Massage bed body position real-time monitoring method based on multi-mode sensing

Abstract

The invention relates to the technical field of medical instrument monitoring, in particular to a massage bed body position real-time monitoring method based on multi-modal sensing, which comprises the following steps: S1, acquiring a contact surface three-dimensional pressure distribution matrix and body surface temperature field data; s2, generating a correction pressure topological graph; s3, extracting a muscle activation degree index through time-frequency analysis; s4, performing space-time alignment to generate a multi-modal fusion parameter set; s5, calculating a body position offset compensation amount according to the multi-modal fusion parameter set; s6, based on the body position offset compensation amount, a servo executing mechanism is driven to adjust the position and posture of the massage bed surface; according to the massage bed, the human body position state is monitored in real time through the multi-mode sensing technology, the position offset compensation amount is accurately calculated in combination with an ergonomic optimization algorithm, and the position and posture of the massage bed are dynamically adjusted through the servo execution mechanism, so that personalized and accurate position optimization is achieved, and the massage curative effect and the comfort degree of a patient are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical device monitoring, and particularly to a real-time monitoring method for the body position of a massage bed based on multi-modal sensing. Background Art

[0002] During massage treatment and rehabilitation training, the precise adjustment of the human body position has an important impact on the effectiveness of massage techniques and the comfort of patients. At present, massage beds mainly rely on manual adjustment or preset mechanical pose adjustment modes, and it is difficult to adapt to the individual needs of different patients in real time. In addition, traditional body position adjustment methods often lack precise measurement of human physiological characteristics and only rely on simple height adjustment or angle adjustment, unable to effectively combine the biomechanical state of the human body, resulting in body position deviation or unstable support during massage, affecting the massage effect and possibly causing secondary injuries.

[0003] There are still multiple problems in the body position adjustment of massage beds in the prior art. First, there is a lack of precise perception of the contact pressure between the human body and the massage bed, making the body position adjustment lack data support and difficult to accurately match the patient's body shape. Second, the existing body position monitoring methods mainly rely on a single sensor (such as a pressure sensor or a vision sensor), unable to accurately capture the changes in muscle state, resulting in body position adjustment based only on passive support and ignoring the dynamic response of muscle groups. In addition, the existing body position optimization algorithms usually do not consider ergonomic parameters, making it difficult for the body position adjustment scheme to balance physiological comfort and massage mechanics requirements. Summary of the Invention

[0004] Based on the above purposes, the present invention provides a real-time monitoring method for the body position of a massage bed based on multi-modal sensing.

[0005] The real-time monitoring method for the body position of a massage bed based on multi-modal sensing includes the following steps:

[0006] S1: Obtain a three-dimensional pressure distribution matrix of the contact surface through an embedded pressure sensor array, and synchronously collect the body surface temperature field data of the infrared thermal imaging module;

[0007] S2: Perform temperature compensation processing on the pressure distribution matrix to generate a corrected pressure topology map;

[0008] S3: Obtain the electrophysiological characteristic vectors of the target muscle groups based on the electromyography signal sensor group, and extract the muscle activation degree index through time-frequency analysis;

[0009] S4: Align the corrected pressure topology map generated in S2 and the muscle activation degree index extracted in S3 in space and time to generate a multi-modal fusion parameter set;

[0010] S5: Apply an ergonomic optimization algorithm to calculate the body position deviation compensation amount according to the multi-modal fusion parameter set;

[0011] S6: Based on the body position offset compensation amount, drive the servo actuator to adjust the pose of the massage bed surface to achieve body position optimization.

[0012] Optionally, the S1 specifically includes:

[0013] S11: Embed a flexible thin-film resistive pressure sensor array arranged in a matrix within the bed surface of the massage bed. Each sensor unit will output a corresponding resistance signal according to the magnitude of the applied force.

[0014] S12: After the resistance signals generated by each resistive pressure sensor unit are converted into voltage signals through a bridge circuit, they are synchronously sampled and digitally processed through a multi-channel analog-to-digital conversion circuit, and sent to the processor unit through a data interface.

[0015] S13: The processor unit constructs a three-dimensional space pressure distribution matrix of the contact surface between the massage bed and the human body based on the spatial position coordinates of each sensor unit and the corresponding digital pressure values.

[0016] S14: Set an infrared thermal imaging module above the massage bed. The infrared thermal imaging module uses a non-cooled focal plane array infrared detector to collect the infrared radiation energy of the human body surface in real time; and converts the collected infrared radiation energy into body surface temperature data through a built-in calibration function to form a body surface temperature field data matrix.

[0017] Optionally, the expression of the calibration function is: where T is the calculated body surface temperature; E is the infrared radiation energy measured by the infrared detector; ε is the surface emissivity of the measured target; σ is the Stefan-Boltzmann constant; C is the correction coefficient obtained during equipment calibration.

[0018] Optionally, the S2 specifically includes:

[0019] S21: Synchronously register the three-dimensional space pressure distribution matrix and the body surface temperature field data matrix obtained in S1 at the same moment to obtain a data matrix in which the pressure values and the corresponding position temperature values correspond one by one.

[0020] S22: Based on the temperature drift characteristics of the flexible thin-film resistive pressure sensor, construct a fitting mathematical model of the output resistance value of the sensor changing with temperature, and calculate the temperature drift compensation coefficient according to this model.

[0021] S23: Use the temperature drift compensation coefficient to correct each original measurement value of each pressure sensor unit in the three-dimensional space pressure distribution matrix point by point to obtain the pressure measurement value after eliminating temperature interference.

[0022] S24: Based on the corrected pressure measurement values, reconstruct the pressure data on the contact surface of the massage bed through a spatial interpolation algorithm to generate a corrected pressure topology map for reflecting the true contact pressure distribution of the human body.

[0023] Optionally, the specific steps of S3 are as follows:

[0024] S31: Place a surface electromyogram (sEMG) sensor group on the surface of the target muscle group of the human body to collect the original sEMG signal data generated during the contraction of the muscle group, and transmit it to the data acquisition device through a wire;

[0025] S32: Perform band-pass filtering on the collected original sEMG signal data. Use a finite impulse response (FIR) filter to filter the sEMG signals in the range of 20 Hz to 450 Hz to remove baseline drift and high-frequency interference, and obtain the filtered pure sEMG signals;

[0026] S33: Perform frame addition and windowing processing on the filtered sEMG signal data. Set the sliding time window length to 200 ms and the window overlap rate to 50%, and obtain a number of sEMG signal analysis frames;

[0027] S34: Calculate the root mean square value in the time domain based on each analysis frame to obtain an electrophysiological feature vector;

[0028] S35: Perform short-time Fourier transform time-frequency analysis on the filtered sEMG signals, calculate the instantaneous frequency spectrum energy distribution of the sEMG signals, and obtain a frequency domain energy distribution matrix;

[0029] S36: Based on the frequency domain energy distribution matrix, calculate the frequency domain energy integral value of the target muscle group in a specific frequency band to obtain the muscle activation index E m 。

[0030] Optionally, the specific steps of S4 are as follows:

[0031] S41: Based on a unified synchronous clock source, determine the data acquisition timestamps of the corrected pressure topology map generated by S2 and the muscle activation index extracted by S3 to ensure that the two types of data are completely corresponding in the time dimension;

[0032] S42: Based on the human anatomical coordinate system, establish a spatial mapping relationship between the spatial coordinates of the corrected pressure topology map on the massage bed surface and the placement positions of the sEMG sensors of the target muscle group, and obtain a spatial matching relationship matrix between the pressure data and the sEMG signal data;

[0033] S43: Based on the timestamps determined by S41 and the spatial matching relationship matrix determined by S42, perform spatio-temporal data interpolation and fusion on the corrected pressure topology map data and the muscle activation index to generate a spatio-temporally aligned multi-modal fusion data matrix, that is, a multi-modal fusion parameter set.

[0034] Optionally, S5 specifically includes:

[0035] S51: Input the multi-modal fusion parameter set generated in S4 into the ergonomic optimization algorithm to determine the position difference between the actual feature points of the current human body position and the theoretical feature points of the standard ergonomic body position;

[0036] S52: Determine the direction and magnitude of the body position offset based on the position difference in S51, and establish a coordinate reference system based on the human physiological structure to clarify the spatial vector of the offset in this coordinate system;

[0037] S53: Based on the standard ergonomic body position database, set the body position compensation optimization objective function and construct a mathematical optimization problem with the optimization objective of minimizing the deviation between the actual body position feature points and the standard theoretical position;

[0038] S54: Use the particle swarm optimization algorithm to obtain the solution to the optimization problem in S53 through iterative search, obtain the optimal body position offset compensation amount, and obtain a clear optimization value.

[0039] Optionally, S52 specifically includes:

[0040] S521: Based on human anatomical features, with the standard supine position of the human body as the basis, construct a rectangular coordinate system with the center of gravity of the human body as the origin, the head-to-foot direction of the human body defined as the longitudinal axis Z, the left and right sides of the human body defined as the transverse axis X, and the front and back directions of the chest and abdomen of the human body defined as the vertical axis Y;

[0041] S522: Based on the position difference calculated in S51, project the actual feature point coordinate vector Q(x, y, z) into the human physiological structure coordinate system respectively to obtain the difference components corresponding to the X, Y, and Z axis directions, which are denoted as Δx, Δy, and Δz respectively;

[0042] S53: Determine the spatial direction vector V of the body position offset and the magnitude of the offset ΔL according to the components Δx, Δy, and Δz of the position difference in the three axes of the coordinate system. The formulas are: V = (Δx, Δy, Δz) and where: V is the spatial body position offset vector, which clarifies the specific direction of the body position offset; |V| is the modulus of the spatial body position offset vector, indicating the specific magnitude of the body position offset.

[0043] Optionally, S53 specifically includes:

[0044] S531: Obtain the theoretical position coordinates of multiple feature points from the standard ergonomic body position database and define them as the theoretical coordinate point vector set in meters, where the subscript i is the feature point number, i = 1, 2, 3,..., k;

[0045] S532: Obtain the actual measurement coordinate positions corresponding to the above key feature points according to the multi-modal fusion parameter set, which are defined as the set of actual coordinate vectors and denoted as Q i (x i , y i , z i ), with the unit of meter;

[0046] S532: Based on the differences between the actual coordinate points and the theoretical coordinate points, with the goal of minimizing the body position deviation, construct the objective function of the mathematical optimization problem, and the expression is:

[0047] where F is the optimization objective function, representing the overall position deviation between the actual feature point position and the theoretical position of the standard feature point; (x i , y i , z i ) is the spatial coordinate of the i-th actual measurement feature point; (x 0i , y 0i , z 0i ) is the theoretical position coordinate of the i-th standard feature point; i is the serial number of the feature point.

[0048] Optionally, the S6 specifically includes:

[0049] S61: According to the body position offset compensation amount calculated by S5, obtain the pose adjustment target values of the massage table surface in the X, Y, and Z directions, and clarify the distances and angles that the massage table surface needs to move or tilt in each direction;

[0050] S62: Input the pose adjustment target values into the multi-axis servo control system, and parse to obtain the operation instructions of each servo actuator, including the motion direction, displacement amount, and motion speed parameters;

[0051] S63: The multi-axis servo actuators act in sequence according to the operation instructions, and through linear movement and angular rotation, achieve synchronous or step-by-step adjustment of the massage table surface in the three dimensions of X, Y, and Z, ensuring that the table surface pose matches the expected compensation amount.

[0052] Advantages of the present invention:

[0053] In the present invention, through the collaborative work of the embedded pressure sensor array, the infrared thermal imaging module, and the electromyogram signal sensor group, the contact pressure distribution between the human body and the massage table, the body surface temperature field data, and the electrophysiological characteristics of the target muscle groups are obtained in real time; the pressure data is corrected by combining the temperature compensation algorithm, the muscle activation degree index is extracted by time-frequency analysis, and the multi-modal data is fused based on the space-time alignment method, making the body position monitoring more comprehensive and accurate; compared with the traditional single-sensor method, it can accurately reflect the human body position state, ensure the stability of the body position support during the massage process, and improve the patient comfort and treatment effect.

[0054] In the present invention, by combining an ergonomic optimization algorithm, the body position offset compensation amount is calculated according to a multi-modal fusion data matrix, the particle swarm optimization method is used to determine the optimal body position adjustment parameters, and a multi-axis servo actuator is used to achieve precise adjustment of the massage bed posture; by this method, not only can the body position offset be corrected in real time, but also the physiological characteristics of different individuals can be dynamically adapted, the degree of intelligence of body position adjustment can be improved, and the decline of massage mechanical effect or patient discomfort caused by improper body position can be avoided. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only those of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0056] Figure 1 Schematic diagram of the real-time monitoring method for the body position of the massage bed in the embodiment of the present invention;

[0057] Figure 2 Schematic diagram of the method for calculating the body position offset compensation amount in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0058] The present invention will be described in detail below in conjunction with the drawings and specific embodiments. At the same time, it should be noted here that in order to make the embodiments more detailed, the following embodiments are the best and preferred embodiments. For some well-known technologies, those skilled in the art can also adopt other alternative methods for implementation; moreover, the drawings are only for more specific description of the embodiments, and are not intended to specifically limit the present invention.

[0059] It should be pointed out that in the specification, it is mentioned that "one embodiment", "embodiment", "exemplary embodiment", "some embodiments", etc. indicate that the described embodiments may include specific features, structures or characteristics, but not necessarily every embodiment includes the specific features, structures or characteristics. In addition, when combining embodiments to describe specific features, structures or characteristics, implementing such features, structures or characteristics in combination with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the relevant art.

[0060] Generally, terms can be understood at least in part from their use in context. For example, depending at least in part on the context, the term "one or more" as used herein can be used to describe any feature, structure, or property in the singular sense, or can be used to describe a combination of features, structures, or properties in the plural sense. Additionally, the term "based on" can be understood as not necessarily intended to convey an exclusive set of factors, but rather can alternatively, at least in part depending on the context, allow for the existence of other factors that are not necessarily explicitly described.

[0061] As Figure 1 - Figure 2 shown, a real-time monitoring method for the body position of a massage bed based on multi-modal sensing includes the following steps:

[0062] S1: Obtain a three-dimensional pressure distribution matrix of the contact surface through an embedded pressure sensor array, and synchronously collect the body surface temperature field data of the infrared thermal imaging module;

[0063] S2: Perform temperature compensation processing on the pressure distribution matrix to generate a corrected pressure topology map;

[0064] S3: Obtain the electrophysiological feature vectors of the target muscle groups based on the electromyogram signal sensor group, and extract the muscle activation degree index through time-frequency analysis;

[0065] S4: Align the corrected pressure topology map generated in S2 and the muscle activation degree index extracted in S3 in space and time to generate a multi-modal fusion parameter set;

[0066] S5: Apply an ergonomic optimization algorithm to calculate the body position offset compensation amount according to the multi-modal fusion parameter set;

[0067] S6: Based on the body position offset compensation amount, adjust the position and pose of the massage bed surface by driving the servo actuator to achieve body position optimization.

[0068] S1 specifically includes:

[0069] S11: Embed a flexible thin-film resistive pressure sensor array arranged in a matrix within the surface of the massage bed, where each sensor unit outputs a corresponding resistance signal according to the magnitude of the applied force;

[0070] S12: After the resistance signals generated by each resistive pressure sensor unit are converted into voltage signals through a bridge circuit, they are synchronously sampled and digitally processed through a multi-channel analog-to-digital conversion circuit, and sent to the processor unit through a data interface;

[0071] S13: The processor unit constructs a three-dimensional space pressure distribution matrix of the contact surface between the massage bed and the human body based on the spatial position coordinates of each sensor unit and the corresponding digital pressure values;

[0072] S14: An infrared thermal imaging module is set above the massage bed. The infrared thermal imaging module uses a non-cooled focal plane array infrared detector to collect the infrared radiation energy of the human body surface in real time; and converts the collected infrared radiation energy into body surface temperature data through a built-in calibration function, forming a body surface temperature field data matrix, laying a data foundation for subsequent temperature compensation processing. The above steps clearly show how to accurately obtain the pressure distribution data of the contact area between the human body and the massage bed through the embedded pressure sensor array, and how to use the infrared thermal imaging technology to obtain the body surface temperature field data, forming a dual-channel collaborative data acquisition method of pressure and temperature, ensuring the accuracy of subsequent temperature compensation processing and the accuracy of multi-modal fusion data.

[0073] The expression of the calibration function is: where, T is the calculated body surface temperature (unit: K); E is the infrared radiation energy measured by the infrared detector (unit: W / m 2 ); ε is the surface emissivity of the measured target; σ is the Stefan-Boltzmann constant (5.670×10 -8 W / m 2 K 4 ); C is the correction coefficient obtained during equipment calibration (unit: K), used to compensate for environmental effects and sensor system errors; in this solution, the infrared thermal imaging module undergoes multi-point temperature calibration before leaving the factory to obtain the correction coefficient C in different temperature ranges. During temperature measurement, the infrared detector obtains the infrared radiation energy E in real time, combines the preset surface emissivity ε of the target skin, and applies the above calibration function to calculate the body surface temperature T, thereby generating the body surface temperature field data matrix.

[0074] S2 specifically includes:

[0075] S21: Synchronously register the three-dimensional space pressure distribution matrix and the body surface temperature field data matrix obtained in S1 at the same moment to obtain a data matrix in which the pressure value and the temperature value at the corresponding position correspond one by one;

[0076] S22: Based on the temperature drift characteristics of the flexible thin-film resistive pressure sensor, construct a fitting mathematical model of the sensor output resistance value changing with temperature, and calculate the temperature drift compensation coefficient according to this model;

[0077] The fitting mathematical model adopts the polynomial regression model, and the expression is as follows: R(T) = R0·(1 + α·(T - T0) + β·(T - T0) 2 ), where, R(T) is the resistance value of the sensor at temperature T; R0 is the reference resistance value at the reference temperature T0; α is the temperature first-order term coefficient; β is the temperature second-order term coefficient; T is the currently measured body surface temperature; T0 is the reference temperature set during calibration; Let the temperature drift compensation coefficient be K(T), and its calculation formula is:

[0078] S23: Use the temperature drift compensation coefficient to correct the original measurement values of each pressure sensor unit in the three-dimensional space pressure distribution matrix point by point to obtain the pressure measurement values after eliminating temperature interference; the correction formula is as follows: P c = P m ·K(T), where P c is the true pressure value after compensation; P m is the original pressure value measured by the sensor without temperature drift compensation; K(T) is the temperature drift compensation coefficient calculated based on the temperature drift model;

[0079] S24: Based on the corrected pressure measurement values, reconstruct the pressure data on the contact surface of the massage bed through a spatial interpolation algorithm to generate a corrected pressure topology map for reflecting the true contact pressure distribution of the human body; specifically, use the inverse distance weighted interpolation algorithm to calculate the pressure values in the unknown area, and the formula is: where P i is the pressure value after interpolation; P j is the corrected pressure value of the known measurement point; w j is the interpolation weight, and the calculation formula is as follows: where d j is the Euclidean distance between the point to be interpolated and the known measurement point j; p is the weight exponent, and the value of p = 2; through the above method, compensating the temperature drift error of the pressure sensor by means of mathematical modeling and combining the interpolation algorithm to optimize the pressure distribution data can ensure accurate pressure measurement results under different temperature conditions, improve the reliability of the pressure distribution data, and provide more accurate basic data support for subsequent body position optimization and adjustment.

[0080] S3 specifically includes:

[0081] S31: Place a surface electromyogram signal sensor group on the surface of the human target muscle group to collect the original electromyogram signal data generated during the muscle group contraction process and transmit it to the data acquisition device through a wire;

[0082] S32: Perform band-pass filtering on the collected original electromyogram signal data, use a finite impulse response filter to filter the electromyogram signals in the range of 20 Hz to 450 Hz, and remove the baseline drift and high-frequency interference to obtain the pure electromyogram signals after filtering;

[0083] S33: Perform frame-by-frame and windowing processing on the filtered electromyogram signal data, set the sliding time window length to 200 ms, and the window overlap rate to 50%, to obtain several electromyogram signal analysis frames;

[0084] S34: Calculate the root mean square value in the time domain based on each analysis frame to obtain the electrophysiological feature vector, and the calculation formula is as follows: Among them, X rms is the root mean square value of the electromyogram signal analysis frame, with the unit of volt; x n is the amplitude of the electromyogram signal corresponding to the nth sampling point, with the unit of volt; N is the total number of sampling points included in each analysis frame;

[0085] S35: When performing short-time Fourier transform time-frequency analysis on the filtered electromyogram signal, calculate the instantaneous frequency spectrum energy distribution of the electromyogram signal to obtain the frequency domain energy distribution matrix. The calculation formula is as follows: Among them, STFT(t, f) is the time-frequency domain energy distribution matrix obtained by short-time Fourier transform calculation, with the unit of volt; x(m) is the data of the filtered electromyogram signal; w(t - m) is the sliding time window function; t is the time variable, with the unit of second; f is the frequency variable, with the unit of hertz; m is the sampling point serial number;

[0086] S36: Based on the frequency domain energy distribution matrix, calculate the frequency domain energy integral value of the target muscle group within a specific frequency band to obtain the muscle activation index E m , and the calculation formula is as follows: Among them: E m is the muscle activation index, with the unit of square volt; f1 is the lower limit frequency of the electromyogram signal analysis of the target muscle group, with the unit of hertz; f2 is the upper limit frequency of the electromyogram signal analysis of the target muscle group, with the unit of hertz; Through the above method, the electrophysiological characteristics of the target muscle group can be effectively obtained, and the muscle activation index can be extracted by time-frequency analysis, so as to improve the effectiveness and accuracy of the electromyogram signal in real-time body position monitoring and provide high-quality electrophysiological information for subsequent multimodal fusion processing.

[0087] S4 specifically includes:

[0088] S41: According to the unified synchronous clock source, determine the data acquisition timestamps of the calibration pressure topology map generated by S2 and the muscle activation index extracted by S3 to ensure that the two types of data are completely corresponding in the time dimension;

[0089] S42: Based on the human anatomical coordinate system, establish the spatial mapping relationship between the spatial coordinates of the calibration pressure topology map on the massage table surface and the placement positions of the electromyogram signal sensors of the target muscle group to obtain the spatial matching relationship matrix between the pressure data and the electromyogram signal data. The specific mapping matrix calculation formula is as follows: Among them, M is the spatial matching relationship matrix, which is used to clarify the spatial position correspondence relationship between the pressure data and the electromyogram data; x p , y p , z p are respectively the spatial coordinate values of the characteristic positions in the pressure topology map, with the unit of meter; x e , y e , z eThey are the spatial coordinate values of the corresponding positions where the electromyography sensors are placed in the human anatomical coordinate system, with the unit of meter;

[0090] S43: Based on the time stamp determined by S41 and the spatial matching relationship matrix determined by S42, perform spatio-temporal data interpolation and fusion on the corrected pressure topology map data and the muscle activation degree index to generate a spatio-temporally aligned multi-modal fusion data matrix, that is, a multi-modal fusion parameter set, and its calculation formula is: F(t, s) = λ · P c (t, s) + (1 - λ) · E m (t, s), where F(t, s) is the multi-modal fusion data matrix, representing the fusion parameter value at time t and spatial position s; P c (t, s) is the pressure value at time t and spatial position s in the corrected pressure value matrix, with the unit of Pascal; E m (t, s) is the muscle activation degree index value at time t and spatial position s in the muscle activation degree index matrix, with the unit of square volt; λ is the fusion weight coefficient, used to adjust the proportion of pressure data and electromyography data in the fusion, and its value range is from 0 to 1; Through the above step method, the precise alignment of pressure data and electromyography data in the time and space dimensions is achieved, ensuring that the fused multi-modal data has good spatio-temporal consistency, providing an accurate and reliable data basis for calculating the body position offset compensation amount.

[0091] S5 specifically includes:

[0092] S51: Input the multi-modal fusion parameter set generated in S4 into the ergonomic optimization algorithm to determine the position difference between the actual feature points of the current human body position and the theoretical feature points of the standard ergonomic body position;

[0093] Specifically, first extract the spatial coordinate position data of the current human body feature points from the multi-modal fusion parameter set obtained in S4, and define it as the measured feature point coordinate vector Q(x, y, z), with the unit of meter;

[0094] At the same time, extract the theoretical spatial coordinates of the corresponding human body target feature points from the standard ergonomic body position database, and define it as the standard feature point coordinate vector Q0(x0, y0, z0), with the unit of meter;

[0095] Subsequently, calculate the Euclidean distance between the actual feature point coordinate vector Q and the standard feature point coordinate vector Q0 to determine the position difference value Δd between the two, and the calculation formula is as follows:

[0096] Wherein, Δd represents the spatial position difference between the actual feature point and the standard feature point, with the unit of meter; x, y, and z respectively represent the three-dimensional spatial coordinate values of the actual feature point in the ergonomic coordinate system, with the unit of meter; x0, y0, and z0 respectively represent the theoretical spatial position coordinate values of the standard feature point, with the unit of meter;

[0097] S52: Determine the direction and magnitude of the body position offset based on the position difference in S51, and establish a coordinate reference system based on the human physiological structure to clarify the spatial vector of the offset amount in this coordinate system;

[0098] S53: Based on the standard human ergonomic body position database, set the body position compensation optimization objective function, and construct a mathematical optimization problem with the optimization objective of minimizing the deviation between the actual body position feature point and the standard theoretical position;

[0099] S54: Use the particle swarm optimization algorithm to obtain the solution to the optimization problem in S53 through iterative search, obtain the optimal body position offset compensation amount, and obtain a clear optimization value; Through the above steps, it is possible to accurately solve the difference between the current human body position and the standard ergonomic body position based on the multi-modal fusion parameters, realize the precise calculation of the body position offset amount, and provide a reliable reference and basis for the subsequent pose adjustment of the massage table surface.

[0100] S52 specifically includes:

[0101] S521: Based on human anatomical characteristics, with the standard supine position of the human body as the basis, construct a rectangular coordinate system with the center of gravity of the human body as the origin, the head-to-foot direction of the human body defined as the longitudinal axis Z, the left and right sides of the human body defined as the transverse axis X, and the front and back directions of the chest and abdomen of the human body defined as the vertical axis Y;

[0102] S522: Based on the position difference calculated in S51, project the actual feature point coordinate vector Q(x, y, z) into the human physiological structure coordinate system respectively to obtain the difference components corresponding to the X, Y, and Z axis directions, which are respectively denoted as Δx, Δy, and Δz; The specific calculation method is: Δx = x - x0; Δy = y - y0; Δz = z - z0, where Δx is the body position offset component in the transverse axis X direction; Δy is the body position offset amount in the vertical axis Y direction; Δz = z - z0 is the body position offset amount in the longitudinal axis direction; (x, y, z) are the spatial coordinate values of the actually measured human feature points; (x0, y0, z0) are the coordinate values of the corresponding human ergonomic standard feature points;

[0103] S53: According to the components Δx, Δy, and Δz of the position difference in the three axes of the coordinate system, determine the spatial direction vector V of the body position offset and the magnitude of the offset ΔL. The formulas are respectively: V = (Δx, Δy, Δz) and Where: V is the spatial body position offset vector, which clarifies the specific direction of the body position offset; |V| is the modulus of the spatial body position offset vector, representing the specific magnitude of the body position offset, with the unit of meter. Through the above method steps, a clear human body coordinate reference system is established, specifically giving the calculation process of the body position offset direction and magnitude, clarifying the difference between the actual body posture and the standard posture, and providing a clear direction and numerical basis for the subsequent precise adjustment of the massage bed pose.

[0104] S53 specifically includes:

[0105] S531: Obtain the theoretical position coordinates of multiple feature points from the standard human ergonomic body position database and define them as a set of theoretical coordinate point vectors in meters, where the subscript i is the feature point number, i = 1, 2, 3, …, k;

[0106] S532: Obtain the actual measured coordinate positions corresponding to the above key feature points according to the multi-modal fusion parameter set, define them as a set of actual coordinate vectors, denoted as Q i (x i , y i , z i ), in meters;

[0107] S532: Based on the difference between the actual coordinate points and the theoretical coordinate points, with the goal of minimizing the body position deviation, construct the objective function of the mathematical optimization problem, and the expression is:

[0108] Where, F is the optimization objective function, representing the overall position deviation between the actual feature point position and the theoretical position of the standard feature point, in meters; (x i , y i , z i ) is the spatial coordinate of the i-th actually measured feature point, in meters; (x 0i , y 0i , z 0i ) is the theoretical position coordinate of the i-th standard feature point, in meters; i is the serial number of the feature point. Through the above method steps, a specific body position optimization mathematical model is constructed, realizing the precise comparison and deviation calculation between the actual measurement data and the standard human ergonomic data, and providing a reliable and clear calculation basis for the subsequent precise determination of the massage bed pose adjustment.

[0109] S6 specifically includes:

[0110] S61: According to the body position offset compensation amount calculated by S5, obtain the pose adjustment target values of the massage bed surface in the X, Y, and Z directions, and clarify the distances and angles that the massage bed surface needs to move or tilt along each direction;

[0111] S62: Input the pose adjustment target value into the multi-axis servo control system, and analyze to obtain the operation instructions for each servo actuator, including the movement direction, displacement, and movement speed parameters;

[0112] S63: The multi-axis servo actuators act in the order of the operation instructions. Through linear movement and angular rotation, synchronous or step-by-step adjustment of the massage table surface in the three dimensions of X, Y, and Z is achieved to ensure that the pose of the table surface matches the desired compensation amount. Through the above steps, the body position offset compensation amount can be intuitively converted into the action instructions of the multi-axis servo actuators, and the pose adjustment of the massage table surface can be accurately completed, providing a comfortable and safe massage support environment that conforms to ergonomics for the human body.

[0113] The present invention covers any alternatives, modifications, equivalent methods, and solutions made within the essence and scope of the present invention. To enable the public to have a thorough understanding of the present invention, specific details are described in detail in the following preferred embodiments of the present invention. However, those skilled in the art can fully understand the present invention without the description of these details. In addition, well-known methods, processes, procedures, components, and circuits are not described in detail to avoid unnecessary confusion to the essence of the present invention.

[0114] The above description is only a preferred embodiment of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements can be made, and these improvements and refinements should also be regarded as the protection scope of the present invention.

Claims

1. A real-time monitoring method for the body position of a massage bed based on multimodal sensing, characterized in that, It includes the following steps: S1: Obtain the three-dimensional pressure distribution matrix of the contact surface through an embedded pressure sensor array, and synchronously collect the body surface temperature field data of the infrared thermal imaging module; S2: Perform temperature compensation processing on the pressure distribution matrix to generate a corrected pressure topology map; S3: Obtain the electrophysiological feature vectors of the target muscle group based on the electromyogram signal sensor group, and extract the muscle activation degree index through time-frequency analysis; S4: Align the corrected pressure topology map generated in S2 and the muscle activation degree index extracted in S3 in space and time to generate a multi-modal fusion parameter set; S5: Apply the ergonomic optimization algorithm to calculate the body position offset compensation amount according to the multi-modal fusion parameter set; S6: Based on the body position offset compensation amount, drive the servo actuator to adjust the pose of the massage bed surface to achieve body position optimization.

2. The real-time monitoring method for the body position of a massage bed based on multi-modal sensing according to claim 1, wherein The specific content of S1 includes: S11: Embed a flexible thin-film resistive pressure sensor array arranged in a matrix in the massage bed surface, and each sensor unit will output a corresponding resistance signal according to the different magnitudes of the applied force; S12: After the resistance signals generated by each resistive pressure sensor unit are converted into voltage signals through a bridge circuit, they are synchronously sampled and digitally processed through a multi-channel analog-to-digital conversion circuit, and sent to the processor unit through a data interface; S13: The processor unit constructs a three-dimensional space pressure distribution matrix of the contact surface between the massage bed and the human body based on the spatial position coordinates of each sensor unit and the corresponding digital pressure values; S14: Set an infrared thermal imaging module above the massage bed. The infrared thermal imaging module collects the infrared radiation energy of the human body surface in real time through a non-cooled focal plane array infrared detector; and converts the collected infrared radiation energy into body surface temperature data through a built-in calibration function to form a body surface temperature field data matrix.

3. The real-time monitoring method for the body position of a massage bed based on multimodal sensing according to claim 2, wherein, The expression of the calibration function is as follows: where T is the calculated body surface temperature; E is the infrared radiation energy measured by the infrared detector; ε is the surface emissivity of the target to be measured; σ is the Stefan-Boltzmann constant; and C is the correction coefficient obtained during the calibration of the device.

4. The real-time monitoring method for the body position of a massage bed based on multi-modal sensing according to claim 1, characterized in that, The specific content of S2 includes: S21: Synchronously register the three-dimensional space pressure distribution matrix and the body surface temperature field data matrix obtained in S1 at the same moment to obtain a data matrix in which the pressure value and the corresponding position temperature value correspond one by one; S22: According to the temperature drift characteristics of the flexible thin-film resistive pressure sensor, construct a fitting mathematical model of the output resistance value of the sensor changing with temperature, and calculate the temperature drift compensation coefficient according to this model; S23: Use the temperature drift compensation coefficient to correct the original measurement values of each pressure sensor unit in the three-dimensional space pressure distribution matrix point by point to obtain the pressure measurement values after eliminating temperature interference; S24: Based on the corrected pressure measurement values, reconstruct the pressure data of the massage bed contact surface through a spatial interpolation algorithm to generate a corrected pressure topology map for reflecting the true contact pressure distribution of the human body.

5. The real-time monitoring method for the body position of a massage bed based on multi-modal sensing according to claim 1, characterized in that, The specific content of S3 includes: S31: Place a surface electromyogram signal sensor group on the surface of the human target muscle group to collect the original electromyogram signal data generated during the contraction of the muscle group, and transmit it to the data acquisition device through a wire; S32: Perform band-pass filtering on the collected original electromyogram signal data, and use a finite impulse response filter to filter the electromyogram signals in the range of 20 Hz to 450 Hz to remove baseline drift and high-frequency interference to obtain the filtered pure electromyogram signals; S33: Perform frame windowing processing on the filtered EMG signal data, set the sliding time window length to 200 ms, and the window overlap rate to 50%, to obtain a number of EMG signal analysis frames; S34: Calculate the root mean square value in the time domain based on each analysis frame to obtain an electrophysiological feature vector; S35: Perform short-time Fourier transform time-frequency analysis on the filtered EMG signal, calculate the instantaneous frequency spectrum energy distribution of the EMG signal, and obtain a frequency domain energy distribution matrix; S36: Calculate the frequency-domain energy integral value of the target muscle group within a specific frequency band based on the frequency-domain energy distribution matrix to obtain the muscle activation index E m .

6. The real-time monitoring method for the body position of a massage bed based on multimodal sensing according to claim 1, characterized in that, The specific steps of S4 include: S41: Based on a unified synchronous clock source, determine the data acquisition timestamps of the calibrated pressure topology map generated by S2 and the muscle activation index extracted by S3, to ensure that the two types of data are completely corresponding in the time dimension; S42: Based on the human anatomical coordinate system, establish a spatial mapping relationship between the spatial coordinates of the calibrated pressure topology map on the massage table surface and the placement positions of the EMG signal sensors of the target muscle groups, to obtain a spatial matching relationship matrix between the pressure data and the EMG signal data; S43: Based on the timestamps determined by S41 and the spatial matching relationship matrix determined by S42, perform spatio-temporal data interpolation fusion on the calibrated pressure topology map data and the muscle activation index, to generate a spatio-temporally aligned multi-modal fusion data matrix, that is, a multi-modal fusion parameter set.

7. The real-time monitoring method for the body position of a massage bed based on multimodal sensing according to claim 1, characterized in that The specific steps of S5 include: S51: Input the multi-modal fusion parameter set generated in S4 into the ergonomic optimization algorithm to determine the position difference between the actual feature points of the current human body position and the theoretical feature points of the standard ergonomic body position; S52: Based on the position difference in S51, determine the direction and magnitude of the body position offset, and establish a coordinate reference system based on the human physiological structure to clarify the spatial vector of the offset in this coordinate system; S53: Based on the standard ergonomic body position database, set an objective function for body position compensation optimization, and construct a mathematical optimization problem with the optimization objective of minimizing the deviation between the actual body position feature points and the standard theoretical positions; S54: Use the particle swarm optimization algorithm to obtain the solution to the optimization problem in S53 through iterative search, to obtain the optimal body position offset compensation amount and obtain a clear optimization value.

8. The real-time monitoring method for the body position of a tuina bed based on multi-modal sensing according to claim 7, wherein, The specific steps of S52 include: S521: Based on human anatomical features, with the standard supine position of the human body as the basis, construct a rectangular coordinate system with the center of gravity of the human body as the origin, the head-to-foot direction of the human body defined as the longitudinal axis Z, the left and right sides of the human body defined as the horizontal axis X, and the front and back directions of the chest and abdomen of the human body defined as the vertical axis Y; S522: Based on the position difference calculated in S51, project the actual feature point coordinate vector Q(x, y, z) into the human physiological structure coordinate system respectively to obtain the difference components corresponding to the X, Y, and Z axis directions, denoted as Δx, Δy, and Δz respectively; S53: Determine the spatial direction vector V of the body position offset and the magnitude of the offset ΔL according to the components Δx, Δy, and Δz of the position difference in the three axes of the coordinate system. The formulas are: V = (Δx, Δy, Δz) and where: V is the spatial body position offset vector, which clarifies the specific direction of the body position offset; |V| is the modulus of the spatial body position offset vector, indicating the specific magnitude of the body position offset.

9. The real-time monitoring method for the body position of a massage bed based on multi-modal sensing according to claim 8, wherein, The specific steps of S53 include: S531: Obtain the theoretical position coordinates of multiple feature points from the standard human ergonomic position database and define them as a set of theoretical coordinate point vectors The unit is meter, where the subscript i is the feature point number, i = 1, 2, 3, …, k; S532: Obtain the actual measurement coordinate positions corresponding to the above key feature points according to the multi-modal fusion parameter set, which are defined as the set of actual coordinate vectors and denoted as Q i (x i , y i , z i ), with the unit of meter; S532: Based on the difference between the actual coordinate points and the theoretical coordinate points, with the goal of minimizing the body position deviation, construct the objective function of the mathematical optimization problem, and the expression is: Among them, F is the optimization objective function, representing the overall position deviation between the actual feature point position and the theoretical position of the standard feature point; (x i , y i , z i ) are the spatial coordinates of the i-th actually measured feature point; (x 0i , y 0i , z 0i ) are the theoretical position coordinates of the i-th standard feature point; i is the serial number of the feature point.

10. The real-time monitoring method for the body position of a massage bed based on multi-modal sensing according to claim 1, wherein, The specific steps of S6 include: S61: According to the body position offset compensation amount calculated in S5, obtain the pose adjustment target values of the massage table surface in the X, Y, and Z directions, and clarify the distances and angles that the massage table surface needs to move or tilt in each direction; S62: Input the pose adjustment target value into the multi-axis servo control system, and parse to obtain the operation instructions for each servo actuator, including the movement direction, displacement, and movement speed parameters; S63: The multi-axis servo actuators act in the order of the operation instructions, and through linear movement and angular rotation, achieve synchronous or step-by-step adjustment of the massage tabletop in the three dimensions of X, Y, and Z, ensuring that the tabletop pose matches the expected compensation amount.

Images