A posture sensing type alarm call method and device for a bedside apparatus

Abstract

The application discloses a posture sensing type alarm calling method and device of a bedside device, and relates to the field of intelligent health monitoring.The method comprises the following steps: acquiring posture information data of a user; preprocessing the posture information data, and constructing posture safety features according to the preprocessed posture information data; performing threshold value judgment according to the sleep posture stability features, the sitting posture stability features and the posture mutation detection features; when the preset safety threshold condition is not met, performing safety early warning; when the preset safety threshold condition is met, calculating posture safety scores; performing threshold value judgment according to the posture safety scores and preset posture safety scores; and when the posture safety scores are higher than the preset posture safety score threshold values, performing safety early warning.Through multi-dimensional sensor data fusion and comprehensive evaluation of user posture features, the safety state of the user is monitored in real time to identify falling risks, false positives are reduced, and the accuracy of safety early warning is improved.

Description

Technical Field

[0001] This invention relates to the field of intelligent health monitoring, specifically to a posture-sensing alarm call method and device for bedside equipment. Background Technology

[0002] With the increasing prevalence of an aging population, more and more elderly people, patients, and others with special needs require long-term home care or hospitalization. Due to the gradual decline in physical function, the elderly and some patients are prone to falls, especially at night when sleeping or getting out of bed. Falls can lead to serious injuries such as fractures, as well as other complications, and in severe cases, even death. Therefore, effective human posture detection and early warning mechanisms have become an important issue in the field of health monitoring.

[0003] Traditional human posture detection methods mostly rely on physical monitoring devices or manual care monitoring. Common human posture detection technologies include: video monitoring methods: using cameras to monitor patients' activities in real time. However, this method is limited by privacy issues and has poor monitoring performance in low-light environments or when patients are covered, making it easy to miss detections; sound monitoring methods: using sound sensors to detect the sounds generated when a fall occurs. However, this method is greatly affected by noisy environments and cannot effectively determine whether a patient is in a dangerous state or has an abnormal posture; touch and pressure sensing monitoring methods: installing pressure-sensitive pads, mattresses, and other sensors to detect whether a patient has fallen from a bed or chair. However, these devices lack flexibility, cannot cover all activity scenarios, and are easily affected by environmental changes.

[0004] These traditional methods cannot provide continuous, real-time, and efficient monitoring in complex and ever-changing daily environments, especially when patients are in bed or changing positions, often making it difficult to detect potential fall risks in a timely manner. With the continuous development of sensing technology, IoT technology, and artificial intelligence technology, sensor-based intelligent fall monitoring solutions are gradually emerging. By collecting user posture information in real time through multiple sensors (such as accelerometers, pressure sensors, and angle sensors), and combining this with intelligent algorithms to analyze and process the data, more accurate and efficient fall monitoring can be achieved. Some existing smart bedside devices and wearable devices can already monitor user activities, but these devices generally suffer from the following problems: Insufficient accuracy of data analysis: It fails to fully utilize multi-dimensional data, which can easily lead to false alarms or omissions.

[0005] Poor adaptability to different postures: Existing monitoring systems can usually only work effectively in certain specific postures, making it difficult to accurately judge the changes in a user's posture under different environments and activities.

[0006] System response lag: Alarm mechanisms in traditional systems often have a certain lag, failing to issue an alarm in a timely manner after a user falls unexpectedly. Summary of the Invention

[0007] Based on the shortcomings of the prior art described above, the purpose of this invention is to provide a posture-sensing alarm call method and device for bedside equipment to solve the above-mentioned technical problems.

[0008] To achieve the above objectives, the present invention provides the following technical solution: a posture-sensing alarm call method for a bedside device, comprising: The user's posture information data is obtained by detecting the user's posture information through sensors on the bedside equipment; The posture information data is preprocessed, and posture safety features are constructed based on the preprocessed posture information data. The posture safety features include: sleeping posture stability features, sitting posture stability features, and posture change detection features. Threshold judgments are made based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the detection characteristics of the posture change. When the preset safety threshold conditions are not met, a safety warning is issued. When the preset safety threshold condition is met, the posture safety score is calculated based on the sleep posture stability feature, the sitting posture stability feature and the posture change detection feature. A threshold is determined based on the attitude safety score and a preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

[0009] The present invention is further configured such that the attitude information data includes acceleration, angle, and pressure.

[0010] The present invention is further configured to construct attitude safety features based on the preprocessed attitude information data, including: The stability characteristics of the sleeping posture are calculated based on the acceleration, the angle, and the pressure. The stability characteristics of the sitting posture are calculated based on the acceleration, the angle, and the pressure. The posture change detection features are calculated based on the acceleration and the angle.

[0011] The present invention is further configured such that the calculation logic for the attitude safety score is as follows: ,in, To score safely, For the characteristics of sleep posture stability, For characteristics of sitting and standing stability, For the detection of posture mutation features, To obtain and The maximum value in These are the weighting coefficients.

[0012] The present invention is further configured to calculate an attitude safety score according to the calculation logic of the attitude safety score, compare the attitude safety score with a preset attitude safety score for threshold judgment, and issue a safety warning when the attitude safety score is higher than the preset attitude safety score threshold.

[0013] The present invention is further configured such that the calculation logic of the posture change detection feature is as follows: the sleep posture stability feature, the sitting posture stability feature and the posture change detection feature are respectively mapped to the corresponding discretized bin intervals; Maintain the historical distribution of discretized bin intervals within a sliding window of a preset length; Identify pose combinations that occur less frequently than the rare pattern threshold in the historical distribution, mark them as rare pose combinations, and trigger an alarm call.

[0014] The present invention is further configured such that the corresponding discretized binning interval mapping logic includes: For the sleep posture stability feature value collected at each moment, the preset sleep bin boundary sequence is traversed sequentially to locate the smallest index i, such that the i-th item in the boundary sequence is ≤ sleep posture stability feature value < i+1-th item, and the corresponding i-th bin interval is confirmed. The sitting posture stability feature value is iterated through the preset sitting bin boundary sequence to confirm the corresponding j-th bin interval. The pose mutation detection feature values ​​are iterated through the preset mutation bin boundary sequence to confirm the corresponding k-th bin interval; The binning index in the form of (i,j,k) triples is used as the mapping result.

[0015] The present invention is further configured to maintain the historical distribution of discretized bin intervals within a sliding window of a preset length, including: Set the maximum capacity L of the sliding window, and initialize a circular queue to store the discretized bin index triplets. Initialize a three-dimensional combination counting matrix Counts with size N×M×K, and set all elements to zero. Receive the latest mapped triple index (i,j,k), append (i,j,k) to the tail of the circular queue; and increment the corresponding Counts[i][j][k]. If the length of the circular queue exceeds L, pop the earliest enqueued triple (i0,j0,k0) from the front of the queue and decrement Counts[i0][j0][k0].

[0016] The present invention is further configured to identify pose combinations that occur less frequently than a rare pattern threshold in historical distributions, mark them as rare pose combinations, and trigger an alarm call, including: After updating Counts, check the latest mapping index triple (i,j,k) and obtain Counts[i][j][k] from the three-dimensional combination counting matrix; The frequency of the obtained combination is compared with the rare pattern threshold. If Counts[i][j][k] is less than the rare pattern threshold, the triple is marked as a rare pose combination. For triples marked as rare pose combinations, record the timestamp and original feature value, generate an alarm message, call the alarm module, and initiate an alarm call to a predefined receiver.

[0017] The present invention also provides a posture-sensing alarm call device for bedside equipment, the device comprising: Data acquisition module: used to acquire user posture information data by detecting user posture information through sensors on the bedside device; Feature construction module: used to preprocess the posture information data and construct posture safety features based on the preprocessed posture information data, wherein the posture safety features include: sleeping posture stability features, sitting posture stability features and posture change detection features; The first early warning module is used to make a threshold judgment based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the change in posture detection characteristics. When the preset safety threshold conditions are not met, a safety warning is issued. The scoring calculation module is used to calculate a posture safety score based on the sleeping posture stability feature, the sitting posture stability feature, and the posture change detection feature when the preset safety threshold condition is met. The second early warning module is used to make a threshold judgment based on the attitude safety score and the preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

[0018] This invention provides a posture-sensing alarm call method and device for a bedside device. The method acquires user posture information data by detecting user posture information through sensors on the bedside device; preprocesses the posture information data and constructs posture safety features based on the preprocessed posture information data. These posture safety features include: sleep posture stability features, sitting posture stability features, and posture change detection features; threshold judgment is performed based on the sleep posture stability features, sitting posture stability features, and posture change detection features; a safety warning is issued when a preset safety threshold condition is not met; when the preset safety threshold condition is met, a posture safety score is calculated based on the sleep posture stability features, sitting posture stability features, and posture change detection features; a threshold judgment is performed between the posture safety score and a preset posture safety score; a safety warning is issued when the posture safety score is higher than the preset posture safety score threshold. The beneficial effects include: 1. Real-time monitoring of user posture and immediate safety alerts: Sensors on the bedside device detect user posture information in real time, accurately capturing changes in posture during sleep and sitting. Through the collection and analysis of multi-dimensional data, it can quickly identify whether the user is at risk of falling or unstable posture, and issue timely safety alerts when abnormal situations occur, greatly improving user safety. 2. Multi-dimensional sensor data fusion improves monitoring accuracy: The fusion and analysis of data from multiple sensors, including acceleration, angle, and pressure, provides a more comprehensive understanding of the user's actual state. Different sensor data complement each other, effectively avoiding the limitations that may arise from using a single sensor alone. For example, acceleration sensors may be affected by external environmental interference, while pressure sensors can supplement the accurate perception of changes in the user's posture. This fusion of multi-source data not only improves the system's robustness but also more accurately determines whether the user is in an unsafe posture. 3. Comprehensive assessment of user safety status to reduce false alarm rate: By combining and analyzing multiple features such as sleeping posture stability, sitting posture stability, and posture change detection, the system avoids false alarms that may result from judging a single feature. The device not only accurately detects the stability and changes in user posture, but also uses threshold judgments and comprehensive scoring of multiple features to make the safety score assessment more accurate, thereby reducing false alarms and improving alarm accuracy.

[0019] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 A flowchart illustrating an attitude-sensing alarm call method for a bedside device, as shown in an exemplary embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the structure of a posture-sensing alarm call device for a bedside device, as shown in an exemplary embodiment of the present invention. Detailed Implementation

[0021] The embodiments of the present invention will be described below with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention and not for limiting the scope of protection of the present invention.

[0022] It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Therefore, the drawings only show the components related to the present invention and are not drawn according to the actual number, shape and size of the components in the actual implementation. In the actual implementation, the form, quantity and proportion of each component can be arbitrarily changed, and the layout of the components may also be more complex.

[0023] In the following description, numerous details are explored to provide a more thorough explanation of embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention.

[0024] Example 1 A posture-sensing alarm call method for bedside devices, such as Figure 1 As shown, it includes: The user's posture information data is obtained by detecting the user's posture information through sensors on the bedside equipment; The posture information data is preprocessed, and posture safety features are constructed based on the preprocessed posture information data. The posture safety features include: sleeping posture stability features, sitting posture stability features, and posture change detection features. Threshold judgments are made based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the detection characteristics of the posture change. When the preset safety threshold conditions are not met, a safety warning is issued. When the preset safety threshold condition is met, the posture safety score is calculated based on the sleep posture stability feature, the sitting posture stability feature and the posture change detection feature. A threshold is determined based on the attitude safety score and a preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

[0025] Specifically, sensors in the headboard unit, combined with sensors under the mattress, detect and acquire real-time user posture data, including acceleration, angle, and pressure. Acceleration data is acquired through accelerometers located in the headboard unit and under the mattress, used to monitor the user's body movement, turning over, sitting stability, dynamic changes, and acceleration changes during sudden posture changes. Angle data is acquired through angle sensors and gyroscopes in the headboard unit and under the mattress, used to detect changes in the angle between the body and the bed surface, monitor angle changes during rapid rotation or tilting, and the body tilt angle when sitting. Pressure data is acquired through pressure sensors located inside the mattress, used to monitor changes in pressure on the mattress or mattress surface, and detect the stability of pressure distribution when sitting.

[0026] The attitude information data undergoes preprocessing to ensure the accuracy, reliability, and real-time performance of subsequent data analysis and judgment. Preprocessing includes noise filtering and denoising, standardization and normalization, outlier detection and processing, feature extraction, and data dimensionality reduction. Noise filtering removes useless information from the collected data through filtering and smoothing methods, making the data smoother and more reliable. Standardization and normalization adjust all acquired data to a unified standard, facilitating subsequent analysis and comparison and improving the fusion effect of data from different sensors. Outlier detection and processing detect and correct abnormal data generated by sensors to avoid affecting the system's accuracy. Feature extraction and data dimensionality reduction extract key safety features from a large amount of raw attitude information data, simplifying subsequent processing and analysis.

[0027] The present invention is further configured to construct attitude safety features based on the preprocessed attitude information data, including: The sleep posture stability characteristics are calculated based on the acceleration, angle, and pressure; the sitting posture stability characteristics are calculated based on the acceleration, angle, and pressure; and the posture change detection characteristics are calculated based on the acceleration and angle. Specifically, the calculation logic for the sleep posture stability characteristics is as follows: ,in, for Characteristics of sleep posture stability at any given time The influence coefficient of acceleration change on stability score, For the change in acceleration, For the sleep instability function, The bed surface disturbance index represents the change in acceleration. The calculation logic is as follows: ,in, The sampling time interval is the sleep instability function. The calculation logic is as follows: ,in, For slope parameters, For offset parameters, The change in attitude angle is represented by the bed surface disturbance index. The calculation logic is as follows: ,in, The first adjustment parameter, For pressure change; sleep instability function By capturing the user's reaction to sudden changes in posture through nonlinear response to angle changes, the system can detect changes in the user's body position in a timely manner, helping to more accurately determine whether there is a risk of postural instability during sleep; Bed surface disturbance index By monitoring changes in pressure, it helps detect changes in the user's posture, promptly identifying the impact of potential sleeping instability or movements such as turning over on the sleep posture; sleep posture stability characteristics It can quantify the user's postural stability and combine data from multiple dimensions to assess whether the user is in a safe sleeping posture. A value closer to 1 indicates a more unstable user posture. Used to control the sensitivity of acceleration to attitude stability, with a value range of [0,1]; and Used to determine the sensitivity of posture changes in the sleep instability function. The value range is [0.1, 10]. The value is adjusted according to the required sensitivity to change; The sensitivity used to control the bed surface disturbance index ranges from [0,1]. Through multi-dimensional data fusion, the system can more accurately monitor changes in the user's sleep posture and issue an alarm in a timely manner when postural instability is detected. The calculation logic for the sitting posture stability feature is as follows: ,in, For characteristics of sitting and standing stability, For smoothing parameters, For dynamic center of gravity fluctuation, For acceleration angle linkage, For dynamic pressure fluctuations and dynamic center of gravity fluctuations The calculation logic is as follows: ,in, For the center of gravity offset, Sensitivity to stability when controlling changes in center of gravity displacement For threshold parameters, acceleration and angle are linked. The calculation logic is as follows: ,in, For the change in acceleration, For the change in posture angle, The second adjustment parameter is dynamic pressure fluctuation. The calculation logic is as follows: ,in, For the third adjustment parameter, The pressure distribution non-uniformity index; the seated posture stability feature is used to evaluate the user's postural stability in a seated state. It is evaluated by combining center of gravity shift, acceleration change, posture angle change, and pressure distribution. When the seated posture stability feature... The closer the value is to 1, the more unstable the sitting posture. For the specific calculation logic, please refer to the calculation formula above. In the above calculation formula, dynamic center of gravity fluctuation... Used to detect the degree of shift in the user's center of gravity when sitting, with acceleration and angle linkage. Used to describe the interaction between changes in acceleration and changes in attitude angle, dynamic pressure fluctuations. Used to reflect uneven pressure distribution in mattresses; Used to control the smoothness of the scoring, with values ​​ranging from [0.1, 1]. Used to control the sensitivity of stability to changes in center of gravity displacement, with values ​​ranging from [0.1, 5]. Used to set the sensitivity of the center of gravity offset, with a value range of [0.1, 5]; Used to control the impact of pressure changes on sitting stability, with values ​​ranging from [0.05, 2]; by comprehensively considering dynamic center of gravity fluctuations, acceleration angle linkage, and dynamic pressure fluctuations, it more accurately assesses the user's sitting posture stability, promptly detects potential risks, and prevents falls; the calculation logic for posture change detection features is as follows: ,in, For the detection of posture mutation features, For threshold, For the characteristics of combined acceleration and angle changes, The relative abrupt change factor between angle and acceleration, For adaptive threshold function, the joint variation characteristics of acceleration and angle. The calculation logic is as follows: ,in, For the change in acceleration, For attitude angle change, For acceleration decay time constant, The angle decay time constant is the abrupt change factor between angle and acceleration. The calculation logic is as follows: ,in, A constant, adaptive threshold function The calculation logic is as follows: ,in, The empirical coefficient is used to determine whether a user's posture has changed; this is done by evaluating changes in acceleration and angle. When the posture change detection feature... A value of 1 indicates a sudden change in attitude at time t. The specific calculation logic is explained in the formula above. The formula also includes the characteristics of the combined change in acceleration and angle. Used to represent the degree of change in acceleration and attitude, the angle relative to the abrupt change in acceleration. An adaptive threshold function is used to measure the relative intensity of acceleration and angular changes. Used to adjust the threshold for detecting sudden changes in posture; Used to determine whether the mutation condition is met, with a value range of [0,1]. The decay rate used to control the effect of acceleration changes on mutation detection is in the range of [0.1, 10]. The decay rate used to control the effect of pose angle changes on mutation detection is in the range of [0.1, 10]. The sensitivity used to determine the adaptive threshold ranges from [0.1, 10]. By combining changes in acceleration and posture angle, it can accurately detect abrupt changes in user posture and promptly identify unsafe postures. The adaptive threshold function... It can dynamically adjust the threshold according to different environments and users, improving the system's adaptability in various situations.

[0028] The present invention is further configured such that the calculation logic for the attitude safety score is as follows: ,in, To score safely, For the characteristics of sleep posture stability, For characteristics of sitting and standing stability, For the detection of posture mutation features, To obtain and The maximum value in The weighting coefficients are used to comprehensively assess the user's current posture safety, and the score is based on different postural characteristics, including sleeping posture stability, sitting posture stability, and posture change detection features. The smaller the value, the more... The user's posture is safer at all times; the specific calculation logic is explained in the calculation formula above. In the above calculation formula, the stability characteristics of the sleeping posture are considered. This is used to reflect the user's postural stability and sitting posture stability during sleep. Used to reflect the stability of a user's posture while seated, and to detect abrupt changes in posture. Used to detect if the user's posture changes abruptly. The system is used to control the impact of posture changes on posture safety scores, with values ​​ranging from [0,10]. By integrating multiple features such as sleep posture stability, sitting posture stability, and posture change detection, the system can more comprehensively and accurately assess the user's posture safety. By taking the maximum value and weighting coefficient, the system can adjust the score according to different user states, making the safety assessment closer to actual needs.

[0029] The present invention is further configured to calculate an attitude safety score according to the calculation logic of the attitude safety score, compare the attitude safety score with a preset attitude safety score for threshold judgment, and issue a safety warning when the attitude safety score is higher than the preset attitude safety score threshold; specifically, the calculated attitude safety score is compared with the preset attitude safety score threshold, which is a safety standard value defined in advance according to system settings or user needs; when the attitude safety score is lower than or equal to the preset attitude safety score threshold, it indicates that the user's posture or state is relatively safe; when the attitude safety score is higher than the preset attitude safety score threshold, it indicates that there is a high safety risk, and a safety warning is issued.

[0030] The present invention is further configured to map the sleep posture stability features, sitting posture stability features, and posture change detection features to corresponding discretized bin intervals, respectively. Specifically, based on historical distribution, the bin boundaries of each feature are pre-defined, and each boundary sequence is arranged in order of size. After acquiring the three new types of features, their respective intervals are located in the corresponding boundary sequences, and index triples are generated to represent the position of the current posture in the three-dimensional discrete space. The historical distribution of discretized bin intervals is maintained within a sliding window of a preset length. Specifically, the latest index triplet is appended to a circular queue with a capacity of L, and the corresponding element is incremented by 1 in the three-dimensional combination counting matrix. When the queue length exceeds L, the earliest triplet is popped and its corresponding element is decremented by 1 in the counting matrix, thus ensuring that the matrix always reflects the distribution of the most recent L mapping results. The system identifies pose combinations whose frequency in historical distribution is less than a rare pattern threshold, marks them as rare pose combinations, and triggers an alarm call. Specifically, it extracts the current count value of the latest triplet in the counting matrix and compares it with a pre-set rare pattern threshold. If the count is lower than the threshold, the triplet is determined to be a rare pose combination. The system records the timestamp and original feature value of the identified rare pose combinations, generates an alarm message, and initiates an alarm call to a predefined receiver. To prevent duplicate alarms, already triggered combinations can be marked in a queue or counting matrix. A real-time anomaly detection framework is constructed based on the discretization mapping and frequency statistics of three types of posture safety features. First, the stability features of sleeping posture, the stability features of sitting posture, and the abrupt change detection features are mapped to corresponding discretized bin intervals to transform continuous feature values ​​into fixed-dimensional index triples. Second, a sliding buffer of preset length is used to queue the latest mapping results, and the occurrence frequency of each index triple is maintained synchronously in a three-dimensional combination counting matrix. Finally, by comparing with a rare pattern threshold, index triples with low occurrence frequencies are identified and marked as rare posture combinations, triggering an alarm call.

[0031] The present invention is further configured such that the corresponding discretized binning interval mapping logic includes: For each moment of the collected sleep posture stability feature value, the preset sleep bin boundary sequence is traversed sequentially to locate the smallest index i, such that the i-th item in the boundary sequence is ≤ sleep posture stability feature value < i+1-th item, and the corresponding i-th bin interval is confirmed; specifically, the values ​​of each item in the preset sleep bin boundary sequence are read sequentially; the smallest index i is found to satisfy: the i-th item boundary is ≤ current sleep feature value < i+1-th item boundary; this i-th value is used as the bin interval identifier for the sleep feature; For the sitting posture stability feature value, traverse the preset sitting bin boundary sequence to identify the corresponding j-th bin interval; specifically, traverse the sitting bin boundary sequence and locate the minimum j value such that the j-th boundary ≤ the current sitting feature value < the (j+1)-th boundary; use this j value to identify the bin interval of the sitting feature. The pose mutation detection feature value is traversed through the preset mutation binning boundary sequence to confirm the corresponding k-th binning interval; specifically, the mutation binning boundary sequence is traversed to locate the minimum k value that meets the condition that the k-th boundary ≤ the current mutation feature value < the (k+1)-th boundary; this k value is used as the binning interval identifier for the mutation feature. The binning index in the form of (i,j,k) triples is used as the mapping result.

[0032] The present invention is further configured to maintain the historical distribution of discretized bin intervals within a sliding window of a preset length, including: Set the maximum capacity L of the sliding window and initialize a circular queue to store the discretized bin index triples. Initialize a three-dimensional combination counting matrix Counts of size N×M×K with all elements set to zero. Specifically, set the maximum window capacity L to limit the length of the historical records; create a circular queue of capacity L to store the latest L discretized bin index triples; construct a three-dimensional combination counting matrix Counts of size N×M×K and initialize all elements to 0 to count the number of times each index triple appears in the window. Receive the latest mapped triple index (i,j,k), append (i,j,k) to the tail of the circular queue; increment the corresponding Counts[i][j][k]; specifically, receive the current mapped index triple (i,j,k); append (i,j,k) to the tail of the circular queue; increment the corresponding Counts[i][j][k] by 1 to reflect the new occurrence of this combination within the window; If the length of the circular queue exceeds L, the earliest enqueued triplet (i0, j0, k0) is popped from the front of the queue, and Counts[i0][j0][k0] is decremented. Specifically, it checks whether the current length of the circular queue exceeds L; if it does, the earliest enqueued triplet (i0, j0, k0) is popped from the front of the queue; the value of Counts[i0][j0][k0] is decremented by 1 to remove the statistical contribution of past moments; this ensures that the Counts matrix always accurately reflects the frequency of occurrence of all triplets in the circular queue; the distribution overview of the most recent L pose indices can be obtained by reading Counts at any time, which can be used for rare pattern detection or dynamic threshold adjustment.

[0033] The present invention is further configured to identify pose combinations that occur less frequently than a rare pattern threshold in historical distributions, mark them as rare pose combinations, and trigger an alarm call, including: After updating Counts, check the latest mapping index triple (i,j,k) and obtain Counts[i][j][k] from the three-dimensional combination counting matrix; The frequency of the acquired combinations is compared with the rare pattern threshold. If Counts[i][j][k] is less than the rare pattern threshold, the triplet is marked as a rare pose combination. Specifically, after updating the sliding window and the three-dimensional combination counting matrix, Counts[i][j][k] corresponding to the latest mapping index triplet (i,j,k) is directly read from the matrix. This value reflects the number of times the pose combination has appeared in the most recent L mappings. The frequency of the acquired combinations is compared with the preset rare pattern threshold. If Counts[i][j][k] is lower than the rare pattern threshold, the index triplet is determined to be a rare pose combination, indicating that the current pose pattern rarely appears in normal user behavior, which may indicate an anomaly. Furthermore, the rare pattern threshold is a preset integer threshold used to distinguish between common and rare pose combinations; combinations below this threshold are considered rare or abnormal patterns. For triples marked as rare pose combinations, a timestamp and original feature value are recorded, an alarm message is generated, and the alarm module is invoked to initiate an alarm call to a predefined receiver. Specifically, for the marked rare pose combinations, the corresponding timestamp and three types of original feature values ​​are recorded, and an alarm message containing the above information is generated. This message is transmitted to the alarm module, which initiates an alarm call according to the predefined receiver configuration to ensure that abnormal poses are responded to in a timely manner.

[0034] The above technical solution enables fall detection for users. Sensors installed at the head of the bed and on the mattress monitor the user's fall status. When the acceleration sensors at the head of the bed and on the mattress detect a sharp change, and the angle sensor detects a tilt angle greater than a preset threshold, and the pressure distribution on the mattress changes significantly, a fall is determined. The detection of abnormal actions such as a user pounding their chest is achieved by capturing frame images through a camera in the bedside device and uploading them to the cloud for anomaly detection processing. This is existing technology and will not be elaborated on further.

[0035] The proposed posture-sensing alarm method for a bedside device also includes fall detection for tilting or wobbling of the bedside device. When a user is in an uncomfortable state, they may attempt to touch or push the device, causing it to tilt or tip over. When the bedside device tilts or experiences an unstable posture change, sensors such as accelerometers and angle sensors monitor these changes in real time. An algorithm is then used to determine whether the device has reached an alarm threshold. When the device tilts to a preset safety risk level, an alarm is automatically triggered, alerting relevant personnel to take timely intervention measures to prevent potential accidents.

[0036] Example 2 Please see Figure 2 The exemplary bedside device's attitude-sensing alarm call device includes: Data acquisition module: used to acquire user posture information data by detecting user posture information through sensors on the bedside device; Feature construction module: used to preprocess the posture information data and construct posture safety features based on the preprocessed posture information data, wherein the posture safety features include: sleeping posture stability features, sitting posture stability features and posture change detection features; The first early warning module is used to make a threshold judgment based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the change in posture detection characteristics. When the preset safety threshold conditions are not met, a safety warning is issued. The scoring calculation module is used to calculate a posture safety score based on the sleeping posture stability feature, the sitting posture stability feature, and the posture change detection feature when the preset safety threshold condition is met. The second early warning module is used to make a threshold judgment based on the attitude safety score and the preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

[0037] It should be noted that the posture-sensing alarm call device for bedside equipment provided in the above embodiments and the posture-sensing alarm call method for bedside equipment provided in the above embodiments belong to the same concept. The specific operation methods of each module and unit have been described in detail in the method embodiments and will not be repeated here. In practical applications, the posture-sensing alarm call device for bedside equipment provided in the above embodiments can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above. This is not a limitation here.

[0038] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0039] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0040] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of single or multiple items. For example, at least one of a, b, or c can mean: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0041] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0042] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0043] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0044] In the several embodiments provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0045] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0046] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0047] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0048] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A posture-sensing alarm call method for a bedside device, characterized in that, include: The user's posture information data is obtained by detecting the user's posture information through sensors on the bedside equipment; The posture information data is preprocessed, and posture safety features are constructed based on the preprocessed posture information data. The posture safety features include: sleeping posture stability features, sitting posture stability features, and posture change detection features. Threshold judgments are made based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the detection characteristics of the posture change. When the preset safety threshold conditions are not met, a safety warning is issued. When a preset safety threshold condition is met, a posture safety score is calculated based on the sleep posture stability feature, the sitting posture stability feature, and the posture change detection feature. The calculation logic for the posture safety score is as follows: ,in, To score safely, For the characteristics of sleep posture stability, For characteristics of sitting and standing stability, For the detection of posture mutation features, To obtain and The maximum value in These are the weighting coefficients; A threshold is determined based on the attitude safety score and a preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued. The sleep posture stability feature, sitting posture stability feature, and posture change detection feature are mapped to corresponding discretized bin intervals. The corresponding discretized bin interval mapping logic includes: for the sleep posture stability feature value collected at each time step, traversing the preset sleep bin boundary sequence sequentially, locating the smallest index i, such that the i-th item in the boundary sequence ≤ sleep posture stability feature value < i+1-th item, and confirming the corresponding i-th bin interval; traversing the preset sitting bin boundary sequence for the sitting posture stability feature value, and confirming the corresponding j-th bin interval; traversing the preset change bin boundary sequence for the posture change detection feature value, and confirming the corresponding k-th bin interval; and using the bin index in the form of (i,j,k) triples as the mapping result. Maintain the historical distribution of discretized bin intervals within a sliding window of a preset length, including: setting the maximum capacity L of the sliding window, initializing a circular queue to store discretized bin index triples, initializing a three-dimensional combination counting matrix Counts of size N×M×K, and setting all elements to zero; receiving the latest mapped triple index (i,j,k), appending (i,j,k) to the tail of the circular queue; incrementing Counts[i][j][k] accordingly; if the length of the circular queue exceeds L, popping the earliest enqueued triple (i0,j0,k0) from the head of the queue, and decrementing Counts[i0][j0][k0]. The process involves identifying pose combinations whose frequency in historical distribution is less than a rare pattern threshold, marking them as rare pose combinations, and triggering an alarm call. This includes: after updating Counts, checking the latest mapping index triplet (i,j,k), obtaining Counts[i][j][k] from the three-dimensional combination counting matrix, where Counts[i][j][k] reflects the number of times the pose combination has appeared in the most recent L mappings; comparing the obtained combination frequency with the rare pattern threshold, and if Counts[i][j][k] is less than the rare pattern threshold, marking the triplet as a rare pose combination; recording the timestamp and original feature value of the triplet marked as a rare pose combination, generating an alarm message, calling the alarm module, and initiating an alarm call to a predefined receiver.

2. The posture-sensing alarm call method for a bedside device according to claim 1, characterized in that, The attitude information data includes acceleration, angle, and pressure.

3. The posture-sensing alarm call method for a bedside device according to claim 2, characterized in that, Based on the preprocessed attitude information data, attitude safety features are constructed, including: The stability characteristics of the sleeping posture are calculated based on the acceleration, the angle, and the pressure. The stability characteristics of the sitting posture are calculated based on the acceleration, the angle, and the pressure. The posture change detection features are calculated based on the acceleration and the angle.

4. The posture-sensing alarm call method for a bedside device according to claim 1, characterized in that, An attitude safety score is calculated based on the calculation logic of the attitude safety score. The attitude safety score is compared with a preset attitude safety score for threshold judgment. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

5. A posture-sensing alarm call device for a bedside device, used to implement the posture-sensing alarm call method for a bedside device as described in any one of claims 1-4, characterized in that, include: Data acquisition module: used to acquire user posture information data by detecting user posture information through sensors on the bedside device; Feature construction module: used to preprocess the posture information data and construct posture safety features based on the preprocessed posture information data, wherein the posture safety features include: sleeping posture stability features, sitting posture stability features and posture change detection features; The first early warning module is used to make a threshold judgment based on the stability characteristics of the sleeping posture, the stability characteristics of the sitting posture, and the change in posture detection characteristics. When the preset safety threshold conditions are not met, a safety warning is issued. The scoring calculation module is used to calculate a posture safety score based on the sleeping posture stability feature, the sitting posture stability feature, and the posture change detection feature when the preset safety threshold condition is met. The second early warning module is used to make a threshold judgment based on the attitude safety score and the preset attitude safety score. When the attitude safety score is higher than the preset attitude safety score threshold, a safety warning is issued.

Images