A method and system for quantitatively evaluating the comfort of a double-layer bed net mattress based on pressure distribution testing

Abstract

The application discloses a kind of based on pressure distribution test's double-layer bed net mattress comfort quantitative evaluation method and system, the method is through flexible pressure sensing device acquisition subject's three-dimensional pressure matrix sequence when lying, data processing terminal carries out filtering processing to it, decouples into static average pressure matrix characterized macroscopic load and dynamic pressure fluctuation matrix sequence characterized small change, based on static average pressure matrix, system utilizes standard physiological partition template and structure partition layer to establish the index relationship of physiological region and mattress attribute, calculates overall pressure dispersion index, spine alignment index and partition pressure gradient index.Based on dynamic pressure fluctuation matrix sequence, system constructs dynamic pressure gradient vector field, extracts the phase lag angle between biomechanical excitation function and gradient modulus waveform, calculates double-layer bed net collaborative support index, finally, system is weighted and summarized to each index to generate comfort comprehensive score.The application realizes to double-layer bed net mattress static anatomy support performance and dynamic viscoelastic damping characteristic Comprehensive quantitative evaluation.

Description

Technical Field

[0001] This invention relates to the field of mattress performance testing and ergonomic evaluation technology, specifically to a quantitative evaluation method and system for the comfort of a double-layer mattress based on pressure distribution testing. Background Technology

[0002] Double-layer mattresses typically consist of a lower support layer providing basic support and an upper comfort layer providing a snug, tactile feel. Their physical properties directly impact a user's sleep quality and spinal health. To quantify and evaluate mattress comfort performance, the industry needs to establish objective physical testing indicators to accurately reflect the mechanical interaction at the interface between the human body and mattress materials. This is crucial for mattress material selection, structural optimization, and quality control.

[0003] Current mattress comfort assessment technologies primarily employ pressure distribution testing systems. These systems use a flexible thin-film pressure sensor array placed between the subject and the mattress to record the pressure distribution across the contact surface. Conventional testing procedures focus on collecting and analyzing statistical indicators such as maximum pressure, average pressure, contact area, and pressure gradient distribution under resting conditions. This data is used to determine the mattress's ability to distribute body weight and its support effect on the spine. This static pressure mapping-based analysis method is widely used in mattress structural design and firmness grading.

[0004] However, traditional pressure distribution testing methods primarily treat the human body as a constant static load, focusing only on the final pressure distribution pattern under gravity, neglecting the inherent viscoelastic characteristics of the upper comfort materials (such as memory foam and latex) in double-layer mattresses. These materials exhibit hysteresis response and energy dissipation when subjected to external forces, while the human body continuously undergoes physiological micro-movements during rest, accompanied by breathing and heartbeat. Existing assessment techniques struggle to separate these dynamic micro-movement components from the mixed pressure signals, failing to characterize the mattress's dynamic cushioning performance by analyzing the material's phase hysteresis or damping response to these micro-movement signals. Consequently, mattress comfort assessments are limited to the static geometric support level, failing to truly reflect the dynamic mechanical experience of the human body under sleep physiological rhythms. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a quantitative evaluation method and system for the comfort of double-layer mattresses based on pressure distribution testing, which solves the problem of being unable to quantitatively evaluate the dynamic viscoelasticity (or dynamic cushioning) performance of mattresses.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] The first aspect of this invention provides a quantitative evaluation method for the comfort of a double-layer mesh mattress based on pressure distribution testing. The method configures a data acquisition and control unit to control a flexible pressure sensing device at a set sampling frequency to acquire a three-dimensional pressure matrix sequence of a subject in a standard lying position on the double-layer mesh mattress under test. A data processing terminal receives this three-dimensional pressure matrix sequence, performs time-domain filtering on it, and decouples it into a static average pressure matrix characterizing macroscopic pressure distribution and a dynamic pressure fluctuation matrix sequence characterizing minute pressure changes.

[0008] Based on the static average pressure matrix, the data processing terminal imports a preset standard physiological zoning template and the structural zoning layer of the double-layer mattress to be tested, establishing a positional index relationship between human physiological regions and the physical properties of the double-layer mattress. Based on this positional index relationship and the static average pressure matrix, static basic comfort indices are calculated, including the overall pressure dispersion index, spinal alignment index, and zonal pressure gradient index. Based on the dynamic pressure fluctuation matrix sequence, a dynamic pressure gradient vector field is constructed, the phase lag angle between the biomechanical excitation function and the gradient modulus waveform is extracted, and the double-layer mattress collaborative support index is calculated. Finally, the overall pressure dispersion index, spinal alignment index, zonal pressure gradient index, and double-layer mattress collaborative support index are weighted and summarized to generate a quantitative evaluation report containing a comprehensive comfort score.

[0009] Furthermore, in the data decoupling step, the static average pressure matrix is ​​generated by calculating the arithmetic mean of the pressure values ​​of all time frames of the three-dimensional pressure matrix sequence within the acquisition time. The static average pressure matrix is ​​then subtracted from the instantaneous pressure matrix at each moment using matrix subtraction to obtain the dynamic pressure fluctuation matrix sequence. This sequence is stored in signed floating-point format to record the positive and negative changes in pressure fluctuations.

[0010] Furthermore, in the step of establishing the position index relationship, binarization thresholding and morphological closing operations are performed on the static average pressure matrix to determine the effective contact area, and the subject's contact body length is calculated based on the longitudinal coordinate difference of the effective contact area. Based on the percentage proportions of each part defined in the standard physiological zoning template, combined with the contact body length, the absolute boundaries of the head, shoulder and back, lumbar spine, pelvis, thigh, and calf regions in the pressure matrix coordinate system are calculated to generate a physiological zoning index map. Simultaneously, this physiological zoning index map and the structural zoning layer are read, and a composite attribute matrix is ​​constructed by traversing all effective sensing points. This composite attribute matrix associates the physiological attribution parameters of the current coordinate point with the corresponding viscoelastic coefficient of the upper comfort layer and the elastic modulus of the lower support layer.

[0011] Furthermore, the calculation process of the overall pressure dispersion index is as follows: extract the effective pressure dataset containing only the human body contact surface from the static average pressure matrix, count the number of pixels with pressure values ​​greater than the preset capillary occlusion threshold, calculate the proportion of these pixels to the total number of pixels in the contact area, and define it as the high pressure area proportion; at the same time, calculate the arithmetic mean and standard deviation of all pressure values ​​in the effective pressure dataset, divide the standard deviation by the arithmetic mean to obtain the pressure variation coefficient; based on the linear weighted complementary model, perform a weighted summation of the complement of the high pressure area proportion and the complement of the pressure variation coefficient to obtain the overall pressure dispersion index.

[0012] Furthermore, the calculation process of the spinal alignment index is as follows: for each longitudinal coordinate in the effective contact area, calculate the pressure-weighted centroid coordinates on the transverse section to generate a discrete set of original spinal projection points; use a cubic polynomial model to perform curve fitting on the original spinal projection point set to generate a fitted spinal curve; construct a straight line connecting the center of the scapular region and the center of the pelvic region, define it as the ideal alignment baseline, and calculate the root mean square error of the fitted spinal curve relative to the ideal alignment baseline; use a negative exponential decay model to map the root mean square error into a dimensionless spinal alignment index.

[0013] Further, the calculation process of the partition pressure gradient index is as follows: Identify the boundary line between the lumbar spine region and the pelvic region in the physiological partition index map; extend a preset sampling width to both sides of the boundary line, defining it as the lumbar spine side sampling window and the pelvic spine side sampling window, respectively; calculate the average pelvic pressure within the pelvic side sampling window and the average lumbar spine pressure within the lumbar spine side sampling window, respectively; divide the average pelvic pressure by the average lumbar spine pressure to obtain the support gradient ratio. If the support gradient ratio is less than or equal to a preset support gradient judgment threshold, it is determined as reverse support, and the support gradient ratio is multiplied by a preset penalty coefficient to obtain the partition pressure gradient index; if the support gradient ratio is greater than the support gradient judgment threshold, calculate the square of the difference between the support gradient ratio and the optimal target ratio, and calculate the partition pressure gradient index based on the Gaussian decay score.

[0014] Furthermore, the construction process of the dynamic pressure gradient vector field is as follows: extract a single excitation sample from the dynamic pressure fluctuation matrix sequence, perform spatial differentiation operation on each instantaneous pressure matrix in the single excitation sample; calculate the lateral gradient component and longitudinal gradient component of the target pixel using the central difference method; calculate the gradient direction angle and gradient magnitude based on the lateral gradient component and longitudinal gradient component, and synthesize the dynamic gradient vector field sequence. This sequence is used to characterize the spatial deformation linkage and local shear support characteristics of the mattress material when subjected to micro-motion excitation.

[0015] Further, the extraction process of the phase lag angle is as follows: identify the pixel with the highest accumulated gradient energy in a single excitation sample, define it as the excitation center point, extract the pressure change sequence of the excitation center point on the time axis as the excitation center pressure waveform; define an annular region with the excitation center point as the center as the response propagation region, calculate the average pressure of all pixels in the response propagation region at each moment, and generate the response edge pressure waveform; calculate the cross-correlation function between the excitation center pressure waveform and the response edge pressure waveform, determine the material response time lag, and combine it with the main wave period to convert the material response time lag into a phase lag angle.

[0016] Furthermore, the calculation process of the double-layer bed net collaborative support index is as follows: based on the consistency characteristics of the dynamic pressure gradient vector field, the number of stable points with a standard deviation of the orientation angle less than the preset standard deviation threshold of the orientation angle is counted, and the proportion of these stable points to the total number of pixels is calculated to generate a dynamic stability index; the absolute value of the difference between the phase lag angle and the target value of the optimal damping angle is calculated, and the absolute value is mapped to a viscoelastic damping index using a linear subtraction operation; a weighted summation operation is performed on the spinal alignment index, the zoned pressure gradient index, the dynamic stability index, and the viscoelastic damping index to obtain the double-layer bed net collaborative support index.

[0017] A second aspect of this invention provides a quantitative evaluation system for the comfort measurement of a double-layer mesh mattress based on pressure distribution testing. The system includes a flexible pressure sensing device, a data acquisition and control unit, and a data processing terminal. The flexible pressure sensing device is configured to be laid on the upper surface of the double-layer mesh mattress under test and consists of multiple pressure sensing points arranged in a matrix. The data acquisition and control unit is communicatively connected to the flexible pressure sensing device and is used to convert the analog electrical signals output by the flexible pressure sensing device into digital signals, and its internal sampling frequency is not less than 50 Hz. The data processing terminal is connected to the data acquisition and control unit and is used to receive and process the digitized pressure data stream. The data processing terminal is configured to execute the quantitative evaluation method for the comfort measurement of a double-layer mesh mattress based on pressure distribution testing described in the first aspect.

[0018] This invention provides a quantitative evaluation method and system for double-layer mattress comfort based on pressure distribution testing. It offers the following advantages:

[0019] 1. This invention acquires a three-dimensional pressure matrix sequence by configuring a data acquisition and control unit to obtain the three-dimensional pressure matrix sequence at a high-frequency sampling rate. It then uses time-domain filtering and matrix subtraction to decouple the original signal into a static average pressure matrix and a dynamic pressure fluctuation matrix sequence. In addition, it extracts the phase lag angle between the biomechanical excitation function and the gradient modulus waveform, thereby realizing a quantitative evaluation of the viscoelastic damping characteristics of the upper comfort layer in a double-layer mattress. This effectively solves the technical problem that traditional static pressure testing cannot reflect the energy dissipation and dynamic buffering performance of materials under the micro-movement state of the human body, so that the evaluation results can truly reflect the dynamic response capability of the mattress to the human physiological rhythm.

[0020] 2. This invention imports a standard physiological zoning template and a structural zoning layer of a double-layer mesh mattress, and establishes a positional index relationship between human physiological regions and mattress physical properties based on the subject's contact length. It constructs a composite attribute matrix containing physiological attribution parameters and material mechanical parameters, enabling accurate calculation of the spinal alignment index and zoning pressure gradient index. This allows the evaluation process to identify the spatial matching degree between the subject's anatomical characteristics and the mattress zoning structure, avoiding misjudgments of support performance caused by sleeping posture misalignment or body shape differences. Thus, it provides an objective evaluation standard with anatomical basis for the hardness zoning design and ergonomic matching of double-layer mesh mattresses.

[0021] 3. This invention constructs a dynamic pressure gradient vector field by applying spatial differential operators to the dynamic pressure fluctuation matrix sequence, calculates the gradient direction angle and gradient magnitude to synthesize the vector field sequence, and generates a dynamic stability index by statistically counting stable points based on the consistency characteristics of the vector direction. This enables a quantitative characterization of the risk of local shear deformation and support stability of the mattress surface when subjected to micro-impact from human breathing or heartbeat. Combined with the weighted calculation of the double-layer bed mesh collaborative support index, it can comprehensively reflect the steady-state maintenance capability of the lower support layer and the micro-vibration isolation performance of the upper comfort layer, providing users with a two-dimensional quantification of comfort that covers both static support and dynamic disturbance resistance. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the comfort measurement evaluation method of the present invention. Detailed Implementation

[0023] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] This invention provides a method and system for quantitative evaluation of the comfort of a double-layer mattress based on pressure distribution testing. The system in this embodiment mainly consists of a flexible pressure sensing device, a data acquisition and control unit, and a data processing terminal.

[0025] The flexible pressure sensing device is configured to be laid on the upper surface of the double-layer mattress to be tested, located between the mattress and the subject. The flexible pressure sensing device consists of multiple pressure sensing points arranged in a matrix. The spatial distribution density of the sensing points is set to at least one sensing unit per square centimeter, or the center-to-center distance between adjacent sensing units is less than 10 mm, to meet the needs of capturing minute tremors and local pressure gradient changes in the human body. The substrate of the flexible pressure sensing device is made of elastic fabric or polyester film (PET) material with anisotropic conductivity, which allows the flexible pressure sensing device to bend in accordance with the deformation of the mattress surface, reducing the measurement error introduced by the tension of the sensor itself (i.e., the hammock effect). The effective range of each pressure sensing point covers 0 to 200 mmHg, and the pressure resolution is better than 1 mmHg.

[0026] The data acquisition and control unit communicates with the flexible pressure sensing device via a USB interface, Ethernet interface, or Wi-Fi wireless module. Internally, the data acquisition and control unit integrates a multi-channel analog front-end (AFE) and an analog-to-digital converter (ADC) to convert the analog electrical signals output by the flexible pressure sensing device into digital signals. The sampling frequency of the data acquisition and control unit... The sampling frequency threshold is fixed or programmably configured to be no less than 50 Hz, preferably 100 Hz. The basis for setting this sampling frequency threshold is to satisfy the Nyquist sampling theorem to ensure that the human breathing, heartbeat and muscle micro-movement signals in the frequency range of 0.5 Hz to 10 Hz can be completely reconstructed to prevent spectral aliasing. In addition, the data acquisition control unit is also equipped with a hardware anti-aliasing low-pass filter with a cutoff frequency of 20 Hz to filter out power frequency interference and high-frequency electromagnetic noise before analog-to-digital conversion.

[0027] The data processing terminal is connected to the data acquisition and control unit to receive and process digitized pressure data streams. The terminal stores a preset anatomical model of the human body and the structural parameters of the double-layer mattress to be tested. The data processing terminal constructs a three-dimensional pressure matrix sequence from the received data. ,in and This represents the two-dimensional coordinates of the pressure sensing point on the plane of the flexible pressure sensing device. Indicates the sampling time point. This three-dimensional pressure matrix sequence represents the pressure values ​​at corresponding coordinates and time points. This forms the data basis for subsequent static pressure distribution calculations and dynamic phase hysteresis response analysis.

[0028] For the double-layer mesh mattress under test, the double-layer mesh mattress physically consists of an upper comfort layer and a lower support layer. The upper comfort layer is usually made of viscoelastic materials such as memory foam, latex, or soft polyurethane foam; the lower support layer is usually made of elastic materials such as individual pocket spring arrays, whole-layer springs, or hard fiberboard. The data processing terminal identifies and separates the superimposed response characteristics of the upper comfort layer and the lower support layer along the vertical pressure transmission path through subsequent steps.

[0029] Please see the appendix Figure 1 To ensure that the collected pressure distribution data accurately reflects the physical characteristics of the double-layer bed net and the micro-motion response of the human body, this embodiment has strictly standardized the definition of the test objects, environmental conditions, and deployment steps. The specific implementation process includes the following steps:

[0030] S110: Definition of Physical Properties of the Tested Double-Layer Mattress The test object is selected as a mattress with a clearly defined double-layer mechanical structure. Physically, the double-layer structure is defined as comprising an upper comfort layer and a lower support layer. The upper comfort layer is located at the top of the mattress and is made of a high-viscosity polymer material, including slow-rebound memory foam, natural latex, or gel foam. The physical characteristics of the upper comfort layer are time-dependent responses to stress changes, i.e., phase lag and energy dissipation under dynamic loads. This characteristic forms the physical basis for subsequent phase lag response analysis. The lower support layer is located below the upper comfort layer and is made of a material with a high elastic modulus, including individually pocketed spring assemblies, whole-web spring assemblies, or high-density coir fiberboard. The lower support layer primarily provides linear or nonlinear elastic restoring forces conforming to Hooke's Law to maintain the geometric stability of the overall structure.

[0031] S120: Thermodynamic control of the test environment. Given that the storage modulus and loss modulus of the viscoelastic material in the upper comfort layer are highly sensitive to temperature changes, the temperature of the test environment is strictly controlled within the range of 23 degrees Celsius ± 2 degrees Celsius, and the relative humidity is controlled between 40% and 60%. Before the test, the double-layer mattress to be tested is placed in this standard environment and left to stand for at least 4 hours to ensure that the movement state of the polymer chain segments inside the material reaches thermal equilibrium, avoid local hardness differences caused by temperature gradients, and thus ensure the repeatability of the pressure gradient test results.

[0032] S130: Spatial Anchoring and Laying of the Flexible Pressure Sensing Device. The flexible pressure sensing device is laid flat on the upper surface of the double-layer mesh mattress to be tested. During laying, wrinkles and air bubbles on the surface of the flexible pressure sensing device are eliminated through a physical flattening operation. The effective sensing area of ​​the flexible pressure sensing device covers the human sleep projection area of ​​the double-layer mesh mattress to be tested, with a coverage length of not less than 200 cm and a width of not less than 90 cm. To achieve subsequent structural zoning mapping, a unified spatial coordinate system needs to be established during the laying stage. The geometric center point or specific corner point of the double-layer mesh mattress to be tested is set as the physical origin, and the coordinate zero point of the flexible pressure sensing device coincides with this physical origin. For double-layer mesh mattresses with multi-zone structures (such as soft shoulder zones and hard lumbar zones), it is necessary to ensure that the coordinate axes of the flexible pressure sensing device are parallel and aligned with the physical zone boundaries of the mattress. After laying, a tare operation is performed to zero the weight of the flexible pressure sensing device itself and the initial tension generated during laying in the system.

[0033] S140: Subject Screening and Standardization of Micromotion Excitation Sources. Subjects were selected based on Body Mass Index (BMI), with the standard body type group having a BMI range of 18.5 to 23.9. Subjects wore lightweight cotton test suits without rigid fasteners to minimize interference from clothing friction on micromotion signal extraction. During the data acquisition phase, subjects sat on the flexible pressure sensing device, maintaining a standard supine posture.

[0034] Supine position: The subject's body midline is parallel to the mattress long axis, the head is supported by a standard height test pillow, and the cervical spine is in a natural neutral position.

[0035] Side-lying position: The subject's body is perpendicular to the bed plane, with the flexion angle of the hip and knee joints controlled between 30 and 45 degrees. In order to ensure that the micro-motion excitation signal can be effectively identified by the subsequent frequency domain algorithm, the subject is required to remain relatively still and perform steady rhythmic breathing. The breathing rate needs to be controlled between 12 and 20 breaths per minute (corresponding to a fundamental frequency of 0.2 Hz to 0.33 Hz), and the breathing depth should be kept uniform. This steady breathing movement, combined with the heartbeat, generates a continuous ballistic impact force on the subject's body surface, thereby providing the required biomechanical excitation source for the dynamic mechanical response analysis of the double-layer mesh mattress.

[0036] After hardware deployment is complete and before formal testing begins, an initialization calibration procedure needs to be performed to ensure the accuracy of the data output by the flexible pressure sensing device and to establish a mapping relationship between the mattress structure and human physiology. The specific implementation process includes the following steps:

[0037] S210: Sensor Zero-Point Drift Correction and Tare Weight Reduction. With the subject not in contact with the flexible pressure sensing device and no external load applied, the data acquisition control unit initiates reference signal acquisition. The data acquisition control unit continuously reads the initial voltage value of each pressure sensing point in the flexible pressure sensing device. The continuous reading time is set to 5 to 10 seconds. The data processing terminal calculates the arithmetic mean voltage value of each pressure sensing point during the acquisition period and defines the arithmetic mean voltage value as the reference zero-point voltage. .

[0038] Subsequently, the data processing terminal will use the reference zero-point voltage. Stored in the calibration parameter register. In subsequent real-time tests, for any given time... The acquired raw voltage signal The data processing terminal performs differential operations. This eliminates DC bias errors caused by the weight of the flexible pressure sensor, laying tension, and environmental electromagnetic noise.

[0039] S220: Pressure-Voltage Nonlinear Response Calibration. The sensing materials of flexible pressure sensors typically exhibit nonlinear pressure-voltage response characteristics. To obtain accurate physical pressure values, the data processing terminal loads a preset calibration function onto the effective voltage signal. Mapping is performed. The calibration function uses a third-order polynomial fitting model:

[0040]

[0041] in, The value is the calibrated physical pressure (unit: millimeters of mercury, mmHg). , , The pre-determined fitting coefficients, Through a pressure-voltage nonlinear response calibration step, the system controls the linearity error of the pressure measurement value within ±3% in the range of 0 to 200 mmHg, ensuring the waveform fidelity of small pressure fluctuation signals in subsequent phase hysteresis analysis.

[0042] S230: Import and Mapping of Double-Layer Mattress Structure Partition Layers. The data processing terminal imports the structural parameter file of the double-layer mattress to be tested and generates a structural partition layer based on the spatial coordinate system established during the flexible pressure sensing device installation stage. Structure partition layer The distribution of physical properties of the bunk bed mattress under test on a two-dimensional plane was defined, specifically including:

[0043] Hardness zone boundary coordinates: Determine the start and end Y-axis coordinates of different support areas (including head area, shoulder and back area, lumbar spine area, pelvic area, and leg area) of the double-layer mesh mattress under test in the longitudinal direction.

[0044] Interlayer mechanical property labeling: Each zone is associated with the viscoelastic damping coefficient level of the upper comfort layer and the elastic modulus level of the lower support layer. The import and mapping steps of the double-layer mattress structure zone layers realize the digitization of the physical structure of the double-layer mattress under test, enabling subsequent algorithms to identify the material property background corresponding to each pressure sensing point.

[0045] S240: Preloading of Physiological Zoning Templates: Data processing terminal preloads standard physiological zoning templates based on anthropometry. The standard physiological partitioning template defines the relative proportions of various anatomical regions along the long axis of the body in a standard supine position. The specific proportions are set as follows: head region 0% to 13% of total body length, shoulder and back region 13% to 35%, lumbar spine region 35% to 45%, pelvic region 45% to 55%, thigh region 55% to 75%, and lower leg and foot region 75% to 100%. The pre-loaded standard physiological partitioning template provides an initial spatial reference framework for subsequent adaptive partitioning matching based on actual pressure distribution images.

[0046] S250: Sensor Array Status Self-Check and Defect Repair. The data processing terminal automatically detects the connectivity and background noise amplitude of all pressure sensing points. If the background noise amplitude of a pressure sensing point exceeds 1% of the effective range, or outputs a full-range signal under no-load conditions, the data processing terminal marks the pressure sensing point with abnormal background noise amplitude or outputting a full-range signal as an invalid sensing point. To prevent the numerical change of invalid sensing points from interfering with subsequent gradient field calculations (causing abnormally large gradient values), the data processing terminal uses a local median filtering algorithm to repair the invalid sensing points. Specifically, the data processing terminal selects the median of the pressure values ​​of all effective sensing points within a 3×3 neighborhood around the invalid sensing point and assigns the median to the invalid sensing point to complete the numerical replacement. When the status of all pressure sensing points is confirmed to be normal or has been repaired, the data processing terminal outputs a ready signal.

[0047] The evaluation method described in this invention achieves a comprehensive evaluation of the static support performance and dynamic viscoelastic matching performance of a double-layer mattress through serial data acquisition and parallel data processing. The specific implementation process includes the following steps:

[0048] S310: Multidimensional dynamic pressure data sequence acquisition. With the subject maintaining a standard recumbent posture and stable breathing, the data processing terminal sends an acquisition command to the data acquisition control unit. The data acquisition control unit then performs the acquisition at a preset sampling frequency. (100Hz) Continuous scanning of all pressure sensing points of the flexible pressure sensing device, the acquisition process lasts for [duration missing]. Set the time to 30 to 60 seconds to ensure sufficient data collection time. Capable of covering at least 6 to 10 complete respiratory cycles, the data processing terminal receives and stores a three-dimensional pressure matrix sequence containing time dimension information. .

[0049] S320: Signal preprocessing and static / dynamic data decoupling. The data processing terminal processes the original three-dimensional pressure matrix sequence. Time-domain filtering is performed to remove random noise with frequencies higher than 10 Hz. Subsequently, the data processing terminal decouples the filtered data into two independent data streams:

[0050] Static mean pressure matrix The data processing terminal processes the data by analyzing the acquisition time. The static average pressure matrix is ​​obtained by calculating the arithmetic mean of the pressure values ​​across all time frames. Used to characterize the macroscopic pressure distribution in the body.

[0051] Dynamic pressure fluctuation matrix sequence The data processing terminal obtains the data by subtracting the static average pressure matrix from the instantaneous pressure matrix at each moment. Dynamic pressure fluctuation matrix sequence It is used to characterize minute pressure changes caused by breathing and heartbeat.

[0052] S330: Adaptive physiological zoning and structural alignment data processing terminal based on static mean pressure matrix The pressure contour features are analyzed, and a threshold segmentation algorithm based on pressure gradient or the Canny edge detection operator is used to identify the subject's body contour boundaries. The data processing terminal then processes the data according to the standard physiological partition template preloaded in step S240. The proportional relationship will be used to determine the static average pressure matrix. The body is divided into six physiological regions: head, shoulders and back, lumbar spine, pelvis, thigh, and calf. Simultaneously, the data processing terminal, based on the spatial coordinate system of the flexible pressure sensing device, correlates the six physiological regions with the structural partitioning layer generated in step S230. Spatial superposition is performed to establish a positional index relationship between physiological areas and the physical properties of the double-layer mattress (i.e., the viscoelasticity level of the upper comfort layer and the elastic modulus of the lower support layer).

[0053] S340: The static basic comfort index calculation data processing terminal utilizes the static average pressure matrix. Based on the zoning data determined in step S330, multiple static physical indicators are calculated in parallel. The data processing terminal calculates the Global Pressure Dispersion Index (GPDI) to quantify the pressure uniformity between the subject and the double-layer mattress contact surface. The data processing terminal calculates the Spinal Alignment Index (SAI) to assess the ability to maintain spinal shape by comparing the average pressure ratio between the lumbar region and the pelvic region. The data processing terminal also calculates the Zonal Pressure Gradient Index (PPGI) to assess the smoothness of pressure transition between adjacent physiological regions.

[0054] S350: Micro-motion excitation extraction and gradient field construction data processing terminal for dynamic pressure fluctuation matrix sequencing Further signal processing is performed, which is divided into two parallel paths:

[0055] Path 1: Select the pelvic region on the data processing terminal. The sum of pressure values ​​at all sensing points constitutes a time series, which is then bandpass filtered from 0.2 Hz to 2.0 Hz to extract the biomechanical excitation function in the vertical direction. Biomechanical excitation function It characterizes the periodic vertical loading-unloading effect exerted by the human body on the mattress.

[0056] Path 2: Data processing terminal processes dynamic pressure fluctuation matrix sequence For each frame of the image, a spatial difference operator (such as the Sobel operator) is applied to calculate the dynamic pressure gradient vector field. and its modulus field Modulus field Characterizes the dynamic distribution of shear force risk on the mattress surface over time.

[0057] S360: The data processing terminal performs core phase lag analysis for spatiotemporal phase coupling analysis and collaborative support index calculation. The data processing terminal extracts gradient modulus waveforms within the region of interest where the average gradient modulus is highest. To ensure the physical meaning of the phase analysis, the data processing terminal uses cross-correlation analysis or fast Fourier transform to extract the gradient modulus waveform. With biomechanical excitation function The phase difference at the fundamental respiratory frequency is defined as the phase lag angle. Phase lag angle This reflects the ability of the upper comfort layer of a double-layer mesh mattress to utilize the viscoelasticity of the material to delay the occurrence time of peak shear force. The data processing terminal is based on the phase hysteresis angle. And the stability of gradient modulus, calculate the double-layer bed network cooperative support index (DLSSI).

[0058] S370: The data processing terminal generates a comprehensive assessment report by weighting and summarizing the calculated Overall Pressure Dispersion Index (GPDI), Spinal Alignment Index (SAI), Zonal Pressure Gradient Index (PPGI), and Double Layer Bed Network Synergistic Support Index (DLSSI). The data processing terminal generates a quantitative assessment report that includes a static pressure distribution heatmap, dynamic gradient change curve, and comprehensive comfort score, and presents it to the user through the display unit.

[0059] This embodiment ensures traceability and temporal synchronization in the conversion process from physical sensor signals to final evaluation indicators by defining a well-defined data structure and transmission path. The signal flow process mainly includes four stages: raw signal acquisition, preprocessing and splitting, feature calculation, and comprehensive evaluation.

[0060] During the initial signal acquisition phase, each pressure sensing point of the flexible pressure sensing device generates an analog voltage signal in real time. The multiplexer (MUX) in the data acquisition control unit reads the analog voltage signals sequentially according to the scanning instructions and converts the analog voltage signals into digital voltage frames through an analog-to-digital converter (ADC). Each frame of digital voltage data contains the voltage values ​​of all pressure sensing points and a high-precision timestamp generated by a hardware clock. The data acquisition control unit transmits the sequence of digital voltage frames with high-precision timestamps to the high-speed data buffer of the data processing terminal through a universal serial bus (USB) or network communication interface.

[0061] During the preprocessing and splitting stage, the data processing terminal first performs a calibration operation, reading the zero-point reference voltage from the memory. Using nonlinear fitting coefficients, a point-by-point correction operation is performed on the digital voltage frames within the high-speed data buffer to generate a physical pressure matrix flow. Physical pressure matrix flow The data throughput is kept consistent with the sampling frequency. Subsequently, the data processing terminal uses signal splitting logic to stream the physical pressure matrix. Copy and import the two separate storage structures respectively:

[0062] Static Accumulation Buffer: The static accumulation buffer is used to store pressure data for the entire time period. At the end of the data acquisition, the data processing terminal performs time-domain integration and averaging operations on the data in the static accumulation buffer to generate a static average pressure matrix. .

[0063] Dynamic Ring Buffer: A dynamic ring buffer is configured as a fixed-length first-in-first-out (FIFO) queue, with a capacity set to cover two to three respiratory cycles. The dynamic ring buffer is used to maintain the latest time-series data in real time, providing a data source for subsequent sliding window analysis.

[0064] During the feature extraction stage, the signal flow is further subdivided into a static feature extraction path and a dynamic collaborative analysis path. In the static feature extraction path, the data processing terminal invokes the spatial mapping processing module. The spatial mapping processing module simultaneously reads the static average pressure matrix. Pre-set standard physiological zoning template and the structural partition layer of the mattress under test The spatial mapping processing module uses a coordinate transformation algorithm to transform the static average pressure matrix. Standard physiological zoning template and structural partition layers Alignment is performed within the same Cartesian coordinate system, outputting a partitioned pressure dataset with anatomical and mattress material property labels. Based on the partitioned pressure dataset, the data processing terminal calculates the Global Pressure Dispersion Index (GPDI), Spinal Alignment Index (SAI), and Partitioned Pressure Gradient Index (PPGI).

[0065] In the dynamic collaborative analysis path, the data processing terminal extracts data from the dynamic circular buffer and performs excitation extraction and gradient field calculation in parallel. The data processing terminal locks the channel data corresponding to the pelvic region, performs bandpass filtering, and outputs a one-dimensional vertical excitation force waveform sequence. Simultaneously, the data processing terminal applies the Sobel differential operator to each frame of pressure image within the dynamic circular buffer to generate a dynamic gradient vector field sequence containing horizontal and vertical gradient components. Subsequently, the data processing terminal receives the vertical excitation force waveform sequence. With dynamic gradient vector field sequence Cross-correlation analysis is performed, and the data processing terminal calculates the dynamic gradient vector field sequence. The peak modulus relative to the vertical excitation force waveform sequence Peak time delay The data processing terminal operates according to the formula:

[0066]

[0067] (in (for the respiratory base frequency) time delay Converted to phase lag angle .

[0068] During the comprehensive evaluation phase, the data processing terminal performs weighted fusion calculations. The data processing terminal aggregates GPDI, SAI, and PPGI values ​​from the static feature extraction path, as well as the phase lag angle from the dynamic collaborative analysis path. In conjunction with the Double Bed Network Collaborative Support Index (DLSSI), the data processing terminal uses a linear weighted summation model to calculate the final comprehensive comfort score. :

[0069]

[0070] in, , , , The preset weighting coefficients, and The data processing terminal generates a heat map of static pressure distribution, a dynamic gradient change curve, and a comprehensive comfort score. The quantitative evaluation report is generated and rendered and output through the display unit. Through the above signal flow mechanism, a closed-loop evaluation of the dual characteristics of "static support-dynamic buffering" of the double-layer mattress is realized.

[0071] The dynamic micro-motion data acquisition strategy aims to capture mattress mechanical response signals triggered by physiological micro-motions from the subject's static lying state by combining high-frequency sampling with time-domain filtering. The specific implementation process includes the following steps:

[0072] S410: The high-frequency oversampling data processing terminal for the micro-motion signal sends a configuration command to the data acquisition control unit, setting the sampling frequency of the flexible pressure sensor to 100 Hz. Setting the sampling frequency to 100 Hz provides sufficient time resolution in subsequent phase lag analysis, preventing phase distortion of the respiratory fundamental signal. The data acquisition control unit continuously reads the voltage value of each pressure sensing point in the flexible pressure sensor within the acquisition time window, generating the original pressure time series. .

[0073] S420: Bandpass filtering extraction data processing terminal for physiological frequency bands on raw pressure time series. Digital filtering is performed. To maintain the phase linearity of the waveform (i.e., without introducing nonlinear group delay) while separating the respiratory signal, the data processing terminal constructs and applies a linear-phase finite-impulse response (FIR) bandpass filter. The passband frequency range of the FIR bandpass filter is set to 0.15 Hz to 0.8 Hz. The lower cutoff frequency of 0.15 Hz is used to filter out extremely low-frequency baseline drift caused by changes in the subject's muscle tone, and the upper cutoff frequency of 0.8 Hz is used to filter out heartbeat signals and power line interference. The data processing terminal processes the raw pressure time series... Input a linear-phase finite-impulse-response bandpass filter, output a respiratory component pressure sequence containing only the respiratory excitation component. Respiratory component pressure sequence It includes data on the periodic fluctuations in local pressure caused by the movement of the subject's chest and abdomen.

[0074] S430: Separation of Heartbeat Vibration Signals (Optional Step) In embodiments requiring evaluation of the upper comfort layer's absorption capacity for high-frequency micro-vibrations, the data processing terminal applies a bandpass filter with a passband range of 0.8 Hz to 3.0 Hz. The data processing terminal extracts the original pressure time series... Extracting heart rate component stress sequence Heart rate component stress sequence It characterizes the transmission characteristics of the ballistic impact force generated by heartbeats on the surface of a double-layered mattress.

[0075] S440: Detection and Removal of Motion Artifacts To prevent macroscopic limb movements (such as turning over or raising an arm) during the data acquisition process from interfering with phase analysis, the data processing terminal performs motion artifact detection. First, the terminal calculates the total pressure value at all valid sensing points of the flexible pressure sensor at each moment. Then, it calculates the absolute value of the first difference of the total pressure value over time (i.e., the rate of change of total pressure per unit time). If the absolute value of the first difference of the total pressure value exceeds a preset artifact threshold (e.g., 500 mmHg / s), the terminal marks the time interval before and after the corresponding moment as invalid. In subsequent phase lag analysis, the terminal automatically removes the data marked as invalid segments, retaining only the respiratory component pressure sequence during stable breathing periods. .

[0076] After obtaining the raw pressure time series, the data processing terminal needs to decouple the composite signal, which mixes human body weight (static component) and physiological micro-motion (dynamic component), and then distribute the decoupled data to different storage structures to meet the subsequent parallel computing requirements. The specific implementation process includes the following steps:

[0077] S510: Real-time data stream physical quantity mapping. The data processing terminal receives the raw digital voltage frame sequence from the data acquisition and control unit. The data processing terminal calls the nonlinear calibration curve and performs voltage-pressure conversion operation on each sensing point in each frame of data. The data processing terminal reassembles the converted data into an instantaneous physical pressure matrix stream. Instantaneous physical pressure matrix flow The numerical units are uniformly set to millimeters of mercury (mmHg), and the instantaneous physical pressure matrix flow... The frame rate is consistent with the hardware sampling frequency.

[0078] S520: Construction of the Static Reference Surface To obtain a pressure distribution map reflecting the subject's macroscopic body shape characteristics and static support state, the data processing terminal performs a time-domain integral averaging operation. The data processing terminal allocates a full-size accumulator in memory. The matrix dimension of the full-size accumulator is consistent with the sensor array dimension of the flexible pressure sensing device, and the data type is set to high-precision floating-point numbers. During the acquisition window, the data processing terminal streams the instantaneous physical pressure matrix for each frame. The values ​​are accumulated into a full-size accumulator. When the acquisition ends, the data processing terminal divides the values ​​in the full-size accumulator by the total number of sampling frames to generate a static average pressure matrix. Static average pressure matrix All high-frequency components that fluctuate over time were filtered out, and only the constant gravitational load distribution applied by the subject to the double-layer mesh mattress while in a resting state was retained.

[0079] S530: Differential decoupling data processing terminal for dynamic micro-motion components utilizes static average pressure matrix As a baseline, from the instantaneous physical pressure matrix flow The dynamically changing components are separated from the data. Specifically, the data processing terminal processes each data point at any given time. Perform matrix subtraction on the instantaneous pressure data:

[0080]

[0081] The data processing terminal outputs a dynamic pressure fluctuation matrix sequence through matrix subtraction. To fully record the positive and negative changes in pressure fluctuations (i.e., the increase in pressure during inhalation and the decrease in pressure during exhalation), a dynamic pressure fluctuation matrix sequence is used. The use of signed floating-point format for storage and the technical significance of the differential decoupling step lies in eliminating the large DC component (typically between 20 and 150 mmHg) generated by the body's own weight, thereby preventing the large static background signal from masking the weak AC component (typically only 0.5 to 5 mmHg) caused by breathing and heartbeat, and ensuring the signal-to-noise ratio of subsequent gradient calculations.

[0082] S540: Dual-channel data splitting and cache management. In order to meet the different data structure requirements of static comfort assessment and dynamic phase analysis, the data processing terminal performs splitting and storage operations on the decoupled data.

[0083] Channel 1: Static Feature Storage Channel. The data processing terminal stores the static average pressure matrix generated in step S520. It is stored in non-volatile memory for subsequent calculation of static support indicators.

[0084] Channel 2: Dynamic Feature Buffer Channel. The data processing terminal will process the dynamic pressure fluctuation matrix sequence generated in step S530. Writes are made to a first-in, first-out (FIFO) circular buffer. To support sliding window analysis and prevent memory overflow, the data processing terminal adjusts the depth of the circular buffer. The depth of the annular buffer is calculated based on a minimum respiratory rate of 0.2 Hz and a sampling rate of 100 Hz, and is set to cover three complete respiratory cycles. The frame rate is set to 1500 frames (3×5s×100Hz). When a new data frame is written, if the circular buffer is full, the data processing terminal will automatically overwrite the earliest time frame to ensure that the phase analysis module always obtains the latest continuous micro-motion waveform data.

[0085] The intelligent zoning and structural mapping process aims to establish a spatial correspondence between human anatomical features and the physical properties of double-layer mattresses, providing a data foundation for subsequent assessments of support effectiveness. The specific implementation process includes the following steps:

[0086] S610: Identification and Morphological Correction of Effective Contact Area. The data processing terminal reads the static average pressure matrix. To extract the subject's body contour from background noise, the data processing terminal performs binarization thresholding segmentation. The data processing terminal sets pixels with pressure values ​​less than the effective pressure threshold (e.g., 2 mmHg) in the static average pressure matrix to 0, and pixels with pressure values ​​greater than or equal to the effective pressure threshold to 1, generating an original binary image. Due to the curves of the human back (e.g., lumbar lordosis), some areas may have extremely low contact pressure or even be suspended, resulting in breaks or holes in the original binary image. To obtain a connected and complete body contour, the data processing terminal performs morphological closing operations on the original binary image. The specific execution logic of the morphological closing operation is to first perform a dilation operation on the image, followed by an erosion operation. Through the morphological closing operation, the data processing terminal can fill small holes inside the object and connect adjacent broken areas. After the morphological closing operation, the data processing terminal determines the effective contact area of ​​the subject on the mattress and extracts the minimum and maximum ordinates of the effective contact area along the longitudinal axis of the mattress.

[0087] S620: The adaptive segmentation data processing terminal for physiological regions calculates the subject's contact length based on the difference between the minimum and maximum ordinates. The terminal then calls a preset standard physiological region template. Based on the percentage proportions of each body part defined in the standard physiological region template, and combining the contact length and minimum ordinate, the terminal calculates the absolute boundaries of each physiological region in the pressure matrix coordinate system. Taking the lumbar spine region as an example, the terminal multiplies the contact length by the lumbar spine initiation percentage defined in the standard physiological region template and adds the product to the minimum ordinate to obtain the initiation boundary coordinates of the lumbar spine region. Similarly, the terminal calculates the termination boundary coordinates using the lumbar spine termination percentage. Using the above calculation logic, the terminal sequentially determines the coordinate boundaries of the head, shoulder and back, lumbar spine, pelvis, thigh, and calf regions, and generates a physiological region index map in memory. The value of each pixel in the physiological region index map represents the physiological region code to which the corresponding coordinate point belongs.

[0088] S630: The composite mapping data processing terminal for structure-physiological attributes performs spatial alignment of physiological characteristics with mattress mechanical properties. The data processing terminal simultaneously reads the physiological partition index map and the structural partition layer. By traversing all valid sensing points, the data processing terminal constructs a composite attribute matrix. For each current coordinate point, the data processing terminal associates the following two sets of physical parameters in the composite attribute matrix:

[0089] Physiological attribution parameter: Which human anatomical location the current coordinate point belongs to (e.g., "lumbar spine");

[0090] Material mechanical parameters: viscoelastic coefficient of the upper comfort layer and elastic modulus of the lower support layer in the double-layer mattress area directly below the current coordinate point.

[0091] The construction of the composite attribute matrix realizes the digital alignment of "anatomical requirements" and "material support", providing the necessary parameter index for subsequent calculation of local phase hysteresis based on material viscoelasticity.

[0092] S640: Misalignment Risk Prediction To ensure the validity of the assessment results, the data processing terminal performs misalignment detection before performing complex calculations. The data processing terminal compares the core area (such as the lumbar region) in the physiological zoning index map with the reinforced support area in the structural zoning layer.

[0093] The data processing terminal calculates the spatial overlap between the lumbar spine physiological region and the mattress lumbar support region. Specifically, the terminal counts the number of pixels that simultaneously belong to both the lumbar spine physiological region and the mattress lumbar support region, and calculates the ratio of this number to the total number of pixels in the lumbar spine physiological region. If this ratio is lower than a preset alignment threshold (e.g., 0.5, or 50%), the terminal generates a sleeping posture misalignment marker. This marker will be used as a correction factor in the final comfort score calculation, or it may directly trigger a prompt signal suggesting that the user adjust their sleeping posture.

[0094] The Overall Pressure Dispersion Index (OPI) aims to quantitatively evaluate the ability of a double-layer mesh mattress to eliminate localized high-pressure points and evenly distribute body weight under static support conditions. The specific implementation process includes the following steps:

[0095] S710: Extraction of Effective Pressure Dataset. The data processing terminal reads the static average pressure matrix stored in the static feature storage channel. To eliminate the interference of background noise on the statistical results, the data processing terminal uses effective contact area extraction logic to filter out the effective pressure dataset containing only the human body contact surface from the static average pressure matrix. The data processing terminal counts the total number of pixels in the effective pressure dataset and defines the total number of pixels as the total contact area pixels. .

[0096] S720: Quantitative data processing terminal for high-pressure risk areas sets capillary occlusion threshold. Based on the physiological characteristics of human skin microcirculation, the data processing terminal sets the capillary occlusion threshold. The pressure was set to 32 mmHg. The data processing terminal traversed the effective pressure dataset and counted pressure values ​​greater than the capillary occlusion threshold. The number of pixels was determined, and the statistical result was defined as the number of high-voltage risk pixels. .

[0097] Subsequently, the data processing terminal calculates the percentage of high-voltage area. Specifically, the data processing terminal performs a division operation to calculate the number of high-voltage risk pixels. Total contact area pixels The ratio. Percentage of high-voltage area. It represents the proportion of areas in the mattress support layer that failed to effectively absorb peak pressure.

[0098] S730: Statistical Analysis of Pressure Distribution Uniformity. The data processing terminal performs statistical distribution analysis on the effective pressure dataset to assess the dispersion of pressure across the contact surface. The data processing terminal calculates the arithmetic mean of all pressure values ​​in the effective pressure dataset. and standard deviation .

[0099] To eliminate the influence of differences in body weight among participants on the homogeneity assessment index, the data processing terminal calculated the coefficient of variation (CV) of stress. The data processing terminal then calculated the standard deviation... Divide by the arithmetic mean The pressure coefficient of variation is obtained. As a dimensionless normalized index, the pressure coefficient of variation is used to measure the relative uniformity of pressure distribution.

[0100] S740: Index Synthesis and Normalization Data Processing Terminal Based on High Voltage Area Ratio The overall pressure dispersion index (GPDI) is calculated by combining it with the pressure variability coefficient. To construct an intuitive indicator that a higher value represents better performance, the data processing terminal uses a linear weighted complementary model for calculation.

[0101]

[0102] in, For high-pressure suppression weights, The weights are uniformly distributed. The data processing terminal will suppress the weights with high voltage. Set the weight to 0.6 to distribute the weights evenly. The value is set to 0.4. The data processing terminal stores the calculated overall pressure dispersion index in the assessment report database as a component of the final comfort score.

[0103] The Spinal Alignment Index aims to assess a mattress's ability to maintain support along the body's central axis by analyzing the geometric symmetry and linearity of pressure distribution on the body surface to infer the projected shape of the subject's spine in the horizontal plane. The specific implementation process includes the following steps:

[0104] S810: Extraction of Pressure-Weighted Centroid Trajectory. To transform the two-dimensional static average pressure matrix into a one-dimensional spinal projection curve, the data processing terminal employs a row-by-row pressure-weighted centroid method. The data processing terminal retrieves the effective contact area data.

[0105] For each longitudinal coordinate (corresponding to the length direction of the mattress) within the effective contact area, the data processing terminal calculates the pressure-weighted centroid coordinates on the transverse section. The specific calculation logic is as follows: the data processing terminal traverses all effective pixels in the current row, accumulates the product of the transverse coordinate and the corresponding pressure value to obtain the sum of pressure moments; at the same time, it accumulates the pressure values ​​of all effective pixels in the current row to obtain the total pressure value; finally, it divides the sum of pressure moments by the total pressure value to obtain the pressure-weighted centroid coordinates of the current row.

[0106] The data processing terminal performs the above calculations on all longitudinal coordinates to generate a discrete set of original spinal projection points. The original spinal projection point set represents the line connecting the centers of force along the longitudinal direction of the human back.

[0107] S820: Mathematical Fitting of Physiological Curves The original spinal projection point set usually contains high-frequency noise due to sensor discretization errors and the influence of local clothing wrinkles. In order to restore the true physiological curvature of the spine, the data processing terminal performs polynomial curve fitting on the original spinal projection point set.

[0108] The data processing terminal selects a cubic polynomial model as the fitting function to adapt to the natural physiological curvature of the human spine in supine or lateral positions. The data processing terminal uses the least squares method to calculate the coefficients of the cubic polynomial model, generating a continuous and smooth fitted spinal curve.

[0109] The cubic polynomial model can preserve the nonlinear curvature characteristics caused by pathological scoliosis or improper sleeping posture, and prevent the data smoothing process from masking the abnormal curvature information of the subject's spine.

[0110] S830: Construction of Ideal Alignment Baseline and Deviation Calculation Data Processing Terminal. The data processing terminal constructs a reference baseline to quantify the degree of deviation of the spine. The data processing terminal extracts the physiological zonal index map and locks the two key anatomical anchor points, the center of the scapular region and the center of the pelvic region.

[0111] The data processing terminal constructs a straight line connecting the center of the scapular region and the center of the pelvic region, and defines the straight line connecting the center of the scapular region and the center of the pelvic region as the ideal alignment baseline.

[0112] Subsequently, the data processing terminal calculates the root mean square error (RMSE) of the fitted spinal curve relative to the ideal alignment baseline within the longitudinal range from the shoulder to the hip. The specific calculation logic is as follows: the data processing terminal calculates the square of the difference between the fitted spinal curve coordinates and the ideal alignment baseline coordinates at each sampling point, calculates the arithmetic mean of all squared differences, and then takes the square root of the arithmetic mean to obtain the RMS error. The RMS error value reflects the average distance of the subject's spine from the ideal alignment baseline.

[0113] S840: The normalization mapping of the index. The data processing terminal maps the root mean square error of the physical quantity to a dimensionless index of 0 to 100. The data processing terminal uses a negative exponential decay model to calculate the spine alignment index. The specific calculation logic is as follows: the data processing terminal multiplies the root mean square error by the negative value of the preset sensitivity coefficient (i.e., takes the opposite number) to obtain the exponential term; then it calculates the natural constant e to the power of this exponential term and multiplies the calculation result by the full score constant (e.g., 100).

[0114] The data processing terminal sets the sensitivity coefficient to 0.1. A sensitivity coefficient of 0.1 results in a spine alignment index of approximately 37 points when the root mean square error reaches 10 mm, and a spine alignment index of over 80 points when the root mean square error is less than 2 mm.

[0115] The zoned pressure gradient index aims to assess the pressure transition characteristics of a double-mesh mattress at the junction of different physiological zones (especially between the lumbar spine and pelvis) to determine whether the mattress provides differentiated ergonomic support. The specific implementation process includes the following steps:

[0116] S910: Definition of the sampling window in the boundary area. The data processing terminal calls the physiological zoning index map. In order to accurately assess the pressure change trend between regions, the data processing terminal locks the transition zone between adjacent physiological regions.

[0117] The data processing terminal identifies the boundary between the lumbar spine region and the pelvic region. Centered on the boundary line, the data processing terminal extends a preset sampling width (e.g., 5 cm) along the longitudinal axis of the mattress towards the lumbar spine region, defining it as the lumbar spine side sampling window; simultaneously, it extends the same sampling width towards the pelvic region, defining it as the pelvic side sampling window.

[0118] Setting a sampling window instead of directly using the average value of the entire region can avoid smoothing out the local abrupt changes at the boundary during the full-region averaging operation, thereby ensuring that the data processing terminal focuses on the local pressure comparison on both sides of the boundary.

[0119] S920: The data processing terminal for extracting regional average pressure calculates the arithmetic mean of the pressure of all effective pixels within the lumbar spine sampling window and defines the result as the lumbar spine average pressure. At the same time, the data processing terminal calculates the arithmetic mean of the pressure of all effective pixels within the pelvic side sampling window and defines the result as the pelvic side average pressure.

[0120] S930: The data processing terminal calculates the support gradient ratio based on the average pressure on the pelvic side and the average pressure on the lumbar side. The specific calculation logic is as follows: the data processing terminal performs a division operation, using the average pressure on the pelvic side as the dividend and the average pressure on the lumbar side as the divisor to calculate the support gradient ratio. According to the biomechanical characteristics of the human body in a supine position, the pelvic region, due to the greater weight and protrusion of the bones, should receive a higher reaction force than the lumbar region. Therefore, an ergonomically sound support gradient ratio should be greater than 1.0.

[0121] S940: The data processing terminal supports a gradient ratio mapping of 0 to 100 for partitioned pressure gradient exponents. In order to punish unreasonable pressure distributions, the data processing terminal uses a piecewise penalty function for calculation.

[0122] Section 1: Reverse Support Penalty Logic When the support gradient ratio is less than or equal to 1.0 (i.e., the lumbar pressure is greater than or equal to the hip pressure), the data processing terminal determines that the mattress has a reverse support defect. In this case, the data processing terminal directly multiplies the support gradient ratio by a preset penalty coefficient (e.g., 50) to obtain a lower zone pressure gradient index. This calculation logic ensures that the score will inevitably fail in the case of reverse support.

[0123] Section Two: Positive Support Scoring Logic. When the support gradient ratio is greater than 1.0, the data processing terminal determines that the support direction is correct. The data processing terminal evaluates the degree of closeness between the support gradient ratio and the optimal target ratio. The data processing terminal sets the optimal target ratio to 2.0. The data processing terminal calculates the square of the difference between the support gradient ratio and the optimal target ratio, and calculates the Gaussian decay score based on the square of this difference. The specific calculation logic is as follows: the partition pressure gradient index decreases normally as the support gradient ratio deviates from the optimal target ratio. When the support gradient ratio equals the optimal target ratio, the partition pressure gradient index receives a full score of 100 points; as the deviation between the two increases, the score decays non-linearly.

[0124] The micro-motion excitation signal extraction process is used to separate the dynamic pressure waveform generated by human physiological activities and unconscious micro-movements from a continuous pressure data stream. The specific implementation process includes the following steps:

[0125] S1010: The high-frequency dynamic pressure sequence acquisition and data processing terminal controls the pressure sensor array to acquire pressure data at a preset dynamic sampling frequency. The data processing terminal sets the dynamic sampling frequency to 50 Hz. The dynamic sampling frequency of 50 Hz covers the frequency bands of human respiratory rate (approximately 0.2 Hz to 0.5 Hz), heart rate (approximately 0.8 Hz to 2.0 Hz), and muscle twitching frequency (approximately 5 Hz to 10 Hz). The data processing terminal stores the acquired continuous timestamp pressure data as a dynamic pressure time series.

[0126] S1020: Static gravity baseline removal. The dynamic pressure time series contains static gravity components generated by human body weight. In order to extract dynamic changes, the data processing terminal performs a baseline removal operation. The data processing terminal uses the sliding window averaging method to calculate the instantaneous baseline. The specific calculation logic is as follows: The data processing terminal sets a baseline calculation window (e.g., 2 seconds). The data processing terminal calculates the arithmetic mean of all pressure sample values ​​within the baseline calculation window and defines the arithmetic mean as the static baseline value at the current moment. Subsequently, the data processing terminal performs a subtraction operation, subtracting the static baseline value at the current moment from the original pressure value at the current moment to obtain the baseline-removed dynamic fluctuation signal.

[0127] S1030: Frequency Domain Filtering and Noise Suppression. The data processing terminal performs bandpass filtering on the baseline-removed dynamic fluctuation signal to filter out power frequency interference and sensor thermal noise. The data processing terminal uses a fourth-order Butterworth bandpass filter. The lower cutoff frequency of the fourth-order Butterworth bandpass filter is set to 0.1 Hz, and the upper cutoff frequency is set to 10 Hz. The fourth-order Butterworth bandpass filter has a flat frequency response characteristic in the passband, which can maintain the original waveform characteristics of the micro-motion signal.

[0128] S1040: The data processing terminal for effective excitation event envelope detection identifies micro-motion events with step or pulse characteristics. The terminal calculates the short-time energy envelope of the filtered dynamic fluctuation signal. Specifically, the terminal squares the amplitude of the filtered signal to generate a power sequence. Then, it performs a short-time sliding integral (e.g., with an integration window of 0.2 seconds) on the power sequence to obtain the short-time energy envelope value. The terminal sets an excitation trigger threshold. When the short-time energy envelope value exceeds the threshold, the terminal determines that a valid micro-motion excitation event has been captured. The terminal extracts a data segment from 0.5 seconds before to 2.0 seconds after the effective micro-motion excitation event trigger, marking this segment as a single excitation sample. This single excitation sample records the complete dynamic response process of the mattress material after being subjected to a micro-motion impact from the human body.

[0129] The dynamic pressure gradient vector field is used to transform scalar pressure values ​​into vector descriptions with direction and magnitude, thereby quantifying the spatial deformation linkage and local shear support characteristics of mattress materials under micro-motion excitation. The specific implementation process includes the following steps:

[0130] S1110: Application of Spatial Differential Operator. The data processing terminal reads a single excitation sample, which contains a series of instantaneous pressure matrices arranged in chronological order. The data processing terminal performs spatial differential operation on each instantaneous pressure matrix in the single excitation sample. In order to accurately capture the rate of change of pressure distribution in the horizontal (mattress width direction) and vertical (mattress length direction), the data processing terminal adopts the central difference method. The specific calculation logic is as follows: For any target pixel point inside the instantaneous pressure matrix, the data processing terminal calculates the horizontal pressure difference between the pressure values ​​of the adjacent pixels to the right and the left of the target pixel point, and divides the horizontal pressure difference by twice the horizontal physical spacing of the sensor to obtain the horizontal gradient component. Similarly, the data processing terminal calculates the vertical pressure difference between the pressure values ​​of the adjacent pixels below and above the target pixel point, and divides the vertical pressure difference by twice the vertical physical spacing of the sensor to obtain the vertical gradient component. For pixels at the edge of the instantaneous pressure matrix, the data processing terminal uses a boundary mirroring copy strategy to complete the missing adjacent point data before performing differential calculation to maintain the integrity of the data dimensions.

[0131] S1120: The data processing terminal for synthesizing instantaneous gradient vectors constructs instantaneous gradient vectors based on horizontal and vertical gradient components. The data processing terminal calculates the gradient magnitude of the instantaneous gradient vector. The specific calculation logic is as follows: the data processing terminal calculates the sum of the squares of the horizontal gradient components and the squares of the vertical gradient components, and performs a square root operation on the sum of the squares to obtain the gradient magnitude. The gradient magnitude physically characterizes the steepness of the mattress support surface at the target pixel location. The larger the gradient magnitude, the more severe the material deformation and the more obvious the pulling effect on the neighborhood. The data processing terminal calculates the gradient direction angle of the instantaneous gradient vector. In order to accurately distinguish the specific quadrant of the shear force, the data processing terminal uses the two-parameter arctangent function (atan2) for calculation. The specific calculation logic is as follows: the data processing terminal uses the transverse gradient component and the longitudinal gradient component as input parameters, and determines the specific value of the gradient direction angle in the range of 0 degrees to 360 degrees according to the positive and negative signs of the two components. The gradient direction angle physically indicates the direction in which the material deformation decreases the fastest, that is, the main direction of shear force transmission. Through the above calculation, the data processing terminal transforms the scalar dynamic pressure time series into a vector field sequence containing mechanical transmission information.

[0132] S1130: Temporal Accumulation of Gradient Energy To evaluate the dynamic response strength of the mattress material throughout the entire micro-motion event cycle, the data processing terminal calculates the accumulated gradient energy. The data processing terminal iterates through all time frames of a single excitation sample. For each spatial coordinate point, the data processing terminal accumulates the gradient magnitude of that spatial coordinate point across all time frames. The specific calculation logic is as follows: the data processing terminal sums all gradient magnitude values ​​of a spatial coordinate point over time and defines the summation result as the accumulated gradient energy of that spatial coordinate point. The accumulated gradient energy map reflects the total amount of deformation occurring at various points on the mattress surface during a single micro-motion process. High energy concentration areas correspond to the main stress points of the human body, while the rapid energy decay boundary defines the effective range of the point-to-point support capability of the mattress material.

[0133] S1140: Extraction of Vector Consistency Features. The data processing terminal analyzes the directional stability of the gradient vector on the time axis. The data processing terminal calculates the standard deviation of the orientation angle for each spatial coordinate point within the time period of a single excitation sample. The data processing terminal sets a threshold for the standard deviation of the orientation angle (e.g., 30 degrees). If the standard deviation of the orientation angle of a spatial coordinate point is less than the threshold, the data processing terminal determines that the spatial coordinate point is in the "stable support zone," indicating that the mattress material provides constant directional support at this position. If the standard deviation of the orientation angle of a spatial coordinate point is greater than or equal to the threshold, the data processing terminal determines that the spatial coordinate point is in the "nonlinear distortion zone," indicating that the material has undergone complex nonlinear distortion or lateral slippage during micro-motion.

[0134] The phase lag angle calculation logic aims to quantify the viscoelastic damping ratio of the mattress support layer by analyzing the spatiotemporal delay characteristics of dynamic pressure waves propagating within the mattress material. The specific implementation process includes the following steps:

[0135] S1210: Definition and Extraction of Differential Signal Source To calculate the phase difference, the data processing terminal defines the input and output signal sources in space. The data processing terminal calls a single excitation sample. The data processing terminal identifies the pixel with the highest accumulated gradient energy in the single excitation sample and defines it as the excitation center point. The data processing terminal extracts the pressure change sequence of the excitation center point on the time axis and defines the extracted sequence as the excitation center pressure waveform. The excitation center pressure waveform characterizes the direct compression effect of human body micro-movements on the mattress. Simultaneously, the data processing terminal defines an annular region with an inner radius (e.g., 5 cm) and an outer radius (e.g., 10 cm) centered on the excitation center point, and defines this annular region as the response propagation zone. The data processing terminal calculates the average pressure of all pixels within the response propagation zone at each moment, generating the response edge pressure waveform. The response edge pressure waveform characterizes the delayed response after the mechanical wave is conducted and attenuated by the mattress material.

[0136] S1220: The cross-correlation delay estimation data processing terminal calculates the cross-correlation function between the excitation center pressure waveform and the response edge pressure waveform to determine their time offset. The specific calculation logic is as follows: The data processing terminal keeps the excitation center pressure waveform stationary and slides the response edge pressure waveform point-by-point along the time axis. At each sliding step, the data processing terminal calculates the sum of the dot products of the excitation center pressure waveform and the response edge pressure waveform, generating a cross-correlation coefficient sequence. The data processing terminal retrieves the maximum value in the cross-correlation coefficient sequence and identifies the sliding time corresponding to the maximum value. The data processing terminal defines this sliding time as the material response time lag. The material response time lag reflects the physical delay in the transmission of stress waves from the stress center to the surrounding area.

[0137] S1230: The conversion of the phase lag angle eliminates the influence of differences in respiratory or heart rate among different subjects on the absolute time lag value. The data processing terminal converts the material response time lag into a frequency-independent phase lag angle. The data processing terminal performs a period extraction operation on the excitation center pressure waveform (e.g., through a zero-crossing detection algorithm or a fast Fourier transform algorithm) to obtain the main wave period of the current micro-motion event. The specific calculation logic is as follows: the data processing terminal performs a division operation, dividing the material response time lag by the main wave period to obtain the lag ratio; subsequently, the data processing terminal multiplies the lag ratio by 360 degrees to obtain the phase lag angle.

[0138] S1240: The data processing terminal for determining the damping properties of materials evaluates the viscoelastic characteristics of mattress materials based on the phase hysteresis angle.

[0139] The first judgment logic is as follows: when the phase lag angle is close to 0 degrees (e.g., less than 10 degrees), the data processing terminal determines that the material exhibits pure elastic characteristics. Pure elastic characteristics indicate that the mattress responds quickly and has no significant energy dissipation.

[0140] The second judgment logic: When the phase lag angle is within a preset viscoelastic range (e.g., 10 to 30 degrees), the data processing terminal determines that the material exhibits optimal damping characteristics. Optimal damping characteristics indicate that the mattress material has a balanced ability to absorb micro-impact energy and provide support feedback.

[0141] The third judgment logic: When the phase lag angle is greater than the preset lag threshold (e.g., 45 degrees), the data processing terminal determines that the material exhibits excessive viscous characteristics. Excessive viscous characteristics indicate that the mattress material deforms and recovers slowly, and there is a risk of support layer collapse.

[0142] The double-layer mesh synergistic support index integrates static spinal morphology assessment with dynamic micro-motion response assessment into a comprehensive quantitative index to reflect the synergistic performance of the bottom spring system and the upper comfort layer in a double-layer mesh mattress. The specific implementation process includes the following steps:

[0143] S1310: Dynamic Stability Index Calculation. The data processing terminal acquires the vector field consistency analysis results. To quantify the overall stability of the mattress under micro-motion disturbances, the data processing terminal calculates the dynamic stability index. The data processing terminal counts the number of coordinate points identified as "stable support areas" among all spatial coordinate points corresponding to a single excitation sample, and defines the number of coordinate points identified as "stable support areas" as the stable point count. Simultaneously, the data processing terminal acquires the total pixel count within the effective contact area.

[0144] The specific calculation logic is as follows: the data processing terminal performs a division operation, dividing the stable point count by the total pixel count to obtain the stable region ratio; then, the data processing terminal multiplies the stable region ratio by 100 to obtain the dynamic stability index.

[0145] The dynamic stability index characterizes the proportion of the mattress surface that can maintain a constant shear force direction when subjected to micro-impact.

[0146] S1320: The data processing terminal obtains the phase hysteresis angle by mapping the viscoelastic damping index. In order to convert the physical angle into a percentage evaluation, the data processing terminal calculates the viscoelastic damping index.

[0147] The data processing terminal sets the target value of the optimal damping angle (e.g., 20 degrees) and the damping tolerance range (e.g., 10 degrees). The specific calculation logic is as follows: the data processing terminal calculates the absolute value of the difference between the phase lag angle and the target value of the optimal damping angle. If the absolute value of the difference is less than or equal to the damping tolerance range, the data processing terminal determines that the viscoelastic damping index is 100 points.

[0148] If the absolute value of the difference exceeds the damping tolerance range, the data processing terminal performs a linear deduction calculation. The linear deduction calculation involves subtracting the damping tolerance range from the absolute value of the difference to obtain the excess amount, multiplying the excess amount by a preset deduction coefficient (e.g., 5), and subtracting the product from 100 points to obtain the viscoelastic damping index. This mapping logic ensures that a high score is only obtained when the mattress's material properties are in equilibrium.

[0149] S1330: Construction of the Two-Level Weighted Model. The data processing terminal acquires four sub-indices: spine alignment index, zonal pressure gradient index, dynamic stability index, and viscoelastic damping index. The data processing terminal assigns weight coefficients to the four sub-indices. The weight allocation strategy is as follows:

[0150] Static support layer weights: The data processing terminal sets the weight of the spinal alignment index to 0.4 and the weight of the zoned pressure gradient index to 0.2. These two indicators correspond to the static support capacity of the underlying spring mesh for the human skeleton.

[0151] Dynamic comfort layer weighting: The data processing terminal sets the weight of the dynamic stability index to 0.2 and the weight of the viscoelastic damping index to 0.2. These two indicators correspond to the absorption and isolation capabilities of the upper comfort layer material for micro-movements.

[0152] S1340: The composite index data processing terminal performs a weighted summation operation to generate the double-layer mattress synergistic support index. The specific calculation logic is as follows: the data processing terminal multiplies the spinal alignment index by 0.4, the zoned pressure gradient index by 0.2, the dynamic stability index by 0.2, and the viscoelastic damping index by 0.2; subsequently, the data processing terminal adds these four product terms to obtain the double-layer mattress synergistic support index. The double-layer mattress synergistic support index provides users with an intuitive percentage value, directly reflecting the mattress's comprehensive performance in both static physiological alignment and dynamic micro-vibration isolation.

[0153] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A quantitative evaluation method for the comfort of a double-layer mattress based on pressure distribution testing, characterized in that, Includes the following steps: Obtain the time-series three-dimensional pressure matrix sequence of the subjects in a standard resting position on the double-layer mesh mattress to be tested; The time-series three-dimensional pressure matrix sequence is subjected to signal preprocessing and static / dynamic data decoupling to generate a static average pressure matrix and a dynamic pressure fluctuation matrix sequence. Retrieve the preset standard physiological zone template and the structural zone layer of the double-layer mattress to be tested; associate the static average pressure matrix with the standard physiological zone template and the structural zone layer to establish a spatial mapping relationship between physiological regions and structural attributes; Based on the static average pressure matrix after establishing the mapping relationship, the static basic comfort index is calculated. The static basic comfort index includes the overall pressure dispersion index, the spinal alignment index, and the zone pressure gradient index. Based on the dynamic pressure fluctuation matrix sequence, the micro-motion excitation signal is extracted and a dynamic pressure gradient vector field is constructed. The phase lag angle is calculated, and the double-layer bed network collaborative support index is generated accordingly. Based on the overall pressure dispersion index, spinal alignment index, zoned pressure gradient index, and double-layer bed net collaborative support index, a comprehensive comfort score is generated through weighted fusion calculation.

2. The method for quantitatively evaluating the comfort of a double-layer mattress based on pressure distribution testing as described in claim 1, characterized in that, The signal preprocessing and decoupling from static and dynamic data includes: An arithmetic mean operation is performed on the pressure values ​​of all time frames within the acquisition period to generate the static average pressure matrix used to characterize the macroscopic pressure distribution. Perform matrix subtraction on the instantaneous pressure data at each moment, subtract the static average pressure matrix from the instantaneous pressure data, and generate the dynamic pressure fluctuation matrix sequence to characterize the minute pressure changes caused by breathing and heartbeat. The dynamic pressure fluctuation matrix sequence is stored in a signed floating-point format.

3. The method for quantitatively evaluating the comfort of a double-layer mattress based on pressure distribution testing according to claim 1, characterized in that, The static average pressure matrix is ​​associated with the standard physiological zoning template and the structural zoning layer to establish a spatial mapping relationship between physiological regions and structural attributes, including: Binarization thresholding and morphological closing operations are performed on the static average pressure matrix to identify the effective contact area of ​​the subject, and the minimum and maximum ordinates of the effective contact area along the longitudinal axis of the mattress are extracted. The contact body length is calculated based on the difference between the minimum and maximum ordinates. The coordinate boundaries of the head, shoulder and back, lumbar spine, pelvis, thigh and calf regions are calculated according to the percentage ratio of each part defined in the standard physiological partition template, and a physiological partition index map is generated. The physiological partition index map and the structural partition layer are superimposed in the same Cartesian coordinate system to construct a composite attribute matrix. The composite attribute matrix associates the physiological attribution parameters of each coordinate point with the material mechanical parameters of the double-layer mattress to be tested.

4. The method for quantitatively evaluating the comfort of a double-layer mesh mattress based on pressure distribution testing according to claim 1, characterized in that, The calculation steps for the overall pressure dispersion index include: The effective pressure dataset is extracted from the static average pressure matrix. The number of pixels with pressure values ​​greater than the capillary occlusion threshold is counted, and the proportion of these pixels to the total number of pixels in the contact area is calculated to obtain the high-pressure area ratio. Calculate the arithmetic mean and standard deviation of all pressure values ​​in the effective pressure dataset, and divide the standard deviation by the arithmetic mean to obtain the pressure coefficient of variation; The overall pressure dispersion index is calculated using a linear weighted complementary model based on the high-pressure area ratio and the pressure variation coefficient.

5. A quantitative evaluation method for the comfort of a double-layer mesh mattress based on pressure distribution testing as described in claim 1, characterized in that, The specific steps for calculating the spine alignment index include: For each longitudinal coordinate within the effective contact area, calculate the pressure-weighted centroid coordinates on the transverse section to generate a discrete set of original spinal projection points; A cubic polynomial model is used to perform curve fitting on the original set of spinal projection points to generate a fitted spinal curve. Construct a straight line connecting the center of the scapular region and the center of the pelvic region, and define it as the ideal alignment baseline. Calculate the root mean square error of the fitted spinal curve relative to the ideal alignment baseline. The root mean square error is mapped to the dimensionless spinal alignment index using a negative exponential decay model.

6. The method for quantitatively evaluating the comfort of a double-layer mattress based on pressure distribution testing according to claim 1, characterized in that, The specific steps for calculating the partition pressure gradient index include: Identify the boundary line between the lumbar spine region and the pelvic region in the physiological partition index map, and extend a preset sampling width to both sides of the boundary line as the center, which are respectively defined as the lumbar spine side sampling window and the pelvic side sampling window; The average pressure on the pelvic side within the pelvic side sampling window and the average pressure on the lumbar side within the lumbar side sampling window are calculated respectively. The support gradient ratio is obtained by dividing the average pressure on the pelvic side by the average pressure on the lumbar side. If the support gradient ratio is less than or equal to the preset support gradient determination threshold, it is determined to be reverse support, and the support gradient ratio is multiplied by the preset penalty coefficient to obtain the partition pressure gradient index. If the support gradient ratio is greater than the support gradient determination threshold, the square of the difference between the support gradient ratio and the optimal target ratio is calculated, and the partition pressure gradient index is calculated based on the Gaussian decay score.

7. The method for quantitatively evaluating the comfort of a double-layer mattress based on pressure distribution testing according to claim 1, characterized in that, The specific steps for constructing the dynamic pressure gradient vector field include: Extract a single excitation sample from the dynamic pressure fluctuation matrix sequence, and perform spatial differentiation on each instantaneous pressure matrix in the single excitation sample; The horizontal and vertical gradient components of the target pixel are calculated using the central difference method. Based on the transverse and longitudinal gradient components, the gradient direction angle and gradient magnitude are calculated, and a dynamic gradient vector field sequence is synthesized. The dynamic gradient vector field sequence is used to characterize the spatial deformation linkage and local shear support characteristics of the mattress material when subjected to micro-motion excitation.

8. A quantitative evaluation method for the comfort of a double-layer mesh mattress based on pressure distribution testing according to claim 1, characterized in that, The specific steps for extracting the phase lag angle between the biomechanical excitation function and the gradient modulus waveform include: The pixel with the highest accumulated gradient energy in the single excitation sample is identified and defined as the excitation center point. The pressure change sequence of the excitation center point on the time axis is extracted as the excitation center pressure waveform. A ring-shaped region is defined with the excitation center point as the center, which serves as the response propagation region. The average pressure of all pixels in the response propagation region at each time moment is calculated to generate the response edge pressure waveform. Calculate the cross-correlation function between the excitation center pressure waveform and the response edge pressure waveform, determine the material response time lag, and convert the material response time lag into the phase lag angle by combining the main wave period.

9. A quantitative evaluation method for the comfort of a double-layer mesh mattress based on pressure distribution testing as described in claim 1, characterized in that, The specific steps for calculating the double-layer bed network collaborative support index include: Based on the consistency characteristics of the dynamic pressure gradient vector field, the number of stable points with a standard deviation of the orientation angle less than a preset standard deviation threshold is counted, and the proportion of these stable points to the total number of pixels is calculated to generate a dynamic stability index. Calculate the absolute value of the difference between the phase lag angle and the target value of the optimal damping angle, and map the absolute value to the viscoelastic damping index using a linear subtraction operation; The weighted summation of the spinal alignment index, the zoned pressure gradient index, the dynamic stability index, and the viscoelastic damping index yields the double-layer bed netting collaborative support index.

10. A quantitative evaluation system for the comfort of a double-layer mattress based on pressure distribution testing, characterized in that, include: The flexible pressure sensing device is configured to be laid on the upper surface of the double-layer mattress to be tested, and consists of multiple pressure sensing points arranged in a matrix. The data acquisition and control unit is communicatively connected to the flexible pressure sensing device and is used to convert the analog electrical signal output by the flexible pressure sensing device into a digital signal, and its internal sampling frequency is not less than 50 Hz. A data processing terminal, connected to the data acquisition and control unit, is used to receive and process digitized pressure data streams.

Images