals. The deep learning unit 10 can generate a training model based on the basic information and the physiological information of the monitored person, the deep learning unit 10 generates suggestion information for modifying the data subinterval based on the training model and pushes the suggestion information to the service terminal 5, and the user can confirm whether the data subinterval is modified/updated according to the suggestion information through the manual input unit 9 based on the suggestion information. Preferably, the deep learning module 10 can also access information similar/identical to the monitored person and modify the training model at certain time intervals by connecting to a cloud server through a network. The deep learning unit 10 may employ a neural network algorithm, a regression algorithm, and the like.

Preferably, the characteristic values include at least a vibration amplitude, an energy ratio, and an energy value mean. Wherein, under the condition that the vibration amplitude is successfully matched with the first trigger condition, the decision unit 7 starts the monitoring unit 8 according to the grade based on the vibration amplitude and the matching relation between the vibration amplitude and the corresponding first trigger condition. Under the condition that the vibration amplitude is unsuccessfully matched with the first trigger condition, the processing unit 3 matches the energy ratio with the second trigger condition, and under the condition that the energy ratio is successfully matched with the second trigger condition, the decision unit 7 starts the monitoring unit 8 according to the level based on the energy ratio and the matching relation between the energy ratio and the corresponding second trigger condition. Under the condition that the matching of the energy ratio and the second trigger condition fails, the processing unit 3 matches the energy value mean value with a third trigger condition, and under the condition that the matching of the energy value mean value and the third trigger condition succeeds, the decision unit 7 starts the monitoring unit 8 according to levels based on the vibration amplitude and the matching relation between the vibration amplitude and the corresponding third trigger condition. Preferably, the processing unit 3 includes an I-stage processing unit, a II-stage processing unit, and a III-stage processing unit. The I-stage processing unit is used for obtaining the vibration amplitude and matching the vibration amplitude with the first trigger condition stored in the I-stage processing unit based on the vibration amplitude. In the case of failure of matching the vibration amplitude with the first trigger condition, the II-stage processing unit acquires an energy ratio and performs matching with the second trigger condition stored in the storage unit based on the energy ratio. Wherein, in case the matching of the energy ratio with the second trigger condition fails, the stage III processing unit obtains an energy value mean and matches with a third trigger condition stored in the stage III processing unit based on the energy value mean. Preferably, the I-stage processing unit is configured to: the method is used for obtaining the vibration amplitude of the knocking signal and matching the vibration amplitude with a first trigger condition stored in a sample library based on the vibration amplitude. When the vibration amplitude accords with the first trigger condition, the knocking signal is judged to be effective, and the I-level processing unit sends out warning information to the early warning unit 4. The first trigger condition is obtained through tests and/or statistics for different genders, ages and types of emergency events. The stage III processing unit is configured to: and when the energy ratio does not meet the second trigger condition, triggering, performing short-time Fourier transform on the knocking signal to obtain a time-frequency-energy signal so as to obtain an energy value average value, and matching the energy value average value with a third trigger condition stored in the device. And when the energy value average value meets the third trigger condition, judging that the knocking signal is effective and sending warning information to the early warning unit 4 by the III-level processing unit. The mean value of the energy values is obtained by experiments and/or statistics of different sexes, age groups and types of emergencies.

Since the values of the first trigger condition, the second trigger condition and the third trigger condition are obtained through manual statistics or experience, certain errors exist. In order to prevent the loss of the tapping signal. Preferably, the processing unit 3 comprises an IV-stage processing unit. The stage IV processing unit is used for establishing data connection with the stage I processing unit to acquire a first difference between the vibration amplitude and the first trigger condition, establishing data connection with the stage II processing unit to acquire a second difference between the energy ratio and the second trigger condition, and establishing data connection with the stage III processing unit to acquire a third difference between the energy value mean and the third trigger condition, respectively, in response to the failed matching of the energy value mean and the third trigger condition. The IV-level processing unit respectively matches the first diversity, the second diversity and the third diversity with a fourth trigger condition, and generates a suspected warning signal when the first diversity, the second diversity and the third diversity meet the fourth trigger condition. For example, if the amplitude in the first trigger condition is 20-40 mm and the amplitude generated in the tapping signal is 18mm, the first variance is calculated as | (18-20) |/20=10%. For example, the amplitude is subject to a normal distribution, the mean and standard deviation can be calculated based on statistics, and its confidence interval can be set to 5%. The fourth trigger condition is actually the judgment of the system tolerance rate, so that the loss of the suspected knocking signal is prevented. For example, the fourth trigger condition calculated by percentage may be set to 20%, and if the discrepancy value is less than 20%, it is determined as the suspected-tapping signal. For example, the confidence interval of the fourth trigger condition based on the statistical rule may be set to 5%, and if the difference value is less than 5%, the suspected knocking signal is determined. The arrangement aims to improve the precision of the early warning system, and is used for improving the precision of identifying knocking signals and non-knocking signals, so that rescue information can be sent out in time and non-rescue information can be prevented from being sent out.

Preferably, the mattress body 1 is a mattress. As shown in fig. 2, the sensing unit 2 includes at least a flexible smart fabric sensor. The flexible intelligent fabric sensor comprises a pressure-sensitive layer 2a, an electrode layer 2b and a protective layer 2c, wherein the electrode layer 2b is led out from the surface of the pressure-sensitive layer 2a through at least one lead; the protective layer 2c is provided on the surface opposite to the surface. Wherein, the mattress is divided into an upper mattress body 1a and a lower mattress body 1b by a flexible intelligent fabric sensor; at least one layer of insulating layer 2d is arranged between the upper mattress body 1a and/or the lower mattress body 1b and the flexible intelligent fabric sensor. The sensing unit 2 comprises at least a flexible smart fabric sensor. The flexible smart fabric sensor comprises a pressure sensitive layer 2a, an electrode layer 2b and a protective layer 2c. Preferably, the electrode layer 2b is led out from the surface of the pressure-sensitive layer 2a through at least one lead. The protective layer 2c is provided on the surface opposite to the surface. Preferably, the pressure-sensitive layer 2a is formed by placing the integrated fabric mask in the nano-sensitive functional material liquid, penetrating the nano-sensitive functional material liquid through the interior of the fabric and embedding the nano-sensitive functional material liquid on the surface of the fabric by using ultrasonic waves, and then drying and curing the fabric. The electrode layer 2b is formed by manufacturing a patterned electrode and an electrode lead on one or more of a flexible substrate ethylene-vinyl acetate copolymer, polyethylene glycol and polydimethylsiloxane through a stamping, etching and/or printing method. The patterned electrode and electrode leads may be formed of ultra-thin copper or gold films. The electrode layer 2b is adhered and/or pressed on the surface of the pressure-sensitive layer 2 a. The protective layer 2c is one or more of a polydimethylsiloxane film, an ethylene-vinyl acetate copolymer film, a polyethylene film and a silica gel film. Preferably, the mattress body 1 is divided into an upper mattress body 1a and a lower mattress body 1b by a flexible smart fabric sensor. At least one layer of insulating layer 2d is arranged between the upper mattress body 1a and/or the lower mattress body 1b and the flexible intelligent fabric sensor. The insulating layer 2d may be compounded by a cellulose-based material and an electrically insulating thermoplastic polymer aggregate. The upper mattress body 1 a/the lower mattress body 1b are formed by compounding one or two of a sponge layer and/or a loose cotton layer. The outer layer of the upper mattress body 1 a/the lower mattress body 1b can be wrapped by a chemical fiber cotton felt layer. When the mattress body 1 receives an external load, deformation and stress are generated, and the pressure-sensitive layer 2a generates a piezoelectric signal based on the deformation and stress, which is transmitted to the processing unit 3. According to the flexible intelligent fabric sensor, the nano-scale sensitive functional material penetrates through the fabric material through an ultrasonic fusion technology, so that the linear range of the sensitive functional material to the deformation of the external pressure is widened, and the acquisition of knocking signals is better completed. For example, for the people who suffer from cerebral apoplexy, the mattress body cannot be knocked by strong strength, the generated knocking signal is weak, and the flexible intelligent fabric sensor adopts the nano sensitive functional material, so that the acquisition range is widened, and the knocking signal can be acquired. Of course, the mattress body 1 is not only a mattress, but also at least one of a sofa, a tea table, and a table.

Preferably, the trigger conditions are stored in the condition library 12. The trigger conditions in the condition library 12 can also be updated by the deep learning unit 10. The start-up scenario in the scenario library 11 may also be updated by the deep learning unit 10.

Preferably, the early warning unit 4 comprises a positioning module and a monitoring early warning module. Preferably, the positioning module includes a positioning system in real-time communication with the monitoring and early warning module, the positioning module determines the position information of the knocking signal when the monitoring and early warning module receives the warning information or the suspected warning information, and the monitoring and early warning module generates the early warning signal and pushes the early warning signal to the service terminal 5 when the monitoring and early warning module receives the position information. Preferably, the warning signal at least includes a feature value, position information, a first variance, a second variance, and a third variance. Preferably, the early warning unit 4 transmits the alarm signal to the pre-connected 120, 110 and client terminals in a wired and/or wireless manner.

Example 3

The embodiment discloses a method based on knocking signal early warning. The method can be applied to the mattress recorded by the invention to achieve the expected technical effect. The preferred embodiments of the present invention are described in whole and/or in part in the context of other embodiments, which can supplement the present embodiment, without resulting in conflict or inconsistency.

The method comprises the following steps:

s1: the sensing unit 2 is embedded in the inner/outer edge of the mattress body 1 and can establish data connection with the processing unit 3 in a wired/wireless manner.

S2: the sensing unit 2 is responsive to external stimuli received by the mattress body 1 and collects dynamic information of the mattress body 1 based on the stress and/or strain generated by the mattress body 1. The processing unit 3 identifies the kinetic information based on the screening conditions in the sample library 6 so that the tapping signal can be identified from a plurality of disturbing signals consisting of kinetic information.

S3: the decision unit 7 is responsive to a successful match of the tap signal feature value with its corresponding trigger condition, and in case of a successful match of the feature value with its corresponding trigger condition, the decision unit 7 is able to select at least one start-up scenario from a scenario library for starting the monitoring unit 8. So that the service terminal 5 can rapidly monitor the user state of the knocking signal and/or the environment through the monitoring unit 8 in the form of early warning by the early warning unit 4, and thus the service terminal 5 can make an intervention measure/rescue scheme in advance.

Preferably, the trigger condition includes a plurality of continuous data subintervals, each subinterval corresponds to a starting scheme stored in the scheme library 11, and the starting scheme at least includes a monitoring device type, a monitoring frequency, and a monitoring duration of the monitoring unit 8. In the case that the characteristic value falls into a certain data subinterval range, the decision unit 7 can match a starting scheme corresponding to the data subinterval range from the scheme library 11 and start the monitoring unit 8 according to the starting scheme.

Preferably, the data subintervals are configured in at least one of the following ways: the user logs in the system through the manual input unit 9 and inputs basic information of the monitored person to the system and physiological information of the monitored person at certain time intervals. The deep learning unit 10 can generate a training model based on the basic information and the physiological information of the monitored person, the deep learning unit 9 generates suggestion information for modifying the data subinterval based on the training model and pushes the suggestion information to the service terminal 5, and the user can confirm whether to modify/update the data subinterval according to the suggestion information through the manual input unit 10 based on the suggestion information. The deep learning unit 9 can also access information similar to or the same as monitored personnel and modify the training model in a mode of connecting a cloud server through a network at certain time intervals.

Preferably, the screening conditions comprise a first screening condition and a second screening condition. Preferably, in case the sample data at least meets one of the first or second screening conditions, the processing unit 3 marks the power information as a tapping signal and gets at least one characteristic value of the tapping signal. I.e. the first and second screening conditions are in the logical or relationship, i.e. as long as the power information satisfies one of the screening conditions, the processing unit 3 marks the power information as a tap signal and obtains at least one characteristic value. Preferably, the first screening condition is configured to: and in the sampling time, the ratio of the maximum value MAX of the long-time window to M is greater than or equal to a preset long-time window threshold value. The second screening condition is configured to: within the sampling time, the ratio of MIN to M is greater than or equal to a preset short time window threshold, where MIN is the maximum value of the short time window. Preferably, M is obtained by taking the lag point region as the absolute value of the background noise. Preferably, the sampling time is obtained by multiplying the maximum pulse frequency by the maximum time interval between adjacent pulses plus the number of delay points, where the number of delay points is the number of data points reserved before the first trigger pulse within one sampling time. In the present invention, specific numbers of the length of the sampling time, the length of the long time window, the length of the short time window, the threshold value of the long time window, and the threshold value of the short time window may be obtained by those skilled in the art from experimental data and experience.

It should be noted that the above-mentioned embodiments are exemplary, and that those skilled in the art, having benefit of the present disclosure, may devise various arrangements that are within the scope of the present disclosure and that fall within the scope of the invention. It should be understood by those skilled in the art that the present specification and figures are illustrative only and are not limiting upon the claims. The scope of the invention is defined by the claims and their equivalents.

Claims

1. A signal induction system for triggering early warning is suitable for the crowd suffering from cerebral infarction/myocardial infarction and comprises a mattress body (1), an induction unit (2) and a processing unit (3), wherein the induction unit (2) responds to external excitation received by the mattress body (1) and collects power information of the mattress body (1), and the processing unit (3) identifies the power information based on screening conditions so that the processing unit (3) can identify knocking signals generated by the crowd suffering from cerebral infarction/myocardial infarction from the power information comprising a plurality of disturbing signals;

wherein when the limbs of the cerebral infarction/myocardial infarction crowd can generate the external excitation, the induction unit (2) corresponding to the limbs is started, so that the induction unit (2) can effectively receive the knocking signal and prevent excessive reception of the disturbing signal when the cerebral infarction/myocardial infarction crowd has knocking behavior, and the processing unit (3) can accurately identify the knocking signal,

under the condition that the power information at least meets one of a first screening condition or a second screening condition, the processing unit (3) marks the power information as the knocking signal, obtains at least one characteristic value of the knocking signal, and compares the characteristic value with a corresponding triggering condition, wherein the triggering condition comprises a plurality of continuous data subintervals, each data subinterval corresponds to a starting scheme stored in a scheme library (11),

the trigger condition comprises a plurality of continuous data subintervals, and each data subinterval corresponds to a starting scheme stored in a scheme library (11); under the condition that the characteristic value falls into a certain data subinterval range, the decision unit (7) can match a starting scheme corresponding to the data subinterval range from a scheme library (11) and start the monitoring unit (8) according to the starting scheme; the starting scheme at least comprises the monitoring device type, the monitoring frequency and the monitoring duration of the monitoring unit (8),

the processing unit (3) comprises an I-stage processing unit, a II-stage processing unit and a III-stage processing unit, wherein the I-stage processing unit is used for acquiring the vibration amplitude and matching with the first trigger condition stored in the I-stage processing unit based on the vibration amplitude, the II-stage processing unit acquires the energy ratio and matches with the second trigger condition stored in the II-stage processing unit based on the energy ratio in the case that the vibration amplitude fails to match with the first trigger condition, the III-stage processing unit acquires the energy value mean value and matches with the third trigger condition stored in the III-stage processing unit based on the energy value mean value in the case that the energy ratio fails to match with the second trigger condition,

the processing unit (3) further comprises an IV-level processing unit, responsive to a failed matching of the mean value of energy values and the third trigger condition, for establishing a data connection with the I-level processing unit for obtaining a first variance of the vibration amplitude and the first trigger condition, for establishing a data connection with the II-level processing unit for obtaining a second variance of the energy ratio and the second trigger condition, and for establishing a data connection with the III-level processing unit for obtaining a third variance of the mean value of energy values and the third trigger condition, respectively; the IV-level processing unit matches the first variance, the second variance, and the third variance with a fourth trigger condition, respectively, and generates a suspected alert signal if the first variance, the second variance, and the third variance satisfy the fourth trigger condition.

2. The system according to claim 1, characterized in that in case the processing unit (3) recognizes the tapping signal, the decision unit (7) is capable of selecting at least one activation scheme from a scheme library for activating the monitoring unit (8) based on the characteristic value of the tapping signal, so that the service terminal (5) can quickly monitor the user state and/or the environment of the tapping signal through the monitoring unit (8) in the form of an early warning by the early warning unit (4).

3. The system of claim 1 or 2, wherein the screening conditions include a first screening condition configured such that a ratio of a maximum value MAX of the long time window to M is greater than or equal to a preset long time window threshold value during the sampling time and a second screening condition configured such that a ratio of MIN to M is greater than or equal to a preset short time window threshold value during the sampling time, where MIN is a maximum value of the short time window.

4. The system according to claim 1, characterized in that said mattress body (1) is provided with a plurality of said sensing units (2) according to the range of limbs of the cerebral/myocardial infarction population capable of producing said external stimuli, said sensing units (2) comprising at least a flexible smart fabric sensor;

the mattress is divided into an upper mattress body (1 a) and a lower mattress body (1 b) by the flexible intelligent fabric sensor; at least one layer of insulating layer (2 d) is arranged between the upper mattress body (1 a) and/or the lower mattress body (1 b) and the flexible intelligent fabric sensor.