Mattress zoned adaptive adjustment method and system based on pressure and posture data

Abstract

This invention discloses a method and system for adaptive adjustment of mattress zones based on pressure and posture data. The method involves calibrating the pressure sensitivity coefficients of each zone using physical structure difference parameters, constructing a pressure data correction model, and standardizing the original pressure time-series data. Based on the standardized data, the pressure change gradient is extracted to identify the effective pressure-bearing boundary of the user's body pressure. Combining this boundary with the pressure data, the user's sleeping posture and trunk posture angle are identified. Furthermore, the gravity distribution characteristics under this posture are analyzed, and the morphological deviation between each airbag zone and the target comfort support surface is calculated. Finally, air pressure adjustment commands for the airbag zones are generated based on the morphological deviation, thereby achieving adaptive and precise adjustment of the mattress support shape and effectively improving sleep comfort.

Description

Technical Field

[0001] This invention relates to the field of smart home monitoring and control technology, and in particular to a mattress zone adaptive adjustment method and system based on pressure and posture data. Background Technology

[0002] With the popularization of the concept of healthy sleep, smart mattress technology is developing from simple functional adjustment to personalized, adaptive support. Traditional mattress adjustment usually relies on manual control by the user or preset fixed modes, and cannot dynamically adjust according to the user's changing posture and body pressure distribution in real time during sleep, making it difficult to provide precise and physiologically conforming comfortable support.

[0003] In existing technologies, some smart mattresses have integrated pressure sensors and airbag zones, attempting to achieve automatic adjustment by monitoring pressure distribution. However, these solutions typically have the following technical limitations: First, most solutions directly use raw data collected by pressure sensors for judgment, failing to fully consider the systematic errors caused by the mattress's own physical structure (such as airbag seams, uneven surface fabric tension, and airbag geometric differences) on pressure sensing accuracy, resulting in distorted raw data. Second, existing methods are relatively coarse in recognizing human posture, usually only able to determine basic sleeping positions (such as supine and lateral), lacking the ability to capture fine features such as subtle torso twisting angles and body pressure boundaries, and unable to provide accurate biomechanical support analysis for key parts of the body such as the spine. Furthermore, existing adjustment strategies are often based on empirical models or simple pressure equalization logic, failing to establish a closed-loop control model from the user's real-time posture and body pressure distribution to the ideal physiological support surface, resulting in insufficient accuracy of adjustment and adaptation to individual comfort.

[0004] Therefore, overcoming sensor errors, accurately identifying sleep posture and trunk angle, and calculating adaptive and precise support patterns based on biomechanical models to achieve dynamic and intelligent adjustment of mattress zone support performance has become a key technical challenge for improving the comfort and health support effect of smart mattresses. Summary of the Invention

[0005] To address at least one of the aforementioned technical problems, this invention proposes a mattress zone adaptive adjustment method and system based on pressure and posture data.

[0006] The first aspect of this invention provides a mattress zone adaptive adjustment method based on pressure and posture data, comprising: The raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress are obtained. By obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, the pressure sensing error of each zone is calibrated according to the physical structure difference parameters, and the pressure sensitivity coefficient of each zone is determined. A cross-regional pressure data correction model is constructed based on the pressure sensitivity coefficient. The real-time pressure time series data during the user's sleep process is corrected based on the pressure data correction model to obtain standardized pressure time series data. Extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; The user's current sleep posture and torso posture angle are identified based on the effective pressure-bearing boundary and the standardized pressure time-series data. The gravity distribution characteristics of the user's current sleeping posture and torso posture angle are analyzed, the morphological deviation between each airbag zone and the user's target comfort support surface is calculated, and airbag zone air pressure adjustment commands are generated based on the morphological deviation.

[0007] In this solution, the process involves acquiring the raw pressure time-series data collected by the pressure sensors in each zone of the airbag mattress, obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, calibrating the pressure sensing error of each zone based on the physical structure difference parameters, and determining the pressure sensitivity coefficient of each zone. Specifically: A pressure test was conducted by applying a preset stepped pressure load covering the range of human body weight to the airbag mattress. The original pressure time sequence data of each independent airbag zone on the surface of the airbag mattress was collected in real time within the preset sampling period of the pressure test by an array-type pressure sensing unit. Simultaneously acquire geometric topological data and surface fabric tension distribution data of the airbag mattress surface, and align the geometric topological data and fabric tension distribution data in time and space to construct physical structure difference parameters of the airbag mattress surface, including differences in airbag mattress surface curvature, airbag seam protrusion height and fabric elastic modulus. The original pressure time series data is correlated with the physical structure difference dataset to identify the pressure signal distortion characteristics caused by surface unevenness, and the deviation between the measured pressure value and the theoretical applied pressure value of each zone under different step pressure is extracted. Based on the deviation, a mapping relationship between pressure sensing error and physical structure difference parameters is constructed. By performing nonlinear fitting on the mapping relationship, the pressure sensitivity coefficient of each airbag zone under different deformation states is determined, and the pressure sensing error of each zone is calibrated.

[0008] In this solution, the step of constructing a cross-regional pressure data correction model based on the pressure sensitivity coefficient, and then correcting the real-time pressure time-series data during the user's sleep process using the pressure data correction model to obtain standardized pressure time-series data, specifically involves: Based on the pressure sensitivity coefficient of each airbag zone, the ratio of the pressure sensitivity coefficient of each airbag zone to that of the error-free pressure sensing state is calculated, and the distribution of the ratio on the surface of the airbag mattress is mapped to the initial pressure compensation coefficient matrix required for each zone. A radial basis function neural network is introduced, with the spatial coordinates of each airbag section as input and the initial pressure compensation coefficient of the corresponding position as the expected output, to construct a nonlinear mapping model from spatial coordinates to pressure compensation coefficient. By using the hidden layers of the radial basis function neural network and radial basis functions such as Gaussian function, the input space is nonlinearly transformed. The pressure sensing error data of the airbag mattress under different deformation states is used as the training set to iteratively optimize the connection weights and basis function center and width parameters of the network to form a pressure data correction model. Real-time pressure time-series data during the user's sleep process is acquired, and the real-time pressure time-series data is input into the pressure data correction model for correction to obtain standardized pressure time-series data.

[0009] In this solution, the step of extracting the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identifying the effective pressure-bearing boundary of the user's body contact surface with the mattress based on the pressure change gradient, specifically involves: Based on the pressure value difference sequence of adjacent sampling points of the standardized pressure time series data, the pressure change gradient in the neighborhood of each sampling point is calculated according to the pressure value difference sequence. Based on the adaptive threshold segmentation algorithm, regions where the pressure change gradient is greater than the change threshold are marked as pressure transition regions, and regions where the pressure change gradient is less than the change threshold are marked as pressure stable regions. Connect adjacent pressure transition regions to form a continuous boundary region, calculate the integral value of the pressure change gradient in the continuous boundary region, and determine the candidate effective pressure-bearing boundary profile based on the integral value of the pressure change gradient. Based on the pressure value distribution characteristics on the candidate effective pressure bearing boundary profile according to the standardized pressure time series data, the pressure value distribution characteristics are enhanced by the edge detection operator, the gradient abrupt transition points of the pressure value from non-zero to zero are identified, and the gradient abrupt transition points are connected to construct the preliminary effective pressure bearing boundary curve. Based on the spatial distribution characteristics of the pressure stability region, the average pressure on both sides of the preliminary effective pressure-bearing boundary curve is calculated. Regions with average pressure greater than the pressure threshold are identified as pressure-bearing regions, and regions with average pressure less than the pressure threshold are identified as non-pressure-bearing regions. The preliminary effective pressure-bearing boundary curve is then corrected based on the pressure-bearing regions, and the effective pressure-bearing boundary of the contact surface between the user's body and the mattress is output.

[0010] In this solution, the step of identifying the user's current sleep posture and trunk posture angle based on the effective pressure-bearing boundary and the standardized pressure time-series data specifically includes: Obtain multiple standard sleep posture templates, which include standard pressure distribution characteristic areas of various parts of the human body on the mattress and standard trunk axis in supine, left lateral, right lateral, and prone positions. Based on the effective pressure-bearing boundary and the standardized pressure time series data, a real-time pressure distribution matrix of the user's body is constructed, and the spatial pressure distribution similarity between the real-time pressure distribution matrix and each standard sleep posture template is calculated. The standard sleep posture template with the highest similarity to the spatial pressure distribution of the real-time pressure distribution matrix is ​​selected as the user's current sleep posture; The theoretical principal axis direction of the user's torso on the mattress surface is determined based on the user's current sleeping posture, and the spatial deflection angle of the actual pressure distribution relative to the theoretical principal axis is calculated based on the centroid position and higher-order moment characteristics of the pressure distribution within the effective pressure-bearing boundary. A spatial point set registration method based on singular value decomposition is introduced to register the spatial coordinates of the high-pressure point set in the real-time pressure distribution matrix with the coordinates of the standard pressure point set of the corresponding part in the selected standard sleep posture template, and determine the optimal spatial rotation matrix. The rotation angle component of the user's torso in the horizontal axis direction is extracted based on the optimal spatial rotation matrix. The rotation angle component is then fused with the spatial deflection angle, and the user's torso posture angle is calculated by weighted averaging.

[0011] In this solution, the analysis of the gravity distribution characteristics of the user's current sleeping posture and torso posture angle, the calculation of the morphological deviation between each airbag zone and the user's target comfort support surface, and the generation of airbag zone air pressure adjustment commands based on the morphological deviation are as follows: Based on the standardized pressure time series data, the pressure distribution characteristics of the user's body on the mattress surface are extracted, and a biomechanical coupling model of the human spine-mattress contact surface is established according to the user's current sleeping posture and trunk posture angle. The finite element analysis method is introduced, and the trunk posture angle, pressure distribution characteristics and preset physiological curvature parameters of various parts of the human body are used as inputs to calculate the ideal body pressure distribution required to maintain the natural physiological curvature of the human spine under the user's current sleeping posture. Based on the ideal body pressure distribution, a target comfortable support surface is constructed in the mattress surface space using a thin plate spline interpolation algorithm. Spatially register the target comfort support surface with the current shape surface of the airbag mattress, calculate the elevation difference between the center point of each airbag section on the current shape surface and the corresponding point on the target comfort support surface, and obtain the shape deviation of each airbag section. Based on the morphological deviation of each airbag section, the required air pressure adjustment amount for the corresponding airbag section is generated, and an air pressure adjustment command for each airbag section is generated to adjust the morphology of each airbag section.

[0012] A second aspect of the present invention also provides a mattress zone adaptive adjustment system based on pressure and posture data. The system includes a memory and a processor. The memory includes a mattress zone adaptive adjustment method program based on pressure and posture data. When the processor executes the mattress zone adaptive adjustment method program based on pressure and posture data, it performs the following steps: The raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress are obtained. By obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, the pressure sensing error of each zone is calibrated according to the physical structure difference parameters, and the pressure sensitivity coefficient of each zone is determined. A cross-regional pressure data correction model is constructed based on the pressure sensitivity coefficient. The real-time pressure time series data during the user's sleep process is corrected based on the pressure data correction model to obtain standardized pressure time series data. Extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; The user's current sleep posture and torso posture angle are identified based on the effective pressure-bearing boundary and the standardized pressure time-series data. The gravity distribution characteristics of the user's current sleeping posture and torso posture angle are analyzed, the morphological deviation between each airbag zone and the user's target comfort support surface is calculated, and airbag zone air pressure adjustment commands are generated based on the morphological deviation.

[0013] This invention discloses a method and system for adaptive adjustment of mattress zones based on pressure and posture data. The method involves calibrating the pressure sensitivity coefficients of each zone using physical structure difference parameters, constructing a pressure data correction model, and standardizing the original pressure time-series data. Based on the standardized data, the pressure change gradient is extracted to identify the effective pressure-bearing boundary of the user's body pressure. Combining this boundary with the pressure data, the user's sleeping posture and trunk posture angle are identified. Furthermore, the gravity distribution characteristics under this posture are analyzed, and the morphological deviation between each airbag zone and the target comfort support surface is calculated. Finally, air pressure adjustment commands for the airbag zones are generated based on the morphological deviation, thereby achieving adaptive and precise adjustment of the mattress support shape and effectively improving sleep comfort. Attached Figure Description

[0014] Figure 1 A flowchart of a mattress zone adaptive adjustment method based on pressure and posture data according to the present invention is shown; Figure 2A flowchart illustrating the present invention for determining the pressure sensitivity coefficient of each zone is shown; Figure 3 The flowchart illustrating the present invention for identifying the effective pressure-bearing boundary of the contact surface between the user's body and the mattress is shown. Figure 4 A block diagram of a mattress zone adaptive adjustment system based on pressure and posture data according to the present invention is shown. Detailed Implementation

[0015] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in these embodiments can be combined with each other.

[0016] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0017] Figure 1 The flowchart of a mattress zone adaptive adjustment method based on pressure and posture data according to the present invention is shown.

[0018] like Figure 1 As shown, the first aspect of the present invention provides a mattress zone adaptive adjustment method based on pressure and posture data, comprising: S102, acquire the raw pressure time series data collected by the pressure sensors of each zone of the airbag mattress, acquire the physical structure difference parameters of the surface of each zone of the airbag mattress, calibrate the pressure sensing error of each zone according to the physical structure difference parameters, and determine the pressure sensitivity coefficient of each zone. S104, Construct a cross-regional pressure data correction model based on the pressure sensitivity coefficient, and perform correction operations on the real-time pressure time series data during the user's sleep process based on the pressure data correction model to obtain standardized pressure time series data; S106, extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; S108, Identify the user's current sleep posture and trunk posture angle based on the effective pressure bearing boundary and the standardized pressure time series data; S110, analyze the gravity distribution characteristics of the user's current sleeping posture and torso posture angle, calculate the morphological deviation between each airbag zone and the user's target comfort support surface, and generate airbag zone air pressure adjustment commands based on the morphological deviation.

[0019] It should be noted that by conducting precise physical structure difference analysis and pressure sensing error calibration of each zone of the airbag mattress, systematic measurement errors introduced by factors such as airbag seam protrusions, surface curvature changes, and uneven fabric tension are eliminated, thereby obtaining a high-precision pressure sensitivity coefficient. Furthermore, based on this coefficient, a cross-regional pressure data correction model is constructed, which can dynamically standardize real-time pressure time-series data during sleep, significantly improving the spatial consistency and comparability of pressure data. Combining this boundary with the corrected pressure data, the system can not only accurately identify the user's macroscopic sleep posture but also further analyze the subtle posture angles of the torso, achieving refined perception of human sleeping posture. Finally, by analyzing the gravity distribution characteristics under this posture and calculating the morphological deviation between each airbag zone and the ideal comfortable support surface based on a biomechanical model, precise airbag zone air pressure adjustment commands can be generated. Ultimately, this allows the airbag mattress to dynamically and adaptively conform to the user's physiological curve changes during sleep, effectively alleviating local pressure concentration during sleep and improving sleep quality and comfort. The airbag mattress is a smart mattress composed of multiple independent or interconnected airbag units. Each airbag unit can typically be inflated and deflated independently, and its support height and firmness can be changed by adjusting the internal air pressure.

[0020] Figure 2 The flowchart illustrating the present invention for determining the pressure sensitivity coefficient of each zone is shown.

[0021] According to an embodiment of the present invention, the step of acquiring the raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress, acquiring the physical structure difference parameters of the surface of each zone of the airbag mattress, calibrating the pressure sensing error of each zone based on the physical structure difference parameters, and determining the pressure sensitivity coefficient of each zone, specifically involves: A pressure test was conducted by applying a preset stepped pressure load covering the range of human body weight to the airbag mattress. The original pressure time sequence data of each independent airbag zone on the surface of the airbag mattress was collected in real time within the preset sampling period of the pressure test by an array-type pressure sensing unit. Simultaneously acquire geometric topological data and surface fabric tension distribution data of the airbag mattress surface, and align the geometric topological data and fabric tension distribution data in time and space to construct physical structure difference parameters of the airbag mattress surface, including differences in airbag mattress surface curvature, airbag seam protrusion height and fabric elastic modulus. The original pressure time series data is correlated with the physical structure difference dataset to identify the pressure signal distortion characteristics caused by surface unevenness, and the deviation between the measured pressure value and the theoretical applied pressure value of each zone under different step pressure is extracted. Based on the deviation, a mapping relationship between pressure sensing error and physical structure difference parameters is constructed. By performing nonlinear fitting on the mapping relationship, the pressure sensitivity coefficient of each airbag zone under different deformation states is determined, and the pressure sensing error of each zone is calibrated.

[0022] It should be noted that due to limitations in manufacturing processes, material properties, and structural design, the surface of an airbag mattress is not an ideal plane. The seams between airbags, the curvature of the airbags themselves, the uneven tension distribution of the surface covering fabric, and local differences in elastic modulus introduce significant and nonlinear pressure signal distortion. This causes pressure sensors installed on the same plane to produce different readings when subjected to the same load, due to differences in the physical environment of their location. By applying a stepped load covering the range of human body weight to the mattress, pressure data and physical structural parameters such as geometric topology and fabric tension were simultaneously collected to construct a multi-dimensional dataset of physical structural differences. Through correlation analysis, the pressure signal distortion caused by specific structural features was accurately quantified and separated. Furthermore, a nonlinear mapping between physical structural parameters and sensing errors was established, and a unique pressure sensitivity coefficient for each airbag zone under different deformation states was obtained through fitting.

[0023] According to an embodiment of the present invention, the step of constructing a cross-regional pressure data correction model based on the pressure sensitivity coefficient, and correcting the real-time pressure time-series data during the user's sleep process based on the pressure data correction model to obtain standardized pressure time-series data, specifically includes: Based on the pressure sensitivity coefficient of each airbag zone, the ratio of the pressure sensitivity coefficient of each airbag zone to that of the error-free pressure sensing state is calculated, and the distribution of the ratio on the surface of the airbag mattress is mapped to the initial pressure compensation coefficient matrix required for each zone. A radial basis function neural network is introduced, with the spatial coordinates of each airbag section as input and the initial pressure compensation coefficient of the corresponding position as the expected output, to construct a nonlinear mapping model from spatial coordinates to pressure compensation coefficient. It should be noted that the pressure sensing error caused by differences in the physical structure of the mattress has a highly nonlinear mapping relationship with its spatial location. This relationship is difficult to accurately describe using simple linear interpolation or polynomial fitting. Radial basis function neural networks, with their powerful nonlinear function approximation ability and local response characteristics, can effectively learn this complex spatial variation law. By using the spatial coordinates of each airbag zone as input and the corresponding pressure compensation coefficient derived from the pressure sensitivity coefficient as the desired output, the network can perform a nonlinear transformation of the input space using the Gaussian radial basis functions of the hidden layers, and learn a continuous and smooth mapping function from any position to its required compensation amount through training. The sensitivity coefficient of the error-free pressure sensing state is 1; the larger the error, the smaller the sensitivity coefficient, and the compensation coefficient is the percentage of pressure that needs to be compensated.

[0024] By using the hidden layers of the radial basis function neural network and radial basis functions such as Gaussian function, the input space is nonlinearly transformed. The pressure sensing error data of the airbag mattress under different deformation states is used as the training set to iteratively optimize the connection weights and basis function center and width parameters of the network to form a pressure data correction model. Real-time pressure time-series data during the user's sleep process is acquired, and the real-time pressure time-series data is input into the pressure data correction model for correction to obtain standardized pressure time-series data.

[0025] It should be noted that transforming the pressure sensitivity coefficients obtained from discrete partition calibration into a spatially continuous initial compensation coefficient matrix provides a precise training target for the neural network. Subsequently, leveraging the powerful spatial interpolation and function approximation capabilities of the radial basis function neural network, a correction model is constructed that can directly map accurate pressure compensation coefficients from any spatial location, achieving seamless and continuous correction of pressure readings at any point on the mattress surface. The pressure data correction model, by inputting the coordinates of any airbag partition on the mattress surface, outputs the accurate pressure compensation coefficients of that partition under any deformation state, optimized through learning. The correction operation involves inputting the real-time collected time-series pressure data of each partition into the pressure data correction model and multiplying it by the corresponding accurate pressure compensation coefficients generated in real-time by the model, thus completing the correction.

[0026] Figure 3 The flowchart illustrating the present invention identifies the effective pressure-bearing boundary of the contact surface between the user's body and the mattress.

[0027] According to an embodiment of the present invention, the step of extracting the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identifying the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient, specifically includes: Based on the pressure value difference sequence of adjacent sampling points of the standardized pressure time series data, the pressure change gradient in the neighborhood of each sampling point is calculated according to the pressure value difference sequence. Based on the adaptive threshold segmentation algorithm, regions where the pressure change gradient is greater than the change threshold are marked as pressure transition regions, and regions where the pressure change gradient is less than the change threshold are marked as pressure stable regions. Connect adjacent pressure transition regions to form a continuous boundary region, calculate the integral value of the pressure change gradient in the continuous boundary region, and determine the candidate effective pressure-bearing boundary profile based on the integral value of the pressure change gradient. Based on the pressure value distribution characteristics on the candidate effective pressure bearing boundary profile according to the standardized pressure time series data, the pressure value distribution characteristics are enhanced by the edge detection operator, the gradient abrupt transition points of the pressure value from non-zero to zero are identified, and the gradient abrupt transition points are connected to construct the preliminary effective pressure bearing boundary curve. Based on the spatial distribution characteristics of the pressure stability region, the average pressure on both sides of the preliminary effective pressure-bearing boundary curve is calculated. Regions with average pressure greater than the pressure threshold are identified as pressure-bearing regions, and regions with average pressure less than the pressure threshold are identified as non-pressure-bearing regions. The preliminary effective pressure-bearing boundary curve is then corrected based on the pressure-bearing regions, and the effective pressure-bearing boundary of the contact surface between the user's body and the mattress is output.

[0028] It should be noted that due to the tension of the mattress surface fabric and the local stretching caused by the body's weight, minute pressure values ​​may form in the critical area of ​​the human body's contour edge, which are not in direct contact but can still be detected by sensors. These pseudo-pressure signals generated by fabric stretching or indirect transmission can blur the true body pressure boundary, causing the identified contact contour to be larger than the actual human body projection. Therefore, by calculating the pressure change gradient and adaptively segmenting the pressure transition zone and stable zone, small disturbances can be effectively filtered out, and the possible boundary areas can be initially identified. By connecting the transition zone and calculating the gradient integral, candidate boundary contours are robustly extracted from the noise. Further, the pressure distribution on the candidate boundary is enhanced using an edge detection operator to accurately locate the abrupt change point from pressure to non-pressure, constructing a continuous preliminary boundary curve. Finally, a region verification mechanism based on the pressure mean is introduced to determine the pressure-bearing and non-pressure-bearing areas on the inner and outer sides of the preliminary boundary, thereby correcting and ultimately outputting an accurate effective pressure-bearing boundary. This significantly resists pressure noise interference caused by local deformation of the mattress and slight posture changes, accurately depicting the true contact contour between the human body and the mattress.

[0029] According to an embodiment of the present invention, the step of identifying the user's current sleep posture and trunk posture angle based on the effective pressure-bearing boundary and the standardized pressure time-series data specifically includes: Obtain multiple standard sleep posture templates, which include standard pressure distribution characteristic areas of various parts of the human body on the mattress and standard trunk axis in supine, left lateral, right lateral, and prone positions. Based on the effective pressure-bearing boundary and the standardized pressure time series data, a real-time pressure distribution matrix of the user's body is constructed, and the spatial pressure distribution similarity between the real-time pressure distribution matrix and each standard sleep posture template is calculated. The standard sleep posture template with the highest similarity to the spatial pressure distribution of the real-time pressure distribution matrix is ​​selected as the user's current sleep posture; The theoretical principal axis direction of the user's torso on the mattress surface is determined based on the user's current sleeping posture, and the spatial deflection angle of the actual pressure distribution relative to the theoretical principal axis is calculated based on the centroid position and higher-order moment characteristics of the pressure distribution within the effective pressure-bearing boundary. A spatial point set registration method based on singular value decomposition is introduced to register the spatial coordinates of the high-pressure point set in the real-time pressure distribution matrix with the coordinates of the standard pressure point set of the corresponding part in the selected standard sleep posture template, and determine the optimal spatial rotation matrix. The rotation angle component of the user's torso in the horizontal axis direction is extracted based on the optimal spatial rotation matrix. The rotation angle component is then fused with the spatial deflection angle, and the user's torso posture angle is calculated by weighted averaging.

[0030] It should be noted that by comparing the spatial similarity of the real-time pressure distribution matrix with the standard posture template, basic sleep postures such as supine, lateral, and prone are identified. Building upon this, a spatial point set registration method based on singular value decomposition (SVD) is further introduced. Since SVD can robustly solve for the optimal spatial rotation relationship between the real-time pressure point set and the standard template point set, it effectively overcomes calculation errors caused by irregularities in the contact surface between the human body and the mattress, and local missing or offset pressure distribution, thus accurately resolving the subtle deflection angle of the torso in the horizontal plane. Finally, by fusing the principal axis analysis results based on the pressure distribution moment characteristics with the precise rotation components obtained from point set registration, the calculated torso posture angle not only reflects the overall shape of the pressure distribution but also accurately captures the relative spatial position changes of key pressure-bearing parts. This allows the airbag mattress to sensitively perceive and meticulously compensate for subtle changes in spinal posture caused by the user's turning over or curling up during sleep, based on this high-precision angle information. The higher-order moment features are statistical quantities used to describe the details of the pressure distribution morphology. They mainly include the third-order central moment (skewness) to measure the symmetry of the pressure distribution, and the fourth-order central moment (kurtosis) to characterize the concentration and steepness of the pressure distribution.

[0031] According to an embodiment of the present invention, the analysis of the gravity distribution characteristics of the user's current sleeping posture and torso posture angle, the calculation of the morphological deviation between each airbag zone and the user's target comfort support surface, and the generation of airbag zone air pressure adjustment commands based on the morphological deviation are specifically as follows: Based on the standardized pressure time series data, the pressure distribution characteristics of the user's body on the mattress surface are extracted, and a biomechanical coupling model of the human spine-mattress contact surface is established according to the user's current sleeping posture and trunk posture angle. The finite element analysis method is introduced, and the trunk posture angle, pressure distribution characteristics and preset physiological curvature parameters of various parts of the human body are used as inputs to calculate the ideal body pressure distribution required to maintain the natural physiological curvature of the human spine under the user's current sleeping posture. Based on the ideal body pressure distribution, a target comfortable support surface is constructed in the mattress surface space using a thin plate spline interpolation algorithm. Spatially register the target comfort support surface with the current shape surface of the airbag mattress, calculate the elevation difference between the center point of each airbag section on the current shape surface and the corresponding point on the target comfort support surface, and obtain the shape deviation of each airbag section. Based on the morphological deviation of each airbag section, the required air pressure adjustment amount for the corresponding airbag section is generated, and an air pressure adjustment command for each airbag section is generated to adjust the morphology of each airbag section.

[0032] It should be noted that by establishing a biomechanical coupling model of the human spine-mattress contact surface and introducing the finite element analysis method, the ideal body pressure distribution required to maintain the natural physiological curvature of the human spine can be calculated. The target comfort support surface constructed using the thin-plate spline interpolation algorithm can accurately describe the continuous and smooth three-dimensional shape that the mattress surface should present to conform to the user's current spinal curve. By spatially registering this target surface with the current actual shape surface of the airbag mattress and calculating the elevation difference of the center points of each zone, the obtained shape deviation directly and accurately reflects the height difference between each individual airbag and the ideal support shape. Finally, the system directly generates corresponding airbag zone air pressure adjustment commands based on this shape deviation, driving the actuator to precisely inflate and deflate each airbag, thereby enabling the shape of the entire airbag array to dynamically and adaptively approximate the target comfort support surface, achieving real-time tracking and active adaptation to changes in the user's sleeping posture.

[0033] According to an embodiment of the present invention, it further includes: Based on the standardized pressure time series data, the spatiotemporal characteristics of pressure in each airbag zone are extracted. By comparing and analyzing the rate of change of pressure value over time and the pressure change gradient of each airbag zone with the standard spectrum of physiological pressure fluctuations generated by human respiration and body movement, candidate foreign body regions with abnormal pressure change patterns are identified. Based on the user's current sleep posture and the effective pressure-bearing boundary, the theoretical body pressure distribution of the standardized pressure time series data within the effective pressure-bearing boundary is calculated. The real-time pressure distribution of the candidate foreign body region with pressure anomalies is compared with the theoretical body pressure distribution. If the difference exceeds the preset physiological pressure fluctuation range, it is determined that there is a foreign body in the candidate foreign body region. When a foreign object is detected, the standardized pressure time series data containing the candidate foreign object region is optimized to generate updated standardized pressure time series data.

[0034] According to an embodiment of the present invention, when a foreign object is determined to exist, the standardized pressure time series data containing the candidate foreign object region is optimized to generate updated standardized pressure time series data, specifically as follows: Obtain the initial pressure compensation coefficients in the pressure data correction model corresponding to the candidate foreign object region and its surrounding normal airbag partitions, and construct a pressure compensation coefficient matrix containing spatial neighborhood information; Based on the pressure change gradient and the rate of change of pressure value over time, the spatiotemporal pressure anomaly features of the candidate foreign object region are extracted; The pressure compensation coefficient matrix and the pressure spatiotemporal anomaly features are input into an adaptive filter. The adaptive filter iteratively calculates a set of local compensation weights by minimizing the error between the high-frequency energy component of the pressure data in the candidate foreign object region and the theoretical body pressure distribution. The local compensation weight is combined with the initial pressure compensation coefficient to define the local pressure compensation correction function; The local pressure compensation correction function is applied to the standardized pressure time series data of the candidate foreign object region to perform spatial smoothing and amplitude suppression on the pressure peak. At the same time, the corrected pressure energy is diffused to the neighborhood that conforms to the theoretical body pressure distribution, thereby optimizing the pressure data caused by the foreign object and outputting the updated standardized pressure time series data that is smooth and conforms to the physiological pressure distribution characteristics of the human body.

[0035] It's important to note that during actual use of smart mattresses, there's a typical interference scenario caused by user negligence or habits: users might leave hard objects like phones or glasses cases on the mattress, or pets might lie on the surface. These foreign objects create a localized, high-hardness "pseudo-support point," generating abnormally high-pressure signals on the pressure sensors that resemble the support points of human bones but have vastly different dynamic characteristics. If the system indiscriminately misinterprets this abnormal pressure peak as part of the user's body and adjusts its configuration based on this, the corresponding airbags will inflate abnormally, failing to provide comfortable support that conforms to the body's curves. Instead, this could push up the foreign object, causing user discomfort or even safety risks.

[0036] Therefore, by analyzing the time-varying rate and gradient of pressure and comparing it with the standard spectral model of human physiological activity, abnormal pressure regions caused by static foreign objects and lacking the characteristic rhythms of living organisms are identified. Secondly, after determining the presence of a foreign object, an intelligent adaptive filtering mechanism is introduced. This mechanism takes the output of the pressure correction model and the identified abnormal features as input, and generates a targeted local pressure compensation correction function through iterative learning. This function is essentially an intelligent filter with dual spatial and frequency domain constraints. It can suppress abnormal high-pressure peaks and high-frequency noise caused by foreign objects, while cleverly diffusing the suppressed pressure energy reasonably to the surrounding normal physiological pressure-bearing areas according to the human biomechanical model. Finally, the system outputs updated pressure time-series data. This data maximizes the removal of foreign object interference, restoring the true pressure distribution contour of the human body, thereby ensuring the accuracy and safety of subsequent posture recognition, comfort surface calculation, and airbag adjustment command generation, fundamentally avoiding erroneous adjustments caused by misjudgment of foreign objects. The spatiotemporal anomaly features of pressure include pressure peaks, high-frequency noise components, and pressure discontinuities with the surrounding area; the compensation correction function is a band-stop filter for high-frequency noise induced by foreign objects in the frequency domain and a smooth diffusion kernel in the spatial domain.

[0037] Figure 4 A block diagram of a mattress zone adaptive adjustment system based on pressure and posture data according to the present invention is shown.

[0038] A second aspect of the present invention also provides a mattress zone adaptive adjustment system based on pressure and posture data. The system includes a memory 401, a processor 402, and a communication interface 403. The memory includes a mattress zone adaptive adjustment method program based on pressure and posture data. The communication interface is used for data connection and communication between the memory and the processor. When the processor executes the mattress zone adaptive adjustment method program based on pressure and posture data, it performs the following steps: The raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress are obtained. By obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, the pressure sensing error of each zone is calibrated according to the physical structure difference parameters, and the pressure sensitivity coefficient of each zone is determined. A cross-regional pressure data correction model is constructed based on the pressure sensitivity coefficient. The real-time pressure time series data during the user's sleep process is corrected based on the pressure data correction model to obtain standardized pressure time series data. Extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; The user's current sleep posture and torso posture angle are identified based on the effective pressure-bearing boundary and the standardized pressure time-series data. The gravity distribution characteristics of the user's current sleeping posture and torso posture angle are analyzed, the morphological deviation between each airbag zone and the user's target comfort support surface is calculated, and airbag zone air pressure adjustment commands are generated based on the morphological deviation.

[0039] This invention discloses a method and system for adaptive adjustment of mattress zones based on pressure and posture data. The method involves calibrating the pressure sensitivity coefficients of each zone using physical structure difference parameters, constructing a pressure data correction model, and standardizing the original pressure time-series data. Based on the standardized data, the pressure change gradient is extracted to identify the effective pressure-bearing boundary of the user's body pressure. Combining this boundary with the pressure data, the user's sleeping posture and trunk posture angle are identified. Furthermore, the gravity distribution characteristics under this posture are analyzed, and the morphological deviation between each airbag zone and the target comfort support surface is calculated. Finally, air pressure adjustment commands for the airbag zones are generated based on the morphological deviation, thereby achieving adaptive and precise adjustment of the mattress support shape and effectively improving sleep comfort.

[0040] Those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0041] Alternatively, if the integrated units of this invention are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the embodiments of this invention, or the parts that contribute to the prior art, 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 methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, ROM, RAM, magnetic disks, or optical disks.

[0042] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A mattress zone adaptive adjustment method based on pressure and posture data, characterized in that, Includes the following steps: The raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress are obtained. By obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, the pressure sensing error of each zone is calibrated according to the physical structure difference parameters, and the pressure sensitivity coefficient of each zone is determined. A cross-regional pressure data correction model is constructed based on the pressure sensitivity coefficient. The real-time pressure time series data during the user's sleep process is corrected based on the pressure data correction model to obtain standardized pressure time series data. Extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; The user's current sleep posture and torso posture angle are identified based on the effective pressure-bearing boundary and the standardized pressure time-series data. The gravity distribution characteristics of the user's current sleeping posture and torso posture angle are analyzed, the morphological deviation between each airbag zone and the user's target comfort support surface is calculated, and airbag zone air pressure adjustment commands are generated based on the morphological deviation.

2. The mattress zone adaptive adjustment method based on pressure and posture data according to claim 1, characterized in that, The process involves acquiring raw pressure time-series data collected by pressure sensors in each zone of the airbag mattress, obtaining physical structure difference parameters on the surface of each zone of the airbag mattress, calibrating the pressure sensing error of each zone based on these physical structure difference parameters, and determining the pressure sensitivity coefficient of each zone. Specifically: A pressure test was conducted by applying a preset stepped pressure load covering the range of human body weight to the airbag mattress. The original pressure time sequence data of each independent airbag zone on the surface of the airbag mattress was collected in real time within the preset sampling period of the pressure test by an array-type pressure sensing unit. Simultaneously acquire geometric topological data and surface fabric tension distribution data of the airbag mattress surface, and align the geometric topological data and fabric tension distribution data in time and space to construct physical structure difference parameters of the airbag mattress surface, including differences in airbag mattress surface curvature, airbag seam protrusion height and fabric elastic modulus. The original pressure time series data is correlated with the physical structure difference dataset to identify the pressure signal distortion characteristics caused by surface unevenness, and the deviation between the measured pressure value and the theoretical applied pressure value of each zone under different step pressure is extracted. Based on the deviation, a mapping relationship between pressure sensing error and physical structure difference parameters is constructed. By performing nonlinear fitting on the mapping relationship, the pressure sensitivity coefficient of each airbag zone under different deformation states is determined, and the pressure sensing error of each zone is calibrated.

3. The mattress zone adaptive adjustment method based on pressure and posture data according to claim 1, characterized in that, The process involves constructing a cross-regional pressure data correction model based on the pressure sensitivity coefficient, and then applying the pressure data correction model to correct real-time pressure time-series data during the user's sleep process to obtain standardized pressure time-series data. Specifically: Based on the pressure sensitivity coefficient of each airbag zone, the ratio of the pressure sensitivity coefficient of each airbag zone to that of the error-free pressure sensing state is calculated, and the distribution of the ratio on the surface of the airbag mattress is mapped to the initial pressure compensation coefficient matrix required for each zone. A radial basis function neural network is introduced, with the spatial coordinates of each airbag section as input and the initial pressure compensation coefficient of the corresponding position as the expected output, to construct a nonlinear mapping model from spatial coordinates to pressure compensation coefficient. By using the hidden layers of the radial basis function neural network and radial basis functions such as Gaussian function, the input space is nonlinearly transformed. The pressure sensing error data of the airbag mattress under different deformation states is used as the training set to iteratively optimize the connection weights and basis function center and width parameters of the network to form a pressure data correction model. Real-time pressure time-series data during the user's sleep process is acquired, and the real-time pressure time-series data is input into the pressure data correction model for correction to obtain standardized pressure time-series data.

4. The mattress zone adaptive adjustment method based on pressure and posture data according to claim 1, characterized in that, The step of extracting the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identifying the effective pressure-bearing boundary of the user's body contact surface with the mattress based on the pressure change gradient, specifically involves: Based on the pressure value difference sequence of adjacent sampling points of the standardized pressure time series data, the pressure change gradient in the neighborhood of each sampling point is calculated according to the pressure value difference sequence. Based on the adaptive threshold segmentation algorithm, regions where the pressure change gradient is greater than the change threshold are marked as pressure transition regions, and regions where the pressure change gradient is less than the change threshold are marked as pressure stable regions. Connect adjacent pressure transition regions to form a continuous boundary region, calculate the integral value of the pressure change gradient in the continuous boundary region, and determine the candidate effective pressure-bearing boundary profile based on the integral value of the pressure change gradient. Based on the pressure value distribution characteristics on the candidate effective pressure bearing boundary profile according to the standardized pressure time series data, the pressure value distribution characteristics are enhanced by the edge detection operator, the gradient abrupt transition points of the pressure value from non-zero to zero are identified, and the gradient abrupt transition points are connected to construct the preliminary effective pressure bearing boundary curve. Based on the spatial distribution characteristics of the pressure stability region, the average pressure on both sides of the preliminary effective pressure-bearing boundary curve is calculated. Regions with average pressure greater than the pressure threshold are identified as pressure-bearing regions, and regions with average pressure less than the pressure threshold are identified as non-pressure-bearing regions. The preliminary effective pressure-bearing boundary curve is then corrected based on the pressure-bearing regions, and the effective pressure-bearing boundary of the contact surface between the user's body and the mattress is output.

5. The mattress zone adaptive adjustment method based on pressure and posture data according to claim 1, characterized in that, The step of identifying the user's current sleep posture and trunk posture angle based on the effective pressure-bearing boundary and the standardized pressure time-series data specifically involves: Obtain multiple standard sleep posture templates, which include standard pressure distribution characteristic areas of various parts of the human body on the mattress and standard trunk axis in supine, left lateral, right lateral, and prone positions. Based on the effective pressure-bearing boundary and the standardized pressure time series data, a real-time pressure distribution matrix of the user's body is constructed, and the spatial pressure distribution similarity between the real-time pressure distribution matrix and each standard sleep posture template is calculated. The standard sleep posture template with the highest similarity to the spatial pressure distribution of the real-time pressure distribution matrix is ​​selected as the user's current sleep posture; The theoretical principal axis direction of the user's torso on the mattress surface is determined based on the user's current sleeping posture, and the spatial deflection angle of the actual pressure distribution relative to the theoretical principal axis is calculated based on the centroid position and higher-order moment characteristics of the pressure distribution within the effective pressure-bearing boundary. A spatial point set registration method based on singular value decomposition is introduced to register the spatial coordinates of the high-pressure point set in the real-time pressure distribution matrix with the coordinates of the standard pressure point set of the corresponding part in the selected standard sleep posture template, and determine the optimal spatial rotation matrix. The rotation angle component of the user's torso in the horizontal axis direction is extracted based on the optimal spatial rotation matrix. The rotation angle component is then fused with the spatial deflection angle, and the user's torso posture angle is calculated by weighted averaging.

6. The mattress zone adaptive adjustment method based on pressure and posture data according to claim 1, characterized in that, The analysis of the gravity distribution characteristics of the user's current sleeping posture and torso posture angle, the calculation of the morphological deviation between each airbag zone and the user's target comfort support surface, and the generation of airbag zone air pressure adjustment commands based on the morphological deviation are as follows: Based on the standardized pressure time series data, the pressure distribution characteristics of the user's body on the mattress surface are extracted, and a biomechanical coupling model of the human spine-mattress contact surface is established according to the user's current sleeping posture and trunk posture angle. The finite element analysis method is introduced, and the trunk posture angle, pressure distribution characteristics and preset physiological curvature parameters of various parts of the human body are used as inputs to calculate the ideal body pressure distribution required to maintain the natural physiological curvature of the human spine under the user's current sleeping posture. Based on the ideal body pressure distribution, a target comfortable support surface is constructed in the mattress surface space using a thin plate spline interpolation algorithm. Spatially register the target comfort support surface with the current shape surface of the airbag mattress, calculate the elevation difference between the center point of each airbag section on the current shape surface and the corresponding point on the target comfort support surface, and obtain the shape deviation of each airbag section. Based on the morphological deviation of each airbag section, the required air pressure adjustment amount for the corresponding airbag section is generated, and an air pressure adjustment command for each airbag section is generated to adjust the morphology of each airbag section.

7. A mattress zone adaptive adjustment system based on pressure and posture data, characterized in that, The mattress zone adaptive adjustment system based on pressure and posture data includes a storage device and a processor. The storage device includes a mattress zone adaptive adjustment method program based on pressure and posture data. When the processor executes the mattress zone adaptive adjustment method program based on pressure and posture data, it performs the following steps: The raw pressure time-series data collected by the pressure sensors of each zone of the airbag mattress are obtained. By obtaining the physical structure difference parameters of the surface of each zone of the airbag mattress, the pressure sensing error of each zone is calibrated according to the physical structure difference parameters, and the pressure sensitivity coefficient of each zone is determined. A cross-regional pressure data correction model is constructed based on the pressure sensitivity coefficient. The real-time pressure time series data during the user's sleep process is corrected based on the pressure data correction model to obtain standardized pressure time series data. Extract the pressure change gradient of adjacent sampling points in the standardized pressure time series data, and identify the effective pressure-bearing boundary of the contact surface between the user's body and the mattress based on the pressure change gradient; The user's current sleep posture and torso posture angle are identified based on the effective pressure-bearing boundary and the standardized pressure time-series data. The gravity distribution characteristics of the user's current sleeping posture and torso posture angle are analyzed, the morphological deviation between each airbag zone and the user's target comfort support surface is calculated, and airbag zone air pressure adjustment commands are generated based on the morphological deviation.

Images