Distributed information fusion safety monitoring and accompanying system

Abstract

The invention relates to the technical field of health monitoring, and particularly discloses a distributed information fusion safety monitoring and accompanying system comprising a safety monitoring module comprising a posture monitoring unit and a mentality monitoring unit; the monitoring method of the psychological state monitoring unit further comprises monitoring of changes of heart rate and breath. The bedroom sleep monitoring module is used for monitoring a body in a sleep state and comprises monitoring of the heart rate and monitoring of the breath rate in the sleep process. The kitchen environment monitoring module is used for monitoring environment air in a kitchen and judging whether natural gas and / or carbon monoxide gas in the air in the kitchen is influenced or not; and the early warning module is used for carrying out buzzer early warning and / or dialing the telephone number of the guardian based on the abnormal condition obtained by monitoring, so that all-weather and all-directional monitoring on the guardian is realized, the body safety of the person under guardianship is better ensured, and the condition that the accident of the person under guardianship cannot be found in time is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of health monitoring, and in particular to a distributed information fusion safety monitoring and companionship system. Background Art

[0002] When the elderly have health problems, their children, guardians and friends often find it difficult to find out in the first place, making it difficult for the elderly to get help and care in the first place. Therefore, the intelligent monitoring system is particularly important. Existing intelligent monitoring systems usually only collect data and upload it to the children's mobile phones. The service model and service functions are relatively simple, the cost-effectiveness is low, the monitoring effect is not good, and it is difficult to popularize and promote on a large scale. In addition, the system's accuracy is insufficient, resulting in inaccurate monitoring results. This will affect the assessment and timely intervention of the health status of the elderly, reduce the effectiveness of the system, and fail to respond quickly and effectively to the elderly's condition. Summary of the invention

[0003] The purpose of the present invention is to provide a distributed information fusion safety monitoring and companionship system to solve the above-mentioned background problems.

[0004] The purpose of the present invention can be achieved through the following technical solutions:

[0005] A distributed information fusion safety monitoring and companionship system, comprising:

[0006] Safety monitoring module: including posture monitoring unit and mentality monitoring unit;

[0007] The posture monitoring unit is used to monitor the body state of the human body during movement and determine whether the body posture of the human body during movement is abnormal;

[0008] The mentality monitoring unit is used to monitor the heart rate and breathing state of the human body to determine whether the human body's mentality is stable;

[0009] The monitoring method of the state of mind monitoring unit also includes monitoring changes in heart rate and breathing;

[0010] Bedroom sleep monitoring module: The bedroom sleep monitoring module is used to monitor the body during sleep, including heart rate monitoring and respiratory rate monitoring during sleep;

[0011] Kitchen environment monitoring module: used to monitor the ambient air in the kitchen and determine whether there is any impact from natural gas and / or carbon monoxide gas in the air in the kitchen;

[0012] Early warning module: based on the abnormal conditions obtained through monitoring, it will sound a buzzer warning and / or call the guardian.

[0013] As a further solution of the present invention: the monitoring method of the posture monitoring unit specifically includes:

[0014] The motion data at the current time point t is collected in real time by a three-axis acceleration sensor, wherein the motion data includes a forward acceleration value vxt, a lateral acceleration value vyt, and a vertical acceleration value vzt, and the total value of the moving acceleration vj is calculated; the angular velocity data at the current time point t is collected in real time by a gyroscope, wherein the angular velocity data includes a forward rotation velocity value wxt, a lateral rotation velocity value wyt, and a horizontal rotation velocity value wzt;

[0015] Taking unit time T as the monitoring period, obtain the total value vj of the moving acceleration at two adjacent monitoring time points n and m n and vj m ;By BV m =vj m -vj n Calculate the mobile acceleration change value BV at the monitoring time point m m ;

[0016] Get the forward rotation speed value wx at two adjacent monitoring time points n and m n and wx m , via YBx m =|wx m -wx n |Calculate the positive rotation speed variation value YBx m ; Get the lateral rotation speed value wy at two adjacent monitoring time points n and m n and wy m , via YBy m =|wy m -wy n |Calculate the lateral rotation speed variation value YBy m ; Get the horizontal rotation speed value wz at two adjacent monitoring time points n and m n and wz m , via YBz m =|wz m -wz n |Calculate the horizontal rotation speed variation value YBz m ;

[0017] Move the acceleration change value BV m The moving acceleration change threshold BVth is compared, and at the same time, the positive rotation speed change value YBx m , lateral rotation speed variation value YBy m and horizontal rotation speed variation value YBz m Compare with the rotation angular velocity abnormality threshold value YBth;

[0018] If BV m >BVth and max(YBx m , YBy m , YBz m )>YBth; a fall signal is generated.

[0019] As a further solution of the present invention: the calculation method of the total value vj of the moving acceleration is:

[0020] As a further solution of the present invention: the monitoring method of the mentality monitoring unit includes monitoring the heart rate and breathing in real time, specifically:

[0021] The heart rate value and respiratory rate value at the current time point t are monitored in real time by a heart rate sensor and a thermoelectric sensor, which are marked as XTt and HXt respectively; it is determined whether the real-time monitored heart rate value is within the heart rate stable value range [XTs, XTp], and it is determined whether the real-time monitored respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp];

[0022] If the real-time monitored heart rate value XTt is within the heart rate stable value range [XTs, XTp] and the respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp], keep monitoring; otherwise, generate an abnormal indication.

[0023] As a further solution of the present invention: the monitoring method of the mentality monitoring unit also includes monitoring changes in heart rate and breathing, specifically:

[0024] Taking unit time Q as the monitoring period, obtain the heart rate value XT at two adjacent monitoring time points e and f e and XT f ; Simultaneously obtain the respiratory frequency value HX at two adjacent monitoring time points e and f e and HX f ;

[0025] pass Calculate the heart rate fluctuation ratio XTB f ;pass Calculate the respiratory rate fluctuation ratio HXB f ;

[0026] Compare heart rate fluctuations to XTB f It is compared with the heart rate fluctuation ratio threshold XTBY, and the respiratory rate fluctuation ratio HXB f Compare with the respiratory rate fluctuation ratio threshold HXBY;

[0027] If XTB f ≤XTBY and HXBf ≤HXBY, generating a signal of stable mentality;

[0028] If XTB f >XTBY and / or HXB f >HXBY, generates abnormal mentality signal.

[0029] As a further solution of the present invention: the kitchen environment monitoring module obtains the natural gas solubility value and the carbon monoxide concentration value in the air environment at the current time o through the gas sensor, and marks them as RRo and RCo respectively;

[0030] The real-time measured natural gas concentration value RRo is compared with the natural gas concentration safety value RRA, and the current carbon monoxide concentration value RCo is compared with the carbon monoxide safety value RCA;

[0031] If RRo<RRA and RCo<RCA, a hold monitoring signal is generated;

[0032] If RRo≥RRA and / or RCo≥RCA, a kitchen warning signal is generated, and the buzzer is triggered to notify the guardian. At the same time, the kitchen power off protection is activated.

[0033] As a further solution of the present invention: it also includes a cloud platform: used to realize the acquisition, processing and storage management of data; it includes a data acquisition unit, a data processing unit and a storage management unit.

[0034] As a further solution of the present invention: the data acquisition unit is used to collect motion data, angular velocity data, heart rate value, respiratory rate value, natural gas solubility value and carbon monoxide concentration value through sensors.

[0035] As a further solution of the present invention: the data processing unit is used to calculate and process the motion data, angular velocity data, heart rate data, natural gas solubility value and carbon monoxide concentration value acquired by the data acquisition unit.

[0036] As a further solution of the present invention: the safety monitoring module, bedroom sleep monitoring module, kitchen environment monitoring module, early warning module and cloud platform are connected by electricity and / or communication.

[0037] Beneficial effects of the present invention:

[0038] In the present invention, by monitoring the multi-angle acceleration and angular velocity of the body state during the movement, it is determined whether the body falls during the movement, and at the same time, in conjunction with heart rate and breathing monitoring, the body state during movement, falling or sleeping is monitored, thereby realizing all-weather monitoring and guardianship of the ward to avoid situations where accidents to the ward cannot be discovered in time; in addition, the system can also monitor the living environment of the ward, such as natural gas or carbon monoxide in the air in the kitchen, to avoid safety hazards such as fire poisoning to the ward, thereby realizing all-weather and all-round monitoring of the guardian, and better ensuring the physical safety of the ward. The system monitors angular velocity, angular velocity, heart rate and breathing in many aspects with good accuracy and fast response, and can quickly and effectively respond to accidents of the guardian. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The present invention will be further described below in conjunction with the accompanying drawings.

[0040] Figure 1 is a system block diagram of the present invention;

[0041] Figure 2 It is a schematic diagram of the safety monitoring module in the present invention;

[0042] Figure 3 It is a schematic diagram of the cloud platform in the present invention. DETAILED DESCRIPTION

[0043] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0044] Embodiment 1

[0045] See also Figure 1-Figure 3 As shown, the present invention is a distributed information fusion safety monitoring and companionship system, comprising:

[0046] Safety monitoring module: including posture monitoring unit and mentality monitoring unit;

[0047] The posture monitoring unit is used to monitor the body state of the human body during movement and determine whether the body posture of the human body during movement is abnormal;

[0048] The monitoring method of the posture monitoring unit specifically includes:

[0049] S11: collecting motion data at the current time point t in real time through a three-axis acceleration sensor, wherein the motion data includes a forward acceleration value vxt, a lateral acceleration value vyt, and a vertical acceleration value vzt;

[0050] S12: Based on the motion data at the real-time time point t, the total value vj of the moving acceleration is calculated;

[0051] Specifically: The calculation method of the total value of moving acceleration vj is:

[0052] S13: collecting angular velocity data at the current time point t in real time through a gyroscope, wherein the angular velocity data includes a forward rotation velocity value wxt, a lateral rotation velocity value wyt, and a horizontal rotation velocity value wzt;

[0053] S14: Taking unit time T as the monitoring period, obtain the total value vj of the moving acceleration at two adjacent monitoring time points n and m n and vj m ;

[0054] By BV m =vj m -vj n Calculate the mobile acceleration change value BV at the monitoring time point m m ;

[0055] Get the forward rotation speed value wx at two adjacent monitoring time points n and m n and wx m , via YBx m =|wx m -wx n |Calculate the positive rotation speed variation value YBx m ;

[0056] Get the lateral rotation speed value wy at two adjacent monitoring time points n and m n and wy m , via YBy m =|wy m -wy n |Calculate the lateral rotation speed variation value YBy m ;

[0057] Get the horizontal rotation speed value wz at two adjacent monitoring time points n and m n and wz m , via YBz m =|wz m -wz n |Calculate the horizontal rotation speed variation value YBz m ;

[0058] S15: Move the acceleration change value BV m The moving acceleration change threshold BVth is compared, and at the same time, the positive rotation speed change value YBx m , lateral rotation speed variation value YBy m and horizontal rotation speed variation value YBz m Compare with the rotation angular velocity abnormality threshold value YBth;

[0059] If BV m >BVth and max(YBx m , YBy m , YBz m )>YBth; then a fall signal is generated;

[0060] Otherwise, keep monitoring;

[0061] The mentality monitoring unit is used to monitor the heart rate and breathing state of the human body to determine whether the human body's mentality is stable;

[0062] The monitoring method of the mentality monitoring unit includes monitoring the heart rate and breathing in real time, specifically:

[0063] S21: real-time monitoring of the heart rate value and respiratory rate value at the current time point t by means of a heart rate sensor and a thermoelectric sensor, which are marked as XTt and HXt respectively;

[0064] S22: Determine whether the real-time monitored heart rate value is within the heart rate stable value range [XTs, XTp], and determine whether the real-time monitored respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp];

[0065] If the real-time monitored heart rate value XTt is within the heart rate stable value range [XTs, XTp] and the respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp], keep monitoring;

[0066] Otherwise, an exception indication is generated;

[0067] The monitoring method of the mentality monitoring unit also includes monitoring changes in heart rate and breathing, specifically:

[0068] V21: Take unit time Q as monitoring period, and obtain the heart rate value XT at two adjacent monitoring time points e and f. e and XT f ; Simultaneously obtain the respiratory frequency value HX at two adjacent monitoring time points e and f e and HX f ;

[0069] V22: Passed Calculate the heart rate fluctuation ratio XTB f ;pass Calculate the respiratory rate fluctuation ratio HXB f ;

[0070] V23: Heart rate fluctuations compared to XTB f It is compared with the heart rate fluctuation ratio threshold XTBY, and the respiratory rate fluctuation ratio HXB f Compare with the respiratory rate fluctuation ratio threshold HXBY;

[0071] If XTB f ≤XTBY and HXB f ≤HXBY, generating a signal of stable mentality;

[0072] If XTB f >XTBY and / or HXB f >HXBY, generating abnormal mentality signal;

[0073] Bedroom sleep monitoring module: The bedroom sleep monitoring module is used to monitor the body during sleep, including heart rate monitoring and breathing rate monitoring during sleep.

[0074] Kitchen environment monitoring module: used to monitor the ambient air in the kitchen and determine whether there is any impact from natural gas and / or carbon monoxide gas in the air in the kitchen;

[0075] Specifically, the natural gas solubility value and the carbon monoxide concentration value in the air environment at the current time o are obtained through the gas sensor and marked as RRo and RCo respectively;

[0076] The real-time measured natural gas concentration value RRo is compared with the natural gas concentration safety value RRA, and at the same time, the current carbon monoxide concentration value RCo is compared with the carbon monoxide safety value RCA;

[0077] If RRo<RRA and RCo<RCA, a hold monitoring signal is generated;

[0078] If RRo≥RRA and / or RCo≥RCA, a kitchen warning signal is generated, and the buzzer is triggered to notify the guardian, and at the same time, the kitchen power-off protection is activated;

[0079] The early warning module will sound a buzzer warning and / or call the guardian based on the abnormal conditions obtained by monitoring;

[0080] Cloud platform: used to achieve data acquisition, processing and storage management; including data acquisition unit, data processing unit and storage management unit.

[0081] Data acquisition unit: used to collect motion data, angular velocity data, heart rate value, respiratory rate value, natural gas concentration value and carbon monoxide concentration value through sensors;

[0082] Data processing unit: used to calculate and process the motion data, angular velocity data, heart rate data, natural gas solubility value and carbon monoxide concentration value obtained by the data acquisition unit.

[0083] The system monitors the body's state of movement through multi-angle acceleration and angular velocity to determine whether the body falls during movement. At the same time, it cooperates with heart rate and breathing monitoring to monitor the body's state during movement, falling or sleeping. It can realize all-weather monitoring and supervision of the guardian to avoid accidents of the guardian that cannot be discovered in time. In addition, the system can also monitor the living environment of the guardian, such as natural gas or carbon monoxide in the air in the kitchen, to avoid safety hazards such as fire poisoning of the guardian, and realize all-weather and all-round monitoring of the guardian to better ensure the physical safety of the guardian. Through angular velocity, angular velocity, heart rate and breathing and other aspects of monitoring, the system has good accuracy and fast response, and can quickly and effectively respond to the guardian's accidents.

[0084] Embodiment 2

[0085] Based on the above embodiments, this implementation provides a distributed information fusion safety monitoring and companionship method, including the following steps:

[0086] Step 01: The motion data at the current time point t is collected in real time through a three-axis acceleration sensor, wherein the motion data includes a forward acceleration value vxt, a lateral acceleration value vyt, and a vertical acceleration value vzt; it should be noted that the forward acceleration value vxt represents the acceleration of the body in the front-to-back direction during movement, the lateral acceleration value vyt represents the acceleration of the body in the lateral direction during movement, and vzt represents the acceleration of the body in the up-and-down direction during movement;

[0087] Step 02: Based on the motion data at the real-time time point t, calculate and obtain the total value of movement acceleration vj;

[0088] Specifically: The calculation method of the total value of moving acceleration vj is:

[0089] Step 03: The angular velocity data at the current time point t is collected in real time by the gyroscope, wherein the angular velocity data includes a forward rotation velocity value wxt, a lateral rotation velocity value wyt, and a horizontal rotation velocity value wzt; it should be noted that the forward rotation velocity value wxt is the rotation angular velocity of the body rolling forward and backward, the lateral rotation velocity value wyt is the rotation angular velocity of the body rolling left and right, and the horizontal rotation velocity value wzt is the angular velocity of the body rotating around the vertical axis;

[0090] Step 04: Take the unit time T as the monitoring period and obtain the total value vj of the moving acceleration at two adjacent monitoring time points n and m n and vj m ;

[0091] By BV m =vj m -vj n Calculate the mobile acceleration change value BV at the monitoring time point m m ;

[0092] Get the forward rotation speed value wx at two adjacent monitoring time points n and m n and wx m , via YBx m =|wx m -wx n |Calculate the positive rotation speed variation value YBx m ;

[0093] Get the lateral rotation speed value wy at two adjacent monitoring time points n and m n and wy m , via YBy m =|wy m -wy n |Calculate the lateral rotation speed variation value YBy m ;

[0094] Get the horizontal rotation speed value wz at two adjacent monitoring time points n and m n and wz m , via YBz m =|wz m -wz n |Calculate the horizontal rotation speed variation value YBz m；

[0095] Step 05: Change the moving acceleration value BV m The moving acceleration change threshold BVth is compared, and at the same time, the positive rotation speed change value YBx m , lateral rotation speed variation value YBy m and horizontal rotation speed variation value YBzm Compare with the rotation angular velocity abnormality threshold value YBth;

[0096] If BV m >BVth and max(YBx m , YBy m , YBz m )>YBth; then a fall signal is generated; at this time, based on the fall signal, the guardian is notified;

[0097] Step 06: Based on the fall signal, the heart rate sensor and the pyroelectric sensor are used to monitor the heart rate value and the respiratory rate value at the current time point t in real time, which are marked as XTt and HXt respectively;

[0098] Determine whether the real-time monitored heart rate value is within the heart rate stable value range [XTs, XTp], where XTs is the minimum stable heart rate value and XTp is the maximum stable heart rate value; at the same time, determine whether the real-time monitored respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp], where HXs is the minimum stable respiratory rate value and HXp is the maximum stable respiratory rate value;

[0099] If the real-time monitored heart rate value XTt is within the heart rate stable value range [XTs, XTp] and the respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp], keep monitoring;

[0100] Otherwise, an abnormal indication is generated; after notifying the guardian, the buzzer sounds an early warning and notifies the guardian to perform first aid;

[0101] By monitoring the acceleration and angular velocity of the guardian's body movement state, it is possible to determine whether the body has fallen. When the body state is abnormal, the guardian can be notified promptly and quickly. In addition, the heart rate and respiratory rate of the guardian can be monitored to determine whether the heart rate and respiratory state after the guardian falls are abnormal. When the mentality is abnormal, a buzzer warning is used in time. On the one hand, the guardian or people around can be notified, and on the other hand, the guardian can choose whether to call the police for emergency treatment according to the current situation.

[0102] Embodiment 3

[0103] Based on the above embodiments, this embodiment provides a distributed information fusion safety monitoring and accompanying method, including the following steps:

[0104] Step W1: Taking unit time Q as the monitoring period, obtain the heart rate value XT at two adjacent monitoring time points e and f e and XT f ; Simultaneously obtain the respiratory frequency value HX at two adjacent monitoring time points e and f e and HXf ;

[0105] Step W2: Pass Calculate the heart rate fluctuation ratio XTB f ;pass Calculate the respiratory rate fluctuation ratio HXB f ;

[0106] Step W3: Compare heart rate fluctuation to XTB f It is compared with the heart rate fluctuation ratio threshold XTBY, and the respiratory rate fluctuation ratio HXB f Compare with the respiratory rate fluctuation ratio threshold HXBY;

[0107] If XTB f ≤XTBY and HXB f ≤HXBY, generating a signal of stable mentality;

[0108] If XTB f >XTBY and / or HXB f >HXBY, an abnormal mentality signal is generated; at this time, it means that the heart rate and breathing state of the person being supervised have fluctuated greatly in a short period of time, and the heart rate or breathing may be abnormal, so the guardian will be notified;

[0109] By monitoring the fluctuations in the heart rate and respiratory rate of the ward, the ward's daily condition can be monitored. At the same time, the ward's sleep process can be monitored, achieving 24-hour monitoring, achieving better care, and solving the problem of not being able to monitor the ward at night.

[0110] Embodiment 4

[0111] Based on the above embodiments, this embodiment provides the following embodiments for the technical solution:

[0112] Through the camera, we can directly observe the human body's movements and use the mediapipe algorithm to extract the human body's characteristic joints. We can extract the three joints of the eyes, hips, and heels for angle calculation. When the angle is greater than 170°, it is judged as a fall. Taking into account the height of the human body and the furniture (such as the bed), we divide the picture into possible fall areas and impossible fall areas. The angle calculation is valid only when all the nodes are in the possible fall area. The control of the fatness of the human body and the height of the furniture is the result of the training model.

[0113] Based on the fall detection method: we can perform intelligent threshold adjustment:

[0114] The current method uses fixed thresholds (such as acceleration change thresholds, rotation speed change thresholds) to determine fall events. We can introduce an adaptive threshold mechanism.

[0115] Data-based dynamic thresholds: The thresholds are adjusted by learning the system state in real time. For example, when a person's activity state changes (such as movement, stillness, etc.), the system can dynamically adjust the thresholds to improve the fall detection capability in different scenarios.

[0116] Model-based threshold updates: Use a trained machine learning model to automatically update thresholds based on data streams (such as acceleration, rotation speed, etc. within a time window) without manual intervention.

[0117] Event-based fall detection:

[0118] Event-driven sampling and data processing: By introducing an event-driven sensor data processing method, further complex calculations (such as long data analysis) are initiated only when abnormal acceleration changes or rotation changes are detected, which can reduce the computational burden and improve real-time performance.

[0119] Multi-stage detection: First, potential fall signals (such as a sudden increase in acceleration) are detected, and then multi-stage verification is performed based on the time window of the data, changes in rotation speed, and even changes in posture.

[0120] In the application, R60ABD1 millimeter-wave radar and R60AFD1 millimeter-wave radar are used to detect the bedroom and kitchen environment. The bedroom system includes a large voice model to realize voice interaction function, infrared remote control to realize one-key weather broadcast and one-key phone call function. Self-designed circuit realizes charging operation and power-off protection function.

[0121] Methods for monitoring heart rate and breathing may include the following:

[0122] Adopt a multi-level abnormal judgment mechanism:

[0123] Multi-dimensional anomaly detection: In addition to heart rate and respiratory rate, multi-dimensional abnormality judgment can also be performed by combining heart rate variability (HRV), breathing patterns (such as irregular breathing, deep and shallow breathing, etc.), skin temperature changes and other signals. By introducing these physiological indicators, the accuracy of detecting abnormal mental states (such as anxiety, tension, fear, etc.) can be improved.

[0124] For example, low heart rate variability (HRV) is often associated with negative emotions such as stress and anxiety, while irregular breathing patterns are often associated with tension or acute stress.

[0125] Collaborative analysis based on multiple physiological data can capture changes in mentality more accurately.

[0126] Intelligent optimization based on group data:

[0127] Group data learning: Through cross-user anonymous data analysis and integration, a group-level "mentality fluctuation model" is constructed to understand the emotional fluctuation patterns of different groups in specific environments or situations. For example, the mentality reactions of users of different age groups, genders or lifestyles in similar environments can be analyzed, thereby continuously optimizing the individual mentality monitoring model.

[0128] Group models and individual differences: By analyzing the data of multiple users, these group characteristics are integrated into the individual monitoring model, making the individual model more accurate with the assistance of group data.

[0129] In addition, AI-based predictive models can be designed:

[0130] Specifically, machine learning algorithms (such as random forests and support vector machines) are introduced to predict the trend of gas concentrations, so as to determine whether there is any danger in advance.

[0131] Predicting gas concentration trends: The model is trained with historical data to predict the changing trend of gas concentrations in a short period of time. This means that the system not only relies on the concentration value at the current moment, but can also warn of potential dangers in advance. For example, when it is predicted that the concentration of natural gas or carbon monoxide will exceed the safety threshold in a short period of time, a safety alarm can be triggered in advance.

[0132] Abnormal behavior detection: By monitoring kitchen usage patterns (for example, how long and how often the gas stove is turned on), combined with AI models to detect whether there is abnormal usage behavior (such as forgetting to turn off the gas stove). Combined with sudden changes in gas concentration, the AI ​​model can better identify potential dangers.

[0133] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A distributed information fusion safety monitoring and companionship system, characterized in that: include: Safety monitoring module: including posture monitoring unit and mentality monitoring unit; The posture monitoring unit is used to monitor the body state of the human body during movement and determine whether the body posture of the human body during movement is abnormal; The mentality monitoring unit is used to monitor the heart rate and breathing state of the human body to determine whether the human body's mentality is stable; The monitoring method of the state of mind monitoring unit also includes monitoring changes in heart rate and breathing; Bedroom sleep monitoring module: The bedroom sleep monitoring module is used to monitor the body during sleep, including heart rate monitoring and respiratory rate monitoring during sleep; Kitchen environment monitoring module: used to monitor the ambient air in the kitchen and determine whether there is any impact from natural gas and / or carbon monoxide gas in the air in the kitchen; Early warning module: based on the abnormal conditions obtained through monitoring, it will sound a buzzer warning and / or call the guardian.

2. A distributed information fusion safety monitoring and companionship system according to claim 1, characterized in that: The monitoring method of the posture monitoring unit specifically includes: The motion data at the current time point t is collected in real time by a three-axis acceleration sensor, wherein the motion data includes a forward acceleration value vxt, a lateral acceleration value vyt, and a vertical acceleration value vzt, and the total value of the moving acceleration vj is calculated; the angular velocity data at the current time point t is collected in real time by a gyroscope, wherein the angular velocity data includes a forward rotation velocity value wxt, a lateral rotation velocity value wyt, and a horizontal rotation velocity value wzt; Taking unit time T as the monitoring period, obtain the total value vj of the moving acceleration at two adjacent monitoring time points n and m n and vj m ;By BV m =vj m -vj n Calculate the mobile acceleration change value BV at the monitoring time point m m ; Get the forward rotation speed value wx at two adjacent monitoring time points n and m n and wx m , via YBx m =|wx m -wx n |Calculate the positive rotation speed variation value YBx m ; Get the lateral rotation speed value wy at two adjacent monitoring time points n and m n and wy m , via YBy m =|wy m -wy n Calculate the lateral rotation speed variation value YBy m ; Get the horizontal rotation speed value wz at two adjacent monitoring time points n and m n and wz m , via YBz m =|wz m -wz n |Calculate the horizontal rotation speed variation value YBz m ; Move the acceleration change value BV m The moving acceleration change threshold BVth is compared, and at the same time, the positive rotation speed change value YBx m , lateral rotation speed variation value YBy m and horizontal rotation speed variation value YBz m Compare with the rotation angular velocity abnormality threshold value YBth; If BV m >BVth and max(YBx m , YBy m , YBz m )>YBth; a fall signal is generated.

3. A distributed information fusion safety monitoring and companionship system according to claim 2, characterized in that: The calculation method of the total moving acceleration value vj is:

4. A distributed information fusion safety monitoring and companionship system according to claim 1, characterized in that: The monitoring method of the mentality monitoring unit includes monitoring the heart rate and breathing in real time, specifically: The heart rate value and respiratory rate value at the current time point t are monitored in real time by a heart rate sensor and a thermoelectric sensor, which are marked as XTt and HXt respectively; it is determined whether the real-time monitored heart rate value is within the heart rate stable value range [XTs, XTp], and it is determined whether the real-time monitored respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp]; If the real-time monitored heart rate value XTt is within the heart rate stable value range [XTs, XTp] and the respiratory rate value HXt is within the respiratory rate stable value range [HXs, HXp], keep monitoring; otherwise, generate an abnormal indication.

5. A distributed information fusion safety monitoring and companionship system according to claim 1, characterized in that: The monitoring method of the mentality monitoring unit also includes monitoring changes in heart rate and breathing, specifically: Taking unit time Q as the monitoring period, obtain the heart rate value XT at two adjacent monitoring time points e and f e and XT f ; Simultaneously obtain the respiratory frequency value HX at two adjacent monitoring time points e and f e and HX f ; pass Calculate the heart rate fluctuation ratio XTB f ;pass Calculate the respiratory rate fluctuation ratio HXB f ; Compare heart rate fluctuations to XTB f It is compared with the heart rate fluctuation ratio threshold XTBY, and the respiratory rate fluctuation ratio HXB f Compare with the respiratory rate fluctuation ratio threshold HXBY; If XTB f ≤XTBY and HXB f ≤HXBY, generating a signal of stable mentality; If XTB f >XTBY and / or HXB f >HXBY, generates abnormal mentality signal.

6. A distributed information fusion safety monitoring and companionship system according to claim 1, characterized in that: The kitchen environment monitoring module obtains the natural gas solubility value and carbon monoxide concentration value in the air environment at the current time o through the gas sensor, and marks them as RRo and RCo respectively; The real-time measured natural gas concentration value RRo is compared with the natural gas concentration safety value RRA, and the current carbon monoxide concentration value RCo is compared with the carbon monoxide safety value RCA; If RRo<RRA and RCo<RCA, a hold monitoring signal is generated; If RRo≥RRA and / or RCo≥RCA, a kitchen warning signal is generated, and the buzzer is triggered to notify the guardian. At the same time, the kitchen power off protection is activated.

7. A distributed information fusion safety monitoring and companionship system according to claim 1, characterized in that: It also includes a cloud platform: used to realize the acquisition, processing and storage management of data; including a data acquisition unit, a data processing unit and a storage management unit.

8. A distributed information fusion safety monitoring and companionship system according to claim 7, characterized in that: The data acquisition unit is used to collect motion data, angular velocity data, heart rate value, respiratory rate value, natural gas solubility value and carbon monoxide concentration value through sensors.

9. A distributed information fusion safety monitoring and companionship system according to claim 7, characterized in that: The data processing unit is used to calculate and process the motion data, angular velocity data, heart rate data, natural gas concentration value and carbon monoxide concentration value acquired by the data acquisition unit.

10. A distributed information fusion safety monitoring and companionship system according to claim 9, characterized in that: The safety monitoring module, bedroom sleep monitoring module, kitchen environment monitoring module, early warning module and cloud platform are connected electrically and / or by communication.

Images