Waistband-type falling protection early warning device for elderly people living alone

Abstract

The invention discloses a waistband type falling protection early warning device for old people living alone, and belongs to the technical field of intelligent wearable equipment. The problems that an existing device is single in early warning mode and cannot effectively guarantee the safety of old people living alone are solved, corresponding action results are matched for various data of the old people through the convolutional neural network and the lightweight decision tree model, and if a preliminary judgment result conforms to a normal mode, a judgment result is directly output; if not, triggering a secondary analysis mechanism to further judge whether the old people fall down or not; the GPS positioner is also used for analyzing whether falling behaviors occur at the same position for many times, and if a high-risk position is found, when the old people approach the high-risk position, the warning lamp and the alarm are used for reminding and paying attention to action behaviors; the fall detection accuracy is improved, the possibility of false alarm and missing alarm is reduced, the old people can be actively prevented from falling again in a high-risk area, and a more comprehensive safety guarantee is provided for the old people living alone.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent wearable devices, in particular to a waistband type fall protection and early warning device for the elderly living alone. BACKGROUND

[0002] Fall is one of the common accidents for the elderly living alone, which may cause serious injury or even endanger life. There are some fall detection devices on the market at present, which mainly monitor the motion state of the elderly through acceleration sensors, gyroscopes and other motion sensors. However, the existing fall warning devices mostly have the following problems: 1. The existing fall detection devices mostly adopt the form of wristband or ankle ring, which are worn on the wrist or ankle. This wearing method not only affects the normal activities of the elderly, but also may cause discomfort, reducing the willingness of the elderly to use the device. 2. The existing fall detection devices can only issue an alarm through a simple buzzer or vibration after detecting a fall event, which cannot notify the guardian or emergency rescue agencies in time, and it is difficult to effectively protect the safety of the elderly.

[0003] Therefore, the existing needs are not met, and for this purpose, we propose a waistband type fall protection and early warning device for the elderly living alone. SUMMARY

[0004] The purpose of the present application is to provide a waistband type fall protection and early warning device for the elderly living alone, which matches the corresponding motion results for the various data of the elderly through convolutional neural network and lightweight decision tree model. If the preliminary judgment result conforms to the normal mode, the judgment result is directly output. If it does not conform, the secondary analysis mechanism is triggered to further judge whether the elderly have a fall behavior. It also analyzes whether there are multiple fall behaviors in the same position through the GPS locator. If a high-risk position is found, the warning light and alarm are used to remind the action behavior when the elderly approach the high-risk position. Not only does it improve the accuracy of fall detection and reduce the possibility of false positives and false negatives, but it also actively prevents the elderly from falling again in high-risk areas, providing more comprehensive safety protection for the elderly living alone, and solving the problems raised in the above background technology.

[0005] To achieve the above purpose, the present application provides the following technical solutions: A belt-like fall protection and warning device for elderly people living alone, comprising: a belt body, with a warning light and alarm provided on the outside of the belt body; a microprocessor A provided on the inside of the belt body; and a motion sensor B for real-time monitoring of the elderly person's behavior. The motion sensor B comprises an accelerometer and a gyroscope capable of accurately acquiring information on the elderly person's motion acceleration, angular velocity, and altitude change; a pressure sensor C for acquiring atmospheric pressure signals within the elderly person's exercise environment; a pressure sensor D for monitoring changes in pressure distribution around the elderly person's waist; and a GPS locator E for monitoring the elderly person's location. The microprocessor A includes: a primary analysis module and a secondary analysis module; The primary analysis module is configured to build a convolutional neural network, a lightweight decision tree model, and a long short-term memory network, and extract features from the elderly person's motion data, pressure distribution data, and air pressure mutation data through the convolutional neural network, and output the confidence level of the current action; according to the predefined rules and features of the lightweight decision tree model, the corresponding action result is matched for the confidence level of the current action; The secondary analysis module is configured to further determine whether the elderly person has fallen through the temporal dependency of learning features through a long-short-term memory network.

[0006] Furthermore, after the first-level analysis module matches the corresponding action result for the confidence of the current action, it also includes: Based on the historical training results of the convolutional neural network and lightweight decision tree model, the confidence threshold is preset; If the output confidence is within the threshold range and the pressure distribution and air pressure changes conform to the normal pattern, the current judgment result will be used as the final confirmation result; If the output confidence is lower than the threshold range, or the pressure distribution and air pressure changes do not conform to the normal pattern, the secondary analysis mechanism is triggered; After analyzing the elderly's motion data, pressure distribution data, and air pressure mutation data through the secondary analysis module, if multiple models support fall judgment, it will eventually be confirmed as a fall event and an alarm will be triggered.

[0007] Furthermore, the secondary analysis module includes: The early warning analysis module is configured to collect the locations where the elderly have fallen through the GPS locator E, and analyze whether there are multiple falls at the same location; if so, the location is marked as a high-risk location and fed back to the microprocessor A; The early warning and protection module is configured to obtain the location information of the elderly in real time through the GPS locator E, and when the elderly are close to a high-risk location, the warning light and alarm will remind the elderly to pay attention to their actions.

[0008] Furthermore, marking the location as a high-risk location includes: Real-time collection of fall event data through GPS locator E; For each fall event i Perform three-element representation, i.e., e i =(t i ,lat i ,lon i ), where t i Indicates a fall event i The corresponding event time; lat i Indicates a fall event i Corresponding latitude coordinate; lon i Indicates a fall event i The corresponding longitude coordinate; Real-time monitoring of the translation distance between each fall event and the previous fall event, as well as the walking distance between each fall event and the elderly person's current starting position; The dynamic risk threshold is set using the translation distance between each fall event and the previous fall event and the walking distance between each fall event and the starting position of the elderly person's current walking; The dynamic risk threshold is obtained by the following formula: Among them, Q represents the dynamic risk threshold; L p represents the average translation distance between each fall event and its previous fall event; n represents the number of historical fall events; L i+1 L represents the distance between the fall event corresponding to the i+1th fall and the starting position of the elderly person’s current walking; i represents the distance between the fall event corresponding to the i-th fall and the starting position of the elderly person’s current walking; Using each fall event e i The corresponding three elements set the risk density corresponding to the current fall event; Compare the risk density corresponding to the current fall event with the preset dynamic risk threshold; When the current fall event is multiple falls at the same location, and the corresponding risk density is not lower than a preset dynamic risk threshold, the location is marked as a high-risk location.

[0009] Further, the use of each fall event e i The corresponding three elements set the risk density corresponding to the current fall event, including: Retrieve each fall eventi The corresponding three elements; Using each fall event e i The corresponding three elements obtain the average risk density corresponding to historical fall events; The average risk density corresponding to the historical fall events is obtained by the following formula: Among them, Y(p, t) represents the average risk density corresponding to historical fall events; n represents the number of historical fall events; α i represents the confidence weight corresponding to the i-th historical fall event; β represents the spatial sensitivity parameter, which ranges from 0.001 to 0.005 and is used to control the decay rate of risk with distance; d(p, p i ) represents the location p of the elderly person’s home and the i-th historical fall event p i The spherical distance between each fall event e i The corresponding longitude coordinate lon i and the latitude coordinate lat i The Haversine formula is used to calculate the longitude coordinate lon and latitude coordinate lat of the elderly person's home location p; r represents the time decay rate parameter, which is 0.4-1.3; t i represents the time corresponding to the i-th historical fall event; t represents the current time; Retrieve the three elements corresponding to the current fall event; The risk density corresponding to the current fall event is obtained by combining the three elements corresponding to the current fall event with the average risk density corresponding to the historical fall events; The risk density corresponding to the current fall event is obtained by the following formula: Among them, Y (p d , t) represents the risk density corresponding to the current fall event; α d represents the confidence weight corresponding to the current fall event; d(p, p d ) represents the location p of the elderly at home and the current fall event p i The spherical distance between the two is calculated using the longitude coordinate lon corresponding to the current fall event. d and the latitude coordinate lat d Combine the longitude coordinate lon and latitude coordinate lat of the elderly person’s home location p and use the Haversine formula to calculate; t d Indicates the time corresponding to the current fall event.

[0010] Furthermore, the primary analysis module includes: a data collection module and a data clustering module; The data collection module is configured to collect the volunteers' daily motion data, falling motion data, and pressure mutation data, and preset the elderly's falling pressure threshold and normal pressure threshold; The data clustering module is configured to cluster and divide the fall pressure threshold and the normal pressure threshold, and their corresponding daily action data, fall action data, and pressure mutation data based on a lightweight decision tree model.

[0011] Furthermore, the primary analysis module further includes: a data partitioning module and a model verification module; A data partitioning module is configured to divide the volunteer data into training sets and test sets, and use the training sets and test sets to train and test the convolutional neural network and long short-term memory network; The model verification module is configured to compare the test results of the convolutional neural network and the long short-term memory network with the actual action information of the volunteers, and verify whether the results output by the convolutional neural network and the long short-term memory network are accurate; if accurate, the convolutional neural network and the long short-term memory network are saved and put into actual use.

[0012] Furthermore, the waist belt body is composed of a rear belt body, a front strap, a support member and a tension belt. The two ends of the rear belt body are integrated with the front strap. A plurality of support members for supporting the waist muscles are arranged side by side on the inner side of the rear belt body. Two tension belts are symmetrically sewn between the support member in the middle and the rear belt body. The other ends of the two tension belts pass through the inside of the front strap and extend to the outside of the front strap. Velcro is respectively provided on the inner and outer sides of the front strap. The Velcro on the inner side is used to stick the front straps at both ends, and the Velcro on the outer side is used to stick the two tension belts.

[0013] Furthermore, the microprocessor A further comprises: a device integration module, an information setting module and a data storage module; The device integration module is configured to associate the microprocessor A built into the waistband body with the guardian's mobile terminal via wireless communication technology; An information setting module is configured to input personal information of the elderly person wearing the belt into the microprocessor A based on the guardian's mobile terminal; The data storage module is configured to store the motion data, pressure data, and air pressure mutation data of the elderly in time series, which is used to regularly optimize the learning performance of convolutional neural networks, lightweight decision tree models, and long short-term memory networks.

[0014] Furthermore, the microprocessor A further comprises: a remote assistance module and a power supply module; The remote assistance module is configured to send the fall information and the location information detected by the GPS locator E to a preset contact person or emergency rescue agency through wireless communication technology after determining that the elderly person has fallen; The power module is configured to use a rechargeable battery to provide stable power support for the belt body and is equipped with a low-battery reminder function.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. In the present invention, a convolutional neural network is used to extract features from the elderly's motion data, pressure distribution data, and air pressure mutation data, and the confidence of the current action is output. The corresponding action result is quickly matched in combination with a lightweight decision tree model; if the preliminary judgment result meets the confidence threshold and the normal mode, the judgment result is directly output; if the confidence is low or the pressure distribution and air pressure changes do not meet the normal mode, the secondary analysis mechanism is triggered to further judge whether the elderly have fallen; thereby ensuring that fall events can be reliably detected in complex situations, improving the accuracy of fall detection for the elderly, reducing the possibility of false alarms and missed alarms, and providing strong protection for the safety of elderly people living alone.

[0016] 2. In the present invention, the places where the elderly fall are collected through the GPS locator, and it is analyzed whether there are multiple falls in the same location; if a high-risk location is found, when the elderly are close to the high-risk location, warning lights and alarms are used to remind them to pay attention to the action; thereby, it is possible to actively prevent the elderly from falling again in high-risk areas, effectively reduce the risk of falling, enhance the protective function of the early warning device, and provide more comprehensive safety protection for elderly people living alone. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 This is a structural diagram of the belt-style fall protection and warning device for elderly people living alone of the present invention; Figure 2 This is an inside view of the belt-style fall protection and warning device for elderly people living alone of the present invention; Figure 3 This is a fall protection warning flow chart of the present invention.

[0018] In the figure: 1. Belt body; 11. Warning light; 12. Siren; 2. Rear belt body; 3. Front strap; 4. Support member; 5. Tension belt. DETAILED DESCRIPTION

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0020] To address the technical issues that most existing fall detection devices are in the form of wristbands or anklets, which are not comfortable to wear; and after detecting a fall, they can only issue an alarm through a simple beep or vibration, which cannot notify guardians or emergency rescue agencies in time, making it difficult to effectively protect the safety of elderly people living alone, please refer to Figure 1-Figure 3 , this embodiment provides the following technical solutions: The belt-type fall protection warning device for elderly people living alone includes: a belt body 1, which is the main part of the entire fall protection warning device. In order to consider the comfort and functional integrity of the elderly, the belt body 1 is composed of a rear belt body 2, a front strap 3, a support member 4 and a tension belt 5, so that the belt body 1 can fit closely to the waist and provide necessary support and stability; a warning light 11 and an alarm 12 are set on the outside of the belt body 1, which emit obvious light and sound signals when a fall risk is detected or a fall event is confirmed, so as to promptly remind the elderly to pay attention to safety and notify the guardian or emergency rescue agency; the rear belt body 2 is the main supporting part of the belt, and its two ends are connected to the front belt 3. The straps 3 are connected as a whole to form a complete waist belt structure; a plurality of support members 4 for supporting waist muscles are arranged in parallel on the inner side of the rear belt body 2, which are made of soft and elastic material to support waist muscles, reduce waist burden and improve wearing comfort; two tension belts 5 are symmetrically sewn between the support member 4 in the middle and the rear belt body 2, and the other ends of the two tension belts 5 pass through the inside of the front strap 3 and extend to the outside of the front strap 3, which not only enhances the stability of the waist belt body 1, but also enables the waist belt body 1 to better adapt to the elderly of different body shapes; Velcro is provided on the inner and outer sides of the front strap 3 respectively, and the Velcro on the inner side is used to connect the front straps 3 at both ends The belt body 1 is pasted to ensure that it can fit the waist tightly; the Velcro on the outside is used to paste two tension belts 5. By adjusting the tightness of the tension belts 5, the fit of the belt body 1 can be further optimized to adapt to the waist sizes of different elderly people and improve the comfort and stability of wearing; the inside of the belt body 1 is provided with a microprocessor A and a motion sensor B for real-time monitoring of the behavior status of the elderly. The motion sensor B includes: an acceleration sensor and a gyroscope, which can accurately obtain the elderly's motion acceleration, angular velocity and height change information; a pressure sensor C for obtaining the atmospheric pressure signal in the elderly's exercise environment. For example: when the elderly fall from a height, the air pressure will rise rapidly. High, by monitoring the changes in air pressure, abnormal situations can be discovered in time; the pressure sensor D is used to monitor the changes in the pressure distribution on the waist of the elderly. For example, when an elderly person falls, the pressure distribution on the waist will change dramatically. By monitoring these changes, the fall behavior can be discovered in time, improving the accuracy and reliability of fall detection; the GPS locator E is used to monitor the location of the elderly. By obtaining the elderly's location information in real time, the guardian or emergency rescue agency can be notified in time when a fall occurs, and accurate geographic location information can be provided; in addition, the GPS locator E can also be used to analyze the elderly's activity range and behavior patterns, plan high-risk locations in advance, and prevent the occurrence of falls.

[0021] Microprocessor A includes: a primary analysis module and a secondary analysis module; the primary analysis module includes: a data collection module, a data clustering module, a data partitioning module and a model verification module; the secondary analysis module includes: an early warning analysis module and an early warning protection module.

[0022] The data collection module is configured to collect volunteers' daily movement data, fall movement data, and pressure mutation data, and preset the fall pressure threshold and normal pressure threshold of the elderly as data samples for training convolutional neural networks and long short-term memory networks.

[0023] The data clustering module is configured to cluster the fall pressure threshold and normal pressure threshold, as well as their corresponding daily motion data, fall motion data, and pressure mutation data, based on a lightweight decision tree model. Through cluster analysis, the characteristic distribution of different motion patterns can be better understood.

[0024] The data partitioning module is configured to divide the volunteers' data into training sets and test sets, and train and test the convolutional neural network and long short-term memory network through the training sets and test sets to ensure that the convolutional neural network and long short-term memory network perform well on different data sets.

[0025] The model verification module is configured to compare the test results of the convolutional neural network and the long short-term memory network with the actual action information of the volunteers, respectively, to verify whether the results output by the convolutional neural network and the long short-term memory network are accurate; if accurate, the convolutional neural network and the long short-term memory network are saved and put into actual use to ensure that the belt-type fall protection warning device for elderly people living alone can effectively detect the fall behavior of the elderly, while reducing false alarms and missed alarms, and providing strong protection for the safety of the elderly.

[0026] The first-level analysis module is configured to build a convolutional neural network, a lightweight decision tree model and a long short-term memory network. The convolutional neural network is used to extract features from the elderly's motion data, pressure distribution data and air pressure mutation data, and output the confidence of the current action; and based on the historical training results of the convolutional neural network and the lightweight decision tree model, a confidence threshold is preset; for example: the confidence threshold range is set to ≤65%; the corresponding action result is matched for the confidence of the current action through the predefined rules and features of the lightweight decision tree model; for example: the convolutional neural network outputs a "normal sitting down" action with a confidence of 85%, and the decision tree model matches the "normal sitting down" action result; then the action is "normal sitting down" and there is no need to start the second-level analysis module; if the output confidence is within the threshold range and the pressure distribution and air pressure changes conform to the normal pattern, the current judgment result is used as the final confirmation result; if the output confidence is lower than the threshold range, or the pressure distribution and air pressure changes do not conform to the normal pattern, the second-level analysis mechanism is triggered.

[0027] The secondary analysis module is configured to learn the temporal dependency of features through a long-short-term memory network. For example, the long-short-term memory network can capture the changing trends of acceleration and angular velocity over a period of time, thereby further determining whether the elderly have fallen. If multiple models support fall judgment, it is ultimately confirmed as a fall event and an alarm is triggered. For example, the convolutional neural network outputs a "normal sitting down" action with a confidence level of 60%, and the decision tree model matches the "suspicious action" result. The long-short-term memory network outputs a fall behavior with a confidence level of 90%, combined with abnormal pressure distribution and air pressure changes, and ultimately confirmed as a fall event, triggering an alarm to notify the preset contact or emergency rescue agency.

[0028] The beneficial effects achieved by the above content are: through the convolutional neural network, the motion data, pressure distribution data and air pressure mutation data of the elderly are feature extracted, the confidence of the current action is output, and the corresponding action results are quickly matched in combination with the lightweight decision tree model; if the preliminary judgment result meets the confidence threshold and the normal mode, the judgment result is directly output; if the confidence is low or the pressure distribution and air pressure changes do not conform to the normal mode, the secondary analysis mechanism is triggered to further judge whether the elderly have fallen; thereby ensuring that fall events can be reliably detected in complex situations, improving the accuracy of fall detection for the elderly, reducing the possibility of false alarms and missed alarms, and providing strong protection for the safety of elderly people living alone.

[0029] The early warning analysis module is configured to collect the locations where the elderly have fallen through the GPS locator E and analyze whether there are multiple falls at the same location. If so, the location is marked as a high-risk location and the feedback is fed back to the microprocessor A. The early warning and protection module is configured to obtain the location information of the elderly in real time through the GPS locator E, compare the real-time location information with the stored high-risk location information, and when the elderly are close to the high-risk location, the warning light 11 flashes a light of a specific color and the alarm 12 sounds an alarm to remind the elderly to pay attention to the danger of the current location and pay attention to their actions and behaviors; and provide personalized early warning prompts based on the elderly’s personal behavioral habits and historical fall data; for example: for an elderly person who has fallen multiple times in the bathroom, the system can issue more frequent early warning signals when they enter the bathroom.

[0030] The beneficial effects achieved by the above content are: collecting the locations where the elderly fall through GPS locators, and analyzing whether there are multiple falls in the same location; if a high-risk location is found, warning lights and alarms are used to remind them to pay attention to the action when they are close to the high-risk location; thereby actively preventing the elderly from falling again in high-risk areas, effectively reducing the risk of falling, enhancing the protective function of the early warning device, and providing more comprehensive safety protection for elderly people living alone.

[0031] Specifically, marking the location as a high-risk location includes: Real-time collection of fall event data through GPS locator E; For each fall event i Perform three-element representation, i.e., e i =(t i ,lat i ,lon i ), where t i Indicates a fall event i The corresponding event time; lat i Indicates a fall event i Corresponding latitude coordinate; lon i Indicates a fall event i The corresponding longitude coordinate; Real-time monitoring of the translation distance between each fall event and the previous fall event, as well as the walking distance between each fall event and the elderly person's current starting position; The dynamic risk threshold is set using the translation distance between each fall event and the previous fall event and the walking distance between each fall event and the starting position of the elderly person's current walking; The dynamic risk threshold is obtained by the following formula: Among them, Q represents the dynamic risk threshold; L p represents the average translation distance between each fall event and its previous fall event; n represents the number of historical fall events; L i+1 L represents the distance between the fall event corresponding to the i+1th fall and the starting position of the elderly person’s current walking; i represents the distance between the fall event corresponding to the i-th fall and the starting position of the elderly person’s current walking; Using each fall event e i The corresponding three elements set the risk density corresponding to the current fall event; Compare the risk density corresponding to the current fall event with the preset dynamic risk threshold; When the current fall event is multiple falls at the same location, and the corresponding risk density is not lower than a preset dynamic risk threshold, the location is marked as a high-risk location.

[0032] The technical solution described above achieves the following: GPS locators collect data such as the time, latitude, and longitude of fall events (represented by three elements), while also factoring in the distance traveled between fall events and the distance traveled from the starting location. This allows for a multi-dimensional characterization of fall scenarios from both time and space. Calculating dynamic risk thresholds and risk densities based on this data more accurately captures location risk characteristics. Compared to single-dimensional analysis, this approach can avoid misjudgments and missed detections, making high-risk location identification more relevant to real-world scenarios. For example, it can distinguish between occasional fall locations and truly high-risk locations prone to recurrent falls. The dynamic risk threshold formula incorporates the number of historical fall events, the average distance traveled, and the distance difference between different fall events and the starting location, allowing it to adaptively adjust as fall event data changes. Unlike fixed thresholds, which can be inaccurate due to variations in scenario and elderly individuals' walking habits, this approach better adapts to the identification of fall-risk locations across diverse elderly individuals' daily activities (e.g., varying walking routes and activity ranges), improving the accuracy of high-risk location identification. After accurately identifying high-risk locations, targeted protective measures can be implemented based on this information. For example, adding handrails, anti-slip mats, and warning signs can be used to proactively eliminate or mitigate fall hazards, reducing the likelihood of elderly people falling again in high-risk areas. This environmental intervention ensures the safety of elderly people in activities such as walking, and mitigates the likelihood of fall injuries. Because the dynamic calculation and assessment of fall events is based on real-time data, changes in location risk can be continuously tracked. If falls decrease at high-risk locations after protection measures are implemented, and data such as risk density change, thresholds and assessments will be adjusted accordingly. This enables dynamic monitoring and optimization of risk prevention and control measures at risk locations, forming a closed-loop management process of "identification-protection-reassessment-reoptimization" to ensure a safe environment for elderly people's activities over the long term. The continuously collected data on the three elements of fall events, as well as the translation distance and travel distance involved in the calculation process, can be accumulated into a fall risk database. As this data accumulates, the parameters in the formula (such as the number of historical fall events, n) and the calculation logic can be continuously optimized based on actual feedback. For example, if it is found that the threshold calculation in certain scenarios is not sufficiently adapted to the risk density, the weights of the elements in the formula can be adjusted or new influencing factors can be introduced to allow the entire high-risk location identification algorithm to be continuously iterated and upgraded, thereby improving the intelligence level and application effect, and providing more efficient and accurate technical means for fall prevention and control scenarios in smart elderly care.

[0033] Specifically, the use of each fall event e i The corresponding three elements set the risk density corresponding to the current fall event, including: Retrieve each fall event i The corresponding three elements; Using each fall event ei The corresponding three elements obtain the average risk density corresponding to historical fall events; The average risk density corresponding to the historical fall events is obtained by the following formula: Among them, Y(p, t) represents the average risk density corresponding to historical fall events; n represents the number of historical fall events; α i represents the confidence weight corresponding to the i-th historical fall event; β represents the spatial sensitivity parameter, which ranges from 0.001 to 0.005 and is used to control the decay rate of risk with distance; d(p, p i ) represents the location p of the elderly person’s home and the i-th historical fall event p i The spherical distance between each fall event e i The corresponding longitude coordinate lon i and latitude coordinate lat i The Haversine formula is used to calculate the longitude coordinate lon and latitude coordinate lat of the elderly person's home location p; r represents the time decay rate parameter, which is 0.4-1.3; t i represents the time corresponding to the i-th historical fall event; t represents the current time; specifically, and Simulate the propagation and attenuation characteristics of fall risk in reality. Spatially, the greater the physical distance, the lower the "risk correlation" of historical falls to the current location. For example, in a home environment, the risk impact of a fall near the bedroom door on the living room area decreases rapidly with increasing distance. This spatial correlation decay is quantitatively reflected through an exponential decay function, ensuring that risk calculations conform to the actual laws of spatial propagation. Temporally, falls that occurred too long ago have less reference value for current risk (for example, falls from several months ago have less guiding significance for current risk due to changes in the environment and the elderly person's condition). Using a time decay function to simulate the "expiration" of risk over time allows the average risk density to focus more on the impact of recent and nearby falls, meeting the dynamic changes in risk in engineering scenarios.

[0034] Retrieve the three elements corresponding to the current fall event; The risk density corresponding to the current fall event is obtained by combining the three elements corresponding to the current fall event with the average risk density corresponding to the historical fall events; The risk density corresponding to the current fall event is obtained by the following formula: Among them, Y (p d , t) represents the risk density corresponding to the current fall event; α d represents the confidence weight corresponding to the current fall event; d(p, p d) represents the location p of the elderly at home and the current fall event p i The spherical distance between the two is calculated using the longitude coordinate lon corresponding to the current fall event. d and the latitude coordinate lat d Combine the longitude coordinate lon and latitude coordinate lat of the elderly person’s home location p and use the Haversine formula to calculate; t d表示 The moment corresponding to the current fall event. Specifically, and This method emphasizes the immediate impact of a fall event. Based on the historical average risk density, it separately calculates the spatiotemporal impact of the current fall, reflecting the principle of "new events supplement and update risk perception." For example, if an elderly person has just fallen in a certain location, the impact of this event on the current risk assessment needs to be considered separately. This formula quantitatively calculates the immediate risk contribution in spatial (distance from the current assessment location) and temporal (distance from the current moment), providing a real-time response to dynamic risks.

[0035] The technical effect of the above technical solution is that the existing technology often does not take into account the temporal and spatial correlation of fall risk, and most of them simply count the fall location or a single time dimension feature. However, this solution introduces the attenuation function of spatial distance (spherical distance) and time difference to dynamically incorporate the temporal and spatial impact of historical fall events into the risk density calculation. For example, in space, the closer the historical fall point is to the current position, the greater the impact on the current risk assessment (because It decays rapidly with increasing distance; in terms of time, the closer the historical fall is to the current moment, the higher the weight ( The increase and decay with time difference) can more accurately fit the law of temporal and spatial changes of fall risks in actual scenarios, improve the accuracy of risk identification, and make the judgment of high-risk locations more scientific. Dynamic update and adaptive adjustment calculate the average risk density based on historical fall events, and then update the current risk density based on the current fall event. It is a dynamic iterative process. Different from the static risk assessment model of existing technologies (such as the risk level of a fixed location remains unchanged for a long time), risk awareness can be continuously updated as the elderly move around and the situation of falls changes. For example, if the elderly have been active in a certain area for a long time and new fall events are constantly added, the risk density can reflect the latest risk situation in the area in real time, and have better adaptability to environmental changes (such as changes in home facilities that lead to changes in fall risks), helping to promptly discover new high-risk points or adjust existing risk location judgments. The confidence weight is introduced to distinguish event differences. i (historical fall event confidence weight) and α d(Current fall event confidence weight) can be used to assign different importance to different fall events. Existing technologies often treat all fall events equally, ignoring differences in the events themselves (for example, a serious fall should be given more weight than a minor stumble). In practical applications, weights can be set based on factors such as the severity of the fall injury and the reliability of the monitoring equipment. For example, a high α value can be set for serious falls confirmed by a hospital to increase the impact of these key events on risk density, making risk assessments more accurate and aligning with the actual level of harm, and optimizing the identification of high-risk locations.

[0036] The microprocessor A also includes: a device integration module, an information setting module, a data storage module, a remote help module and a power supply module.

[0037] The device integration module is configured to associate the dedicated application of the microprocessor A built into the belt body 1 with the guardian's mobile terminal through wireless communication technology. The microprocessor A synchronizes the sensor data collected in real time to the guardian's mobile terminal. The guardian can view the elderly's real-time location, movement status and fall warning information through the mobile terminal, so that the guardian can understand the elderly's safety status in real time and take corresponding measures when necessary.

[0038] The information setting module is configured to enter the personal information of the elderly wearing the belt body 1 into the microprocessor A based on the guardian's mobile terminal, such as name, age, gender, health status, emergency contact information, etc.

[0039] The data storage module is configured to store the motion data, pressure data, and air pressure mutation data of the elderly in time series, which is used to regularly optimize the learning performance of convolutional neural networks, lightweight decision tree models, and long short-term memory networks.

[0040] The remote assistance module is configured to send the fall information and the positioning information detected by the GPS locator E to a preset contact or emergency rescue agency through wireless communication technology such as Bluetooth, Wi-Fi or mobile communication network after determining that the elderly person has fallen, so that the contact or emergency rescue agency can take quick action to ensure that the elderly living alone receive timely assistance and reduce the severity of the accident.

[0041] The power module is configured to use a rechargeable battery, such as a lithium-ion battery, to provide stable power support for the belt body 1 and is equipped with a low-battery reminder function.

[0042] Working principle: Convolutional neural networks and lightweight decision tree models are used to match the corresponding action results for various data of the elderly. If the preliminary judgment result conforms to the normal pattern, the judgment result is directly output; if not, the secondary analysis mechanism is triggered to further determine whether the elderly have fallen. The GPS locator is also used to analyze whether there have been multiple falls in the same location. If a high-risk location is found, warning lights and alarms are used to remind the elderly to pay attention to the action when they approach the high-risk location. This not only improves the accuracy of fall detection, but also actively prevents the elderly from falling again in high-risk areas, providing more comprehensive safety protection for elderly people living alone.

[0043] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "including," "having," or any other variations thereof are intended to cover non-exclusive possessors, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or includes elements that are inherent to such process, method, article, or apparatus.

[0044] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions, and alterations may be made to the embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A belt-type fall protection warning device for elderly people living alone, including: A waist belt body (1), characterized in that: a warning light (11) and an alarm (12) are provided on the outside of the waist belt body (1), a microprocessor A is provided on the inside of the waist belt body (1), and a motion sensor B for real-time monitoring of the behavior state of the elderly, a pressure sensor C for obtaining an atmospheric pressure signal in the elderly's exercise environment, a pressure sensor D for monitoring changes in the elderly's waist pressure distribution, and a GPS locator E for monitoring the elderly's location; The microprocessor A includes: a primary analysis module and a secondary analysis module; The primary analysis module is configured to build a convolutional neural network, a lightweight decision tree model, and a long short-term memory network, and extract features from the elderly person's motion data, pressure distribution data, and air pressure mutation data through the convolutional neural network, and output the confidence level of the current action; according to the predefined rules and features of the lightweight decision tree model, the corresponding action result is matched for the confidence level of the current action; The secondary analysis module is configured to further determine whether the elderly person has fallen through the temporal dependency of learning features through a long-short-term memory network.

2. The belt-type fall protection warning device for elderly people living alone according to claim 1 is characterized in that: After the first-level analysis module matches the corresponding action result for the confidence of the current action, it also includes: Based on the historical training results of the convolutional neural network and lightweight decision tree model, the confidence threshold is preset; If the output confidence is within the threshold range and the pressure distribution and air pressure changes conform to the normal pattern, the current judgment result will be used as the final confirmation result; If the output confidence is lower than the threshold range, or the pressure distribution and air pressure changes do not conform to the normal pattern, the secondary analysis mechanism is triggered; After analyzing the elderly's motion data, pressure distribution data, and air pressure mutation data through the secondary analysis module, if multiple models support fall judgment, it will eventually be confirmed as a fall event and an alarm will be triggered.

3. The belt-type fall protection warning device for elderly people living alone according to claim 1 is characterized in that: The secondary analysis module includes: The early warning analysis module is configured to collect the locations where the elderly have fallen through the GPS locator E, and analyze whether there are multiple falls at the same location; if so, the location is marked as a high-risk location and fed back to the microprocessor A; The early warning and protection module is configured to obtain the location information of the elderly in real time through the GPS locator E, and when the elderly are close to a high-risk location, the warning light (11) and the alarm (12) are used to remind the elderly to pay attention to their movements and behaviors.

4. The belt-type fall protection warning device for elderly people living alone according to claim 3 is characterized in that: Marking the location as a high-risk location includes: Real-time collection of fall event data through GPS locator E; For each fall event i Perform three-element representation, i.e., e i =(t i ,lat i ,lon i ), where t i Indicates a fall event i The corresponding event time; lat i Indicates a fall event i Corresponding latitude coordinate; lon i Indicates a fall event i The corresponding longitude coordinate; Real-time monitoring of the translation distance between each fall event and the previous fall event, as well as the walking distance between each fall event and the elderly person's current starting position; The dynamic risk threshold is set using the translation distance between each fall event and the previous fall event and the walking distance between each fall event and the starting position of the elderly person's current walking; Using each fall event e i The corresponding three elements set the risk density corresponding to the current fall event; Compare the risk density corresponding to the current fall event with the preset dynamic risk threshold; When the current fall event is multiple falls at the same location, and the corresponding risk density is not lower than a preset dynamic risk threshold, the location is marked as a high-risk location.

5. The belt-type fall protection warning device for elderly people living alone according to claim 4 is characterized in that: The use of each fall event e i The corresponding three elements set the risk density corresponding to the current fall event, including: Retrieve each fall event i The corresponding three elements; Using each fall event e i The corresponding three elements obtain the average risk density corresponding to historical fall events; Retrieve the three elements corresponding to the current fall event; The risk density corresponding to the current fall event is obtained by combining the three elements corresponding to the current fall event with the average risk density corresponding to the historical fall events.

6. The belt-type fall protection warning device for elderly people living alone according to claim 2 is characterized in that: The primary analysis module includes: a data collection module and a data clustering module; The data collection module is configured to collect the volunteers' daily motion data, falling motion data, and pressure mutation data, and preset the elderly's falling pressure threshold and normal pressure threshold; The data clustering module is configured to cluster and divide the fall pressure threshold and the normal pressure threshold, and their corresponding daily action data, fall action data, and pressure mutation data based on a lightweight decision tree model.

7. The belt-type fall protection warning device for elderly people living alone according to claim 6 is characterized in that: The primary analysis module further includes: a data partitioning module and a model verification module; A data partitioning module is configured to divide the volunteer data into training sets and test sets, and use the training sets and test sets to train and test the convolutional neural network and long short-term memory network; The model verification module is configured to compare the test results of the convolutional neural network and the long short-term memory network with the actual action information of the volunteers, and verify whether the results output by the convolutional neural network and the long short-term memory network are accurate; if accurate, the convolutional neural network and the long short-term memory network are saved and put into actual use.

8. The belt-type fall protection warning device for elderly people living alone according to claim 1 is characterized in that: The waist belt body (1) is composed of a rear belt body (2), a front strap (3), a support member (4) and a tension band (5). The two ends of the rear belt body (2) are integrated with the front strap (3). A plurality of support members (4) for supporting waist muscles are arranged in parallel on the inner side of the rear belt body (2). Two tension bands (5) are symmetrically sewn between the support member (4) and the rear belt body (2) in the middle. The other ends of the two tension bands (5) respectively pass through the inside of the front strap (3) and extend to the outside of the front strap (3). Velcro is respectively provided on the inner and outer sides of the front strap (3). The Velcro on the inner side is used to stick the front straps (3) at both ends, and the Velcro on the outer side is used to stick the two tension bands (5).

9. The belt-type fall protection warning device for elderly people living alone according to claim 1 is characterized in that: The microprocessor A further comprises: a device integration module, an information setting module and a data storage module; A device integration module configured to associate the microprocessor A built into the belt body (1) with the guardian's mobile terminal via wireless communication technology; An information setting module is configured to input personal information of an elderly person wearing the waist belt body (1) into the microprocessor A based on a guardian's mobile terminal; The data storage module is configured to store the motion data, pressure data, and air pressure mutation data of the elderly in time series, which is used to regularly optimize the learning performance of convolutional neural networks, lightweight decision tree models, and long short-term memory networks.

10. The belt-type fall protection warning device for elderly people living alone according to claim 9 is characterized in that: The microprocessor A further includes: a remote assistance module and a power supply module; The remote assistance module is configured to send the fall information and the location information detected by the GPS locator E to a preset contact person or emergency rescue agency through wireless communication technology after determining that the elderly person has fallen; The power module is configured to use a rechargeable battery to provide power support for the belt body (1) and is equipped with a low-battery reminder function.

Images