Mattress comfort evaluation method and system based on force line balance system

Abstract

This invention relates to the field of bedding comfort testing and evaluation. A mattress comfort evaluation method based on a force-line balance system includes the following steps: Step S1, obtaining a subjective sleep quality score; Step S2, collecting data on spinal morphology maintenance, pressure distribution, and muscle relaxation; Step S3, establishing a BP neural network evaluation model; Step S4, repeating Step S2 for other mattresses to be evaluated, inputting them into the trained BP neural network evaluation model, and the model outputs the corresponding sleep quality score. Based on this sleep quality score, a quantitative evaluation of mattress comfort is achieved.

Description

Technical Field

[0001] This invention relates to the field of bedding comfort testing and evaluation, specifically to the testing and evaluation of mattress comfort. Background Technology

[0002] Human force lines refer to the biomechanical transmission paths formed by the coordinated action of bones, joints, and muscles during static or dynamic activities. Their core function is to maintain postural stability and movement efficiency, specifically involving three key aspects: gravity distribution, muscle tension balance, and skeletal joint alignment. These directly impact human athletic performance, spinal health, and overall physical well-being. Force line balance, as a core indicator of human biomechanical health, requires the coordinated action of bones, joints, and muscles to ensure the body is in a proper biomechanical state, reducing chronic damage to bones, joints, and muscles.

[0003] Sleep is an important daily physiological activity for the human body. During sleep, the body is in a static and relaxed state. At this time, the force state of the bones, joints and muscles directly affects the balance of the body's force line. As the most important item for sleep, the reasonable design of the mattress directly determines whether the body's force line can maintain balance during sleep.

[0004] Based on the core requirements of the force line balance system, mattress design needs to be zoned while maintaining appropriate firmness to ensure that the spine is in a state of force line balance in both postures.

[0005] Currently, there is a lack of a scientific and systematic verification method to test whether a mattress can effectively maintain the body's force line balance during sleep. Summary of the Invention

[0006] The purpose of this invention is to provide a mattress comfort evaluation method based on a force line balance system, so as to achieve a scientific, quantitative and accurate evaluation of mattress comfort, and provide a reliable basis for product optimization and consumer choice.

[0007] Another objective of this invention is to provide a mattress comfort evaluation system based on a force line balance system.

[0008] A mattress comfort evaluation method based on a force line balance system is characterized by the following steps: Step S1: Obtain the user's subjective sleep quality score when using the mattress to be tested; Step S2: Collect three types of objective characteristic parameters related to the mattress and force line balance. The three types of objective characteristic parameters include spinal shape maintenance, pressure distribution performance, and muscle relaxation. Step S3: Establish a BP neural network evaluation model: Using the spinal morphology maintenance degree, maximum pressure, average pressure, contact area, root mean square value of electromyography, and average power frequency of electromyography collected in step S2 as input feature parameters, and the sleep quality score obtained in step S1 as output parameters, construct a BP neural network model. Step S4, Mattress Comfort Evaluation: For other mattresses to be evaluated, repeat step S2 to collect six input feature parameters: spinal shape maintenance, maximum pressure, average pressure, contact area, root mean square value of electromyography, and average power frequency of electromyography. Input these parameters into the trained BP neural network evaluation model to obtain the sleep quality score for each tester. Take the average score as the final comfort score of the mattress. The model outputs the corresponding sleep quality score, and the mattress comfort is quantitatively evaluated based on this sleep quality score.

[0009] Preferably, the Pittsburgh Sleep Quality Index questionnaire is used, which allows users to fill out the questionnaire after using the mattress to be evaluated for a preset period of time. The subjective sleep score of the mattress to be tested is calculated based on the questionnaire results.

[0010] Preferred method for collecting spinal morphology maintenance data: The angle between the center line of the spine and the horizontal plane when the subject lies on his side on the mattress being tested, as measured by a three-dimensional optical motion capture system. included angle The smaller the value, the better the spinal posture is maintained when using the mattress; Preferably, marker points are affixed to the spinous processes of the occipital protuberance (M1), the seventh cervical vertebra (M2), the second thoracic vertebra (M3), the fifth thoracic vertebra (M4), the seventh thoracic vertebra (M5), the ninth thoracic vertebra (M6), the twelfth thoracic vertebra (M7), the second lumbar vertebra (M8), and the fifth lumbar vertebra (M9) of the subject's spine. The subject lies on their side on a horizontally placed bed in the laboratory, with their body direction parallel to the zOy plane. The angle between the center line of the spine and the horizontal plane when the subject is lying on their side is recorded. Preferably, the method for collecting pressure distribution performance parameters includes the following steps: Step (1) Place the pressure distribution testing system on the mattress to be tested and test the pressure distribution value of the subject on the mattress to be tested; Step (2) performs mathematical processing on the measured pressure distribution values ​​to calculate the maximum pressure. Average pressure Contact area Among them, the maximum pressure The average pressure is the maximum value among the measured pressure values. The average value of the measured pressure, and the contact area. It refers to the contact area between the test subject and the mattress.

[0011] Preferably, the method for collecting muscle relaxation parameters is as follows: a surface electromyography (EMG) sensor is used to collect EMG signals from the erector spinae, quadratus lumborum, latissimus dorsi, and trapezius muscles. After preprocessing the EMG signals, the evaluation indicators are time-domain correlation indicators and frequency-domain correlation indicators. The time-domain correlation indicators are the root mean square (RMS) value or integrated electromyography (iEMG), and the frequency-domain correlation indicators are the average power frequency (MPF) or median frequency (MF).

[0012] Preferably, the BP neural network model includes an input layer, a hidden layer, and an output layer. The input layer has 6 neurons, corresponding to 6 input feature parameters: spinal morphology maintenance, maximum pressure, average pressure, contact area, RMS electromyography, and MPF electromyography. The output layer has 1 neuron, corresponding to the sleep quality score, and uses a linear activation function. The number of neurons in the hidden layer is determined by trial and error, ranging from 10 to 20. During model training, gradient descent is used to optimize the network weights and thresholds, with a learning rate of 0.01-0.05 and 1000-5000 iterations, until the model error is less than a preset threshold, completing the model training and obtaining the trained BP neural network evaluation model.

[0013] A mattress comfort evaluation system based on a force line balance system is characterized by including a data acquisition module, a data preprocessing module, a model training module, and an evaluation module. The data acquisition module is used to collect subjective sleep quality scores, spinal morphology detection units, pressure distribution detection units, and electromyography signal acquisition units when users use the mattress. The data preprocessing module is used to process the raw data collected by the data acquisition module, including spinal morphology maintenance data and pressure distribution data, to obtain the angle α between the spinal centerline and the horizontal plane, maximum pressure, average pressure, and contact area; to filter, denoise, and normalize the electromyographic signals, and to extract the electromyographic RMS and MPF parameters; after the preprocessing is completed, the data is transmitted to the model training module. The model training module is used to construct a BP neural network model, receive the spinal morphology maintenance degree, maximum pressure, average pressure, contact area, electromyography RMS, and electromyography MPF output by the data preprocessing module, output sleep quality score, train and optimize the BP neural network model, and obtain a trained BP neural network evaluation model. The evaluation module is used to receive the preprocessed input feature parameters corresponding to the mattress to be evaluated, input them into the trained BP neural network evaluation model, obtain the sleep quality score output by the model, output the comfort evaluation result according to the scoring criteria, and store and display the evaluation result.

[0014] Preferably, the spinal morphology detection unit uses a three-dimensional optical motion capture system to detect the three-dimensional spatial coordinates of marker points on the spinal centerline when the user is in a lateral lying position. The pressure distribution detection unit employs an array-type pressure distribution testing system to detect pressure distribution data when a user uses the mattress. The electromyography (EMG) signal acquisition unit uses a surface electromyography (SEMG) sensor to acquire EMG signals from the user's neck muscles and output raw EMG signal data.

[0015] Beneficial Effects: 1. Comprehensive Evaluation Dimensions, Balancing Subjective Experience and Objective Data: This invention focuses on the force line balance system, combining three objective characteristic parameters—spinal shape maintenance, pressure distribution, and muscle relaxation—with the subjective sleep score of the Pittsburgh Sleep Quality Index questionnaire. This avoids the one-sidedness of relying solely on subjective evaluation or objective parameter evaluation, and can truly and comprehensively reflect the actual comfort of the mattress. 2. Scientific, Quantitative, and Highly Precise Evaluation Method: By extracting the root mean square (RMS) and average power frequency (MPF) of electromyography signals to quantify muscle relaxation, and combining this with spinal shape maintenance and pressure distribution parameters, a backpropagation neural network is used to establish a correlation model between these features and the sleep quality score, achieving a quantitative evaluation of comfort. The evaluation system of this invention solves the problems of subjectivity and low accuracy in existing evaluation methods, making the evaluation results more scientific and reliable. 3. The system has a reasonable structure, is easy to operate, and has strong versatility: The evaluation system of this invention integrates data acquisition, preprocessing, model training, and evaluation functions, forming a complete closed loop. It is simple to operate and applicable to different types and specifications of mattress products. It can provide data support for product optimization for enterprises and provide consumers with objective product comfort references. 4. It is highly practical and has a wide range of applications: The feature parameters collected by this invention are easy to obtain, and after the model is trained, it can quickly evaluate the comfort of mattresses. It is applicable to various scenarios such as mattress product development, quality testing, market supervision, and consumer selection. Attached Figure Description

[0016] Figure 1 Image showing the pressure distribution while supine; Figure 2 This is an image showing the pressure distribution in a lateral decubitus position. Detailed Implementation

[0017] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below with reference to specific illustrations.

[0018] The mattress comfort evaluation method based on the force line balance system includes the following steps: Step S1: Obtain the user's subjective sleep quality score while using the mattress under test: The Pittsburgh Sleep Quality Index (PSQI) questionnaire is used. Users complete the questionnaire after using the mattress under test for a preset period of time. The user's sleep quality score is calculated based on the questionnaire results and serves as a subjective reference benchmark for comfort evaluation. The preset period is 7-14 days to ensure users adapt to the mattress under test and to guarantee the authenticity and reliability of the subjective score. Example: 50 subjects are recruited, and each subject uses the mattress under test for 10 consecutive days. After 10 days, subjects complete the Pittsburgh Sleep Quality Index (PSQI) questionnaire to obtain the subjective sleep score of the mattress under test. The questionnaire includes seven dimensions: sleep quality, sleep onset time, sleep duration, sleep efficiency, sleep disorders, hypnotic drugs, and daytime functioning. Each dimension is scored from 0 to 3 points, with a total score range of 0-21 points. The relationship between PSQI total score and sleep quality is as follows: PSQI 0–5: very good sleep quality; PSQI 6–10: fair sleep quality; PSQI 11–15: average sleep quality; PSQI 16–21: very poor sleep quality.

[0019] S2. Collect three types of objective characteristic parameters related to mattress force line balance. These three objective characteristic parameters include spinal shape maintenance, pressure distribution performance, and muscle relaxation. The specific collection method is as follows: S21. Collect parameters for spinal morphology maintenance: Spinal morphology maintenance is measured by a three-dimensional optical motion capture system, which measures the angle between the spinal centerline and the horizontal plane when the subject lies on their side on the mattress being tested. To characterize this, the specific steps are as follows: First, calibrate the three-dimensional motion capture system to determine the orientation of the xyz three-axis coordinate system. Second, attach marker points to the bony landmarks on the spinal centerline of the subject's back (occipital protuberance M1, spinous process of the seventh cervical vertebra M2, spinous process of the second thoracic vertebra M3, spinous process of the fifth thoracic vertebra M4, spinous process of the seventh thoracic vertebra M5, spinous process of the ninth thoracic vertebra M6, spinous process of the twelfth thoracic vertebra M7, spinous process of the second lumbar vertebra M8, spinous process of the fifth lumbar vertebra M9). Have the subject lie on their side on a horizontally placed bed in the laboratory, with their body direction parallel to the zOy plane. The angle between the spinal centerline and the horizontal plane when the subject is lying on their side is recorded. included angle The smaller the value, the better the spinal alignment and the better the spinal shape maintenance when using the mattress in lateral position.

[0020] S22. Collect pressure distribution performance parameters: The specific steps are as follows: (1) Place a pressure distribution testing system on the mattress to be tested and test the pressure distribution values ​​of the subjects on the mattress. The test subjects here are the same ones recruited in step S1. Note that three sleeping positions are required: supine, lateral, and prone. Data will be measured for the three sleeping positions later.

[0021] (2) The measured pressure distribution values ​​are mathematically processed to calculate characteristic values ​​such as maximum pressure, average pressure, and contact area. Among them, the maximum pressure... , represents the maximum value among the measured pressure values. Average pressure This represents the average value of the measured pressure. Contact area It is the contact area between the test subject and the mattress. ; The number of pressure distribution test points, This represents the resolution of the pressure distribution testing system. Maximum pressure indicates whether there is pressure concentration in certain areas. Some experts suggest that the pressure threshold on the human skin surface should not exceed 30 mmHg (4 kPa), otherwise blood circulation in the subcutaneous muscles and soft tissues cannot be guaranteed, leading to pressure sores. If abnormally high pressure points appear in certain areas, it indicates that the mattress provides too much support to some parts of the body while providing too little support to areas that need it, resulting in an unbalanced pressure distribution. In this case, the mattress design does not meet the requirements of a force line balance system. Excessive average pressure indicates that the mattress is generally too hard and has poor pressure distribution. Some test results are shown in the following figures. Figure 1 , Figure 2 As shown.

[0022] S23. Acquiring muscle relaxation parameters: Using surface electromyography (EMG) sensors, the EMG signals of the erector spinae, quadratus lumborum, latissimus dorsi, and trapezius muscles are acquired when the user is supine. After preprocessing the EMG signals (filtering, denoising, and normalization), the evaluation indicators are time-domain correlation indicators (such as root mean square value RMS and integrated electromyography iEMG) and frequency-domain correlation indicators (such as average power frequency MPF and median frequency MF).

[0023] During the test, the electromyographic (EMG) activity levels of the main relevant muscles were measured before the subject used the mattress and after lying still on the mattress for a period of time (e.g., 20 minutes). The root mean square (RMS) and mean power frequency (MPF) of the EMG signals were extracted as core characteristic parameters of muscle relaxation. The RMS reflects the amplitude of the EMG signal; a smaller value indicates lower muscle contraction and better relaxation. The MPF reflects the frequency distribution of the EMG signal; a higher value indicates better muscle relaxation. The specific calculation formula is as follows: These are electromyography (EMG) signal sample values; Indicates the total number of values; It refers to the frequency on the frequency axis; It is the power spectral density of the electromyographic signal.

[0024] An ideal state of force line balance should significantly reduce the resting tension of these muscles, manifested as a decrease in the level of electromyographic activity; at the same time, an ideal state of force line balance should increase the rate of change of electrical potential in muscle activity, manifested as an increase in electromyographic frequency.

[0025] This validation used the MPF (Mean Frequency Detection) value of electromyography (EMG) signals, which represents the average value of the frequency range with the highest power density in the spectrum, reflecting the main distribution frequency of EMG signal energy. MPF value is an important indicator for assessing muscle fatigue. During muscle fatigue, the value typically decreases, which is closely related to physiological changes during muscle activity. Therefore, by monitoring changes in MPF ​​value, the degree of muscle fatigue can be determined. The EMG signal changes of the trapezius, splenius cervicis, and sternocleidomastoid muscles were measured in subjects under conditions of 21±1℃ and 50±5%RH before sleep onset, 20 minutes after sleep onset, and throughout the night.

[0026] S3. Establish a BP neural network evaluation model: Using the spinal morphology maintenance degree, maximum pressure, average pressure, contact area, root mean square value of electromyography (RMS), and average power frequency of electromyography (MPF) collected in step S2 as input feature parameters, and the sleep quality score obtained in step S1 as output parameters, construct a BP neural network model.

[0027] The BP neural network model includes an input layer, a hidden layer, and an output layer. The input layer has 6 neurons (corresponding to 6 input feature parameters, namely spinal morphology maintenance, maximum pressure, average pressure, contact area, RMS electromyography, and MPF electromyography). The output layer has 1 neuron (corresponding to sleep quality score, using a linear activation function). The number of neurons in the hidden layer is determined by trial and error, ranging from 10 to 20. During model training, gradient descent is used to optimize network weights and thresholds, with a learning rate of 0.01-0.05 and 1000-5000 iterations, until the model error is less than a preset threshold (preset threshold is 0.001), completing model training and obtaining a trained BP neural network evaluation model. Example: 50 sets of data (6 input feature parameters + 1 output parameter) were divided into a training set (40 sets) and a test set (10 sets). Gradient descent was used to optimize the network weights and thresholds. The learning rate was set to 0.03, and the number of iterations was 3000 until the model error was less than 0.001, completing the model training. After training, the model was validated using the test set data. The validation results showed that the error between the model's output sleep quality score and the actual score was less than 0.5 points, and the model accuracy met the evaluation requirements.

[0028] S4. Mattress Comfort Evaluation: For other mattresses to be evaluated, repeat step S2 to collect six input feature parameters: spinal shape maintenance, maximum pressure, average pressure, contact area, root mean square value of electromyography (RMS), and average power frequency of electromyography (MPF). Input these parameters into the trained BP neural network evaluation model to obtain the sleep quality score for each tester. Take the average score as the final comfort score of the mattress. The model outputs the corresponding sleep quality score, and the mattress comfort is quantitatively evaluated based on this sleep quality score.

[0029] The comfort evaluation rules are as follows: Predicted PSQI 0–5 points: excellent comfort; predicted PSQI 6–10 points: good comfort; predicted PSQI 11–15 points: average comfort; predicted PSQI 16–21 points: poor comfort.

[0030] A mattress comfort evaluation system based on a force line balance system is used to implement the aforementioned mattress comfort evaluation method based on a force line balance system. The system includes a data acquisition module, a data preprocessing module, a model training module, and an evaluation module. The modules are connected through a data interface to realize data transmission and interaction.

[0031] The data acquisition module is used to collect users' subjective sleep quality scores and three types of objective characteristic parameters when using the mattress, including a questionnaire acquisition unit, a spinal morphology detection unit, a pressure distribution detection unit, and an electromyography signal acquisition unit.

[0032] The questionnaire collection unit is used to present the Pittsburgh Sleep Quality Index (PSQI) questionnaire, receive questionnaire data filled out by users, and calculate a sleep quality score based on the questionnaire data. The questionnaire collection unit uses a tablet computer terminal with a built-in Pittsburgh Sleep Quality Index (PSQI) questionnaire program. Testers fill out the questionnaire through the terminal, and the program automatically calculates and outputs the sleep quality score, while storing the score data in a database.

[0033] The spinal morphology detection unit uses a three-dimensional optical motion capture system to detect the three-dimensional spatial coordinates of marker points on the spinal centerline in the user's supine and lateral positions. Preferably, a Qualisys three-dimensional motion capture system with a sampling frequency of 2000Hz is used.

[0034] The pressure distribution detection unit employs an array-type pressure distribution testing system to detect pressure distribution data when a user uses the mattress. Preferably, the sensor array density is 32×42, the pressure detection range is 0-120 mmHg, the accuracy is ±10%FS, and the sampling frequency is 20Hz, enabling real-time acquisition of pressure distribution data.

[0035] The electromyography (EMG) signal acquisition unit uses a surface electromyography (SEMG) sensor to acquire EMG signals from the user's neck muscles and output raw EMG signal data. Preferably, the sampling frequency is 2000Hz and the input impedance is ≥10MΩ.

[0036] The data preprocessing module is used to process the raw data collected by the data acquisition module, including 1) the angle between the spinal centerline and the horizontal plane. The calculation includes: 1) calculating the maximum pressure, average pressure, and contact area of ​​the pressure distribution parameters; 2) filtering, denoising, and normalizing the electromyographic (EMG) signals to extract the root mean square (RMS) and average power frequency (MPF) of the EMG signals. Preferably, the data preprocessing module uses a microcontroller with a built-in signal processing program to process the spinal morphology maintenance data and pressure distribution data to obtain the angle α between the spinal centerline and the horizontal plane, the maximum pressure, the average pressure, and the contact area; filtering, denoising, and normalizing the EMG signals to extract the EMG RMS and MPF parameters; and transmitting the preprocessed data to the model training module.

[0037] The model training module is used to construct a BP neural network model. It receives input feature parameters (spinal morphology maintenance, maximum pressure, average pressure, contact area, RMS, MPF) and output parameters (sleep quality score) from the data preprocessing module, and trains and optimizes the BP neural network model to obtain a trained BP neural network evaluation model. Preferably, the model training module uses an industrial computer with a built-in BP neural network training program. It can receive data transmitted from the data preprocessing module, construct the BP neural network model, complete the training, optimization, and validation of the model, store the trained model in a model database, and iteratively update the model based on new test data to improve model accuracy.

[0038] The evaluation module receives preprocessed input feature parameters of the mattress to be evaluated, inputs them into a trained BP neural network evaluation model, obtains the sleep quality score output by the model, outputs a comfort evaluation result according to the scoring criteria, and stores and displays the evaluation result. Preferably, the evaluation module includes a data input unit, a model calling unit, and a result display unit; the data input unit receives feature parameter data of the mattress to be evaluated; the model calling unit calls the trained BP neural network evaluation model in the model training module to process the input data and output a sleep quality score; the result display unit uses an LCD screen to display the sleep quality score and the corresponding comfort level, and can also export the evaluation result to an Excel file for easy data statistics and analysis.

[0039] The evaluation system of this invention is simple to operate and highly automated, and can quickly complete data collection, preprocessing, model training and comfort evaluation. It is suitable for mattress product development, quality testing and market supervision.

[0040] The foregoing has shown and described the basic principles and main features of the present invention, as well as its advantages. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A method for evaluating the comfort of a mattress based on a force line balance system, characterized by, Includes the following steps: Step S1: Obtain the user's subjective sleep quality score when using the mattress to be tested; Step S2: Collect three types of objective characteristic parameters related to the mattress and force line balance. The three types of objective characteristic parameters include spinal shape maintenance, pressure distribution performance, and muscle relaxation. Step S3: Establish a BP neural network evaluation model: Using the spinal morphology maintenance degree, maximum pressure, average pressure, contact area, root mean square value of electromyography, and average power frequency of electromyography collected in step S2 as input feature parameters, and the sleep quality score obtained in step S1 as output parameters, construct a BP neural network model. Step S4, Mattress Comfort Evaluation: For other mattresses to be evaluated, repeat step S2 to collect six input feature parameters: spinal shape maintenance, maximum pressure, average pressure, contact area, root mean square value of electromyography, and average power frequency of electromyography. Input these parameters into the trained BP neural network evaluation model to obtain the sleep quality score for each tester. Take the average score as the final comfort score of the mattress. The model outputs the corresponding sleep quality score, and the mattress comfort is quantitatively evaluated based on this sleep quality score.

2. The method of claim 1, wherein the force line balance system-based mattress comfort evaluation method is characterized by, The Pittsburgh Sleep Quality Index questionnaire was used, in which users completed the questionnaire after using the mattress to be evaluated for a preset period of time. The subjective sleep score of the mattress was calculated based on the questionnaire results.

3. The force line balance system based mattress comfort evaluation method according to claim 1, wherein, Method for collecting spinal morphology maintenance data: The angle between the center line of the spine and the horizontal plane when the subject lies on his side on the mattress being tested, as measured by a three-dimensional optical motion capture system. included angle The smaller the value, the better the spinal shape is maintained when using the mattress.

4. The method of claim 3, wherein the force line balance system-based mattress comfort evaluation method is characterized by, Marker points were placed on the occipital protuberance (M1) and the spinous process of the fifth lumbar vertebra (M9) along the central line of the spine on the subject's back. The subject was then placed on his / her side on a horizontally placed bed in the laboratory. The angle between the central line of the spine and the horizontal plane was measured when the subject was lying on his / her side. .

5. The force line balance system based mattress comfort evaluation method according to claim 1, wherein, The method for collecting pressure distribution parameters includes the following steps: Step (1) Place the pressure distribution testing system on the mattress to be tested and test the pressure distribution value of the subject on the mattress to be tested; Step (2) performs mathematical processing on the measured pressure distribution values ​​to calculate the maximum pressure. Average pressure Contact area Among them, the maximum pressure The average pressure is the maximum value among the measured pressure values. The average value of the measured pressure, and the contact area. It refers to the contact area between the test subject and the mattress.

6. The force line balance system based mattress comfort evaluation method according to claim 1, wherein, Methods for collecting muscle relaxation parameters: Surface electromyography (EMG) sensors were used to collect EMG signals from the erector spinae, quadratus lumborum, latissimus dorsi, and trapezius muscles. After preprocessing the EMG signals, the evaluation indicators were time-domain and frequency-domain correlation indicators. The time-domain correlation indicator was the root mean square (RMS) value or integrated electromyography (iEMG), and the frequency-domain correlation indicator was the average power frequency (MPF) or median frequency (MF).

7. The force line balance system based mattress comfort evaluation method according to claim 1, wherein, The BP neural network model includes an input layer, a hidden layer, and an output layer. The input layer has 6 neurons, corresponding to 6 input feature parameters: spinal morphology maintenance, maximum pressure, average pressure, contact area, RMS electromyography, and MPF electromyography. The output layer has 1 neuron, corresponding to the sleep quality score, and uses a linear activation function. The number of neurons in the hidden layer is determined by trial and error, ranging from 10 to 20. During model training, gradient descent is used to optimize network weights and thresholds, with a learning rate of 0.01-0.05 and 1000-5000 iterations, until the model error is less than the preset threshold, completing the model training and obtaining the trained BP neural network evaluation model.

8. A mattress comfort evaluation system based on force line balance system, characterized by, It includes a data acquisition module, a data preprocessing module, a model training module, and an evaluation module; The data acquisition module is used to collect subjective sleep quality scores, spinal morphology detection units, pressure distribution detection units, and electromyography signal acquisition units when users use the mattress. The data preprocessing module is used to process the raw data collected by the data acquisition module, including spinal morphology maintenance data and pressure distribution data, to obtain the angle α between the spinal centerline and the horizontal plane, the maximum pressure, the average pressure, and the contact area. The electromyography (EMG) signals were filtered, denoised, and normalized to extract the EMG RMS and MPF parameters. After preprocessing, the data is transferred to the model training module; The model training module is used to construct a BP neural network model, receive the spinal morphology maintenance degree, maximum pressure, average pressure, contact area, electromyography RMS, and electromyography MPF output by the data preprocessing module, output sleep quality score, train and optimize the BP neural network model, and obtain a trained BP neural network evaluation model. The evaluation module is used to receive the preprocessed input feature parameters corresponding to the mattress to be evaluated, input them into the trained BP neural network evaluation model, obtain the sleep quality score output by the model, output the comfort evaluation result according to the scoring criteria, and store and display the evaluation result.

9. The force line balance system based mattress comfort evaluation system according to claim 8, wherein, The spinal morphology detection unit uses a three-dimensional optical motion capture system to detect the three-dimensional spatial coordinates of the marker points on the center line of the spine when the user is lying on their side. The pressure distribution detection unit employs an array-type pressure distribution testing system to detect pressure distribution data when a user uses the mattress. The electromyography (EMG) signal acquisition unit uses a surface electromyography (SEMG) sensor to acquire EMG signals from the user's neck muscles and output raw EMG signal data.

Images