Intelligent seat regulation and control method and system based on sitting posture recognition

Abstract

The invention relates to an intelligent seat regulation and control method and system based on sitting posture recognition, and belongs to the technical field of seat regulation and control. Comprising the steps that according to the initial state of human body and seat regulation and control, position positioning of a human body is obtained, front-back adjustment is conducted according to obtained seat adjustment parameters, and backrest following is conducted according to backrest inclination angle data and backrest body pressure; acquiring human body pressure data and backrest inclination angle data through a data acquisition module; according to the human body pressure data and the backrest inclination angle data, a user posture is obtained through an established sitting posture recognition model, and seat adjusting parameters are generated according to the user sitting posture. Front and back adjustment of the seat is achieved by obtaining the sitting position information of the human body, adaptation of the height of the seat is completed according to the adjustment distance of the front and back positions, fitting following of the seat and the human body is completed through the set backrest inclination angle and the pressure threshold value, recognition data are collected for human body sitting posture recognition, and seat fine adjustment parameters are obtained; the validity of the collected data is guaranteed, and the accuracy of sitting posture recognition is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of seat control, and in particular relates to an intelligent seat control method and system based on sitting posture recognition. Background Art

[0002] When driving a car, the car seat is the part that the human body directly contacts with the largest area and for the longest time. As can be seen from the human body structure, the state of the tendons will change when the spine moves. When the spine deviates from the natural state, the tendons will increase the activity of the muscles under the action of external forces, making the waist susceptible to fatigue. Due to different physical conditions, different drivers and passengers have different perceptions and feelings about the same seat. Therefore, timely identification of abnormal sitting postures and adjustments are of great significance to relieve the fatigue of drivers and passengers.

[0003] Modern seats include many humanized designs, such as the front and rear position adjustment function of the seat, the angle adjustment function of the seat cushion and backrest, the up and down, front and back adjustment function of the headrest, and the adjustment function of the lumbar support, etc., to meet the needs of different drivers and passengers. However, in the actual adjustment process, due to the driving habits of different groups of people, the collected human posture data is not accurate, so the seat adjustment cannot play the best control strategy, and it cannot adapt to users with different human physiological parameters, and the effectiveness of the control is poor.

[0004] Therefore, there is an urgent need to propose an intelligent seat control method and system based on sitting posture recognition, which can achieve the effectiveness and accuracy of recognition data by adjusting the initial position of the human body. Summary of the invention

[0005] In order to solve the above-mentioned problems existing in the prior art, the present invention provides an intelligent seat control method and system based on sitting posture recognition, which adjusts the seat forward and backward by obtaining the position information of the human body, and adapts the seat height according to the adjustment distance of the front and rear position, and then completes the fitting of the seat and the human body by setting the backrest inclination angle and pressure threshold, thereby ensuring the validity and accuracy of the collected data; then, after the initial adjustment of the seat, the recognition data is collected to perform human sitting posture recognition, so as to obtain the seat fine-tuning parameters according to the user's sitting posture, and adjust the support degree accordingly according to the feedback body pressure distribution after adjustment.

[0006] The purpose of the present invention can be achieved through the following technical solutions: The present disclosure provides an intelligent seat control method based on sitting posture recognition, comprising the following steps: Human body positioning: The human body position is obtained according to the initial state of the human body and the seat adjustment, and then the seat is adjusted to fit the human body; Initial seat adjustment: The seat is adjusted forward and backward according to the acquired seat adjustment parameters, and the seat back follows the backrest according to the backrest tilt angle data and backrest body pressure; Collect and identify data: obtain human body pressure data and backrest inclination angle data through the data acquisition module, where the body pressure data includes buttocks body pressure data and back body pressure data; Obtaining adjustment parameters: According to the human body pressure data and backrest inclination angle data, the user posture is obtained through the established sitting posture recognition model, and the seat adjustment parameters are generated according to the user's sitting posture; The following strategies are included: Adjust the backrest tilt angle according to the backrest pressure distribution so that the backrest body pressure reaches the initial backrest pressure threshold, and set the backrest tilt angle threshold as a restriction condition for backrest following; When the backrest inclination angle reaches the set threshold, the backrest body pressure reaches the initial backrest pressure threshold by adjusting the front and rear position of the seat again; The fore-and-aft position of the seat after the initial adjustment is obtained, and according to the fore-and-aft movement ratio of the seat setting and the correlation between the seat height and the fore-and-aft position of the seat, the seat height is automatically adjusted to complete the adaptation with the fore-and-aft adjustment.

[0007] Furthermore, the human body position positioning comprises the following steps: Pressure data preparation: Establish a pressure distribution matrix through the collected human buttocks pressure distribution data, and record the pressure value of each point in the pressure distribution matrix P i,j ; Solve the center coordinates: According to the collected pressure distribution data, calculate the geometric center point of the hip of the human body ( x 0, y 0); Obtain seat adjustment parameters: Obtain seat adjustment parameters for forward and backward movement based on the calculated difference between the geometric center coordinates of the seated hip and the geometric center of the sensor array.

[0008] Furthermore, the solving of the center coordinates comprises the following steps: Select the origin: According to the sensor matrix set on the seat, use the row and column index of the sensor in the matrix as the acquisition point coordinates, and count the point at the upper left corner of the matrix (1,1); Solve the center point: The calculation formula is: ; ; In the formula, i and j are the row and column indices of the sensor in the matrix, and 1≤ i ≤ m , 1≤j ≤ n , P t is the total pressure, and the calculation formula is: .

[0009] Furthermore, obtaining the correlation between the seat height and the front and rear includes the following steps: Data collection: A group of representative users were selected for testing. In each test, the basic physiological parameters of the users were recorded, including height, leg length, and back length. During the driving process, data is collected: the front and rear position of the seat, the height of the seat, and the user's field of vision; Correlation analysis: Use statistical methods to determine the relationship between seat height and fore-aft position, including: calculating the correlation between seat height and fore-aft position, drawing a scatter plot to visualize data trends, and using a multivariate regression model to build a mathematical model to find the best fit line between height and fore-aft position; Fit verification: Verify the accuracy and fit of the model by testing with a new group of users, observing the users’ comfort and field of view under the new adjustments.

[0010] Furthermore, the obtaining of adjustment parameters comprises the following steps: Establish a sitting posture recognition model: collect the user's body pressure data and back inclination data through the pressure sensor installed on the seat and the monitoring vest worn by the user as the training data of the model, and extract features and establish labels for the data by selecting the mean, standard deviation, skewness and kurtosis statistics; Model construction: including a one-dimensional convolutional neural network layer, a dropout layer to prevent overfitting, a pooling layer for extracting features, a flattening layer for connecting the output layer, and an output layer for outputting the final classification results; The extracted features and established labels are input into the constructed convolutional neural network model for training to complete the sitting posture recognition.

[0011] Furthermore, the step of obtaining the adjustment parameters further includes the following steps: Identify the user's current sitting posture: The user's back pressure data and backrest inclination data are used to determine the state of the user's sitting posture and the seat. When the fit is maintained, the collected user body pressure data and backrest inclination data are input into the sitting posture recognition model to classify the user's sitting posture; Generate seat adjustment parameters: Generate seat adjustment parameters based on the different recognized sitting postures and make fine adjustments to the seat.

[0012] Furthermore, the method further comprises the steps of: Collecting the initial state: The data acquisition module collects the body pressure distribution data of the human buttocks and the seat back tilt angle data to obtain the initial state of the human body and seat control.

[0013] Furthermore, the method further comprises the steps of: Implement control strategy: automatically adjust the seat angle according to the seat adjustment parameters, re-collect body pressure data for feedback, and adjust the support degree of the corresponding parts through the seat control module; When the car is driving, the data of the user's manual adjustment of the seat is monitored, and the adjustment parameters are obtained based on the collected data to determine whether the user's posture is abnormal. When it is in a normal state, the adjustment data is recorded to update the sitting posture recognition model. When it is in an abnormal state, a voice reminder of the user's sitting posture warning is issued.

[0014] The present disclosure also provides an intelligent seat control system based on sitting posture recognition, which uses the intelligent seat control method based on sitting posture recognition as described above, including a data acquisition module, a human body positioning module, a parameter acquisition module and a seat control module, wherein the data acquisition module, the human body positioning module, the parameter acquisition module and the seat control module are sequentially connected in communication; The data acquisition module includes an inertial sensor and a piezoelectric film sensor, which are used to collect seat back inclination data and human body pressure data respectively; The human body positioning module collects the pressure distribution of the human body's buttocks, calculates the center coordinates of the human body's buttocks, and obtains the relative distance between the human body's sitting position and the seat.

[0015] As a preferred technical solution of the present invention, the parameter acquisition module is used to generate seat adjustment parameters and change the body pressure distribution by identifying the sitting posture of the human body; the seat control module includes an electric system and a pneumatic system, wherein the electric system controls the motor to adjust the front and back, height and backrest of the seat, and the pneumatic system controls the air pump to adjust the inflation degree of the airbag; Among them, the seat control module sets seat adjustment rules and controls the seat posture according to the different recognized sitting postures.

[0016] The beneficial effects of the present invention are: The present invention firstly obtains the position information of the human body sitting, adjusts the seat fore and aft, and adapts the seat height according to the adjustment distance of the fore and aft position, and then completes the fitting of the seat and the human body by setting the backrest inclination angle and pressure threshold, thereby ensuring the validity and accuracy of the collected data; then, after the initial adjustment of the seat, the recognition data is collected to identify the sitting posture of the human body, thereby obtaining the adjustment parameters of the seat fine-tuning according to the sitting posture, and adjusting the seat support of different parts of the human body again through the feedback of the body pressure data.

[0017] During the initial adjustment of the seat, the present invention establishes data fitting of the front and rear position and height of the seat in order to ensure the user's driving field of vision, thereby adapting to different human physiological parameters and realizing rapid adjustment; during the sitting posture recognition process, the body pressure data and the seat inclination angle data are simultaneously input into the model for recognition, thereby realizing a combination of the sitting posture recognition method based on a wearable device and the sitting posture recognition method based on a pressure sensor, thereby improving the accuracy of recognition, and at the same time correlating the seat back inclination angle with the human back pressure data, thereby obtaining a more refined basis for adjusting the support degree of human parts and improving human comfort. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0019] Figure 1 A schematic diagram of the steps of an intelligent seat control method based on sitting posture recognition provided by an embodiment of the present invention; Figure 2 A schematic diagram of steps for locating a human body provided by an embodiment of the present invention; Figure 3 A schematic diagram of the steps for obtaining adjustment parameters provided by an embodiment of the present invention; Figure 4 A schematic structural diagram of an intelligent seat control system based on sitting posture recognition provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0020] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects of the present invention are described in detail below in conjunction with the accompanying drawings and preferred embodiments.

[0021] From the perspective of data collection, the recognition of abnormal sitting posture of the human body is divided into an abnormal sitting posture recognition method based on a contact sensor, an abnormal sitting posture recognition method based on a radio frequency signal, and an abnormal sitting posture recognition method based on machine vision. In order to adapt to the application scenario of car driving, this embodiment adopts data collection based on a contact sensor. Since the adjustment of sitting comfort is achieved through seat adjustment, before obtaining the seat adjustment parameters, the present invention first locates the position of the human body and completes the fine-tuning of the seat, so that the human body and the seat are in a fit state, thereby further adjusting the seat.

[0022] Specifically, the seat control based on sitting posture recognition of the present invention is described in detail through the following contents: An intelligent seat control method based on sitting posture recognition, such as Figure 1 As shown, the following steps are included: S1. Collecting the initial state: Collecting the body pressure distribution data of the human buttocks and the seat back tilt angle data through the data acquisition module to obtain the initial state of the human body and the seat control.

[0023] S2. Human body position positioning: The human body position positioning is obtained according to the initial state of the human body and the seat adjustment, and then the seat is adjusted to fit the human body, such as Figure 2 As shown, the specific steps include: S21. Pressure data preparation: Establish a pressure distribution matrix through the collected human buttocks pressure distribution data, and record the pressure value of each point (sensor) in the pressure distribution matrix P i,j ; It is understandable that the collected pressure values ​​are filtered to obtain smoothed pressure values, and abnormal or noisy data are removed.

[0024] S22, solving the center coordinates: according to the collected pressure distribution data, calculate the geometric center point of the buttocks of the human body sitting ( x 0, y 0), including the steps of: Select the origin: According to the sensor matrix set on the seat, use the row and column index of the sensor in the matrix as the acquisition point coordinates, and count the point at the upper left corner of the matrix (1,1); Solve the center point: The calculation formula is: ; ; In the formula, i and j are the row and column indices of the sensor in the matrix (1≤ i ≤ m , 1≤ j ≤ n ), P t is the total pressure, and the calculation formula is: ; It should be noted that the above coordinate system is based on the pressure sensor array. The upper left corner of the sensor matrix is ​​set as the (1,1) point of the coordinate system, that is, the origin of the coordinate system is set as (1,1). If the matrix is m × n An array with the number of columns representing x Direction, row number indicates y direction.

[0025] S23, obtaining seat adjustment parameters: obtaining seat adjustment parameters for forward and backward movement according to the calculated difference between the geometric center coordinates of the seated buttocks and the geometric center coordinates of the sensor array.

[0026] It is understandable that by solving the hip position of the sitting human body, the preliminary adjustment parameters of the seat fit are obtained, but due to different human body parameters, the adjustment of the geometric center of the seat cannot completely fit and adapt to the adjustment needs of the seat. Therefore, the pressure data of the back is used to further fine-tune the front and rear to achieve a fit between the human body and the seat.

[0027] S3. Initial seat adjustment: The seat is adjusted forward and backward according to the acquired seat adjustment parameters, and the seat is followed by the backrest according to the backrest tilt angle data and the backrest body pressure. The backrest following includes the following strategies: Adjust the backrest tilt angle according to the backrest pressure distribution so that the backrest body pressure reaches the initial backrest pressure threshold, and set the backrest tilt angle threshold as a restriction condition for backrest following; When the backrest inclination angle reaches a set threshold, the backrest body pressure reaches the initial backrest pressure threshold by adjusting the front and rear position of the seat again.

[0028] It is understandable that by setting the backrest inclination angle threshold, the backrest can be avoided from being over-adjusted during the backrest following process. At the same time, the backrest pressure threshold is set to ensure that the backrest and the human back are in close contact to collect the back pressure data. Therefore, the initial adjustment of the seat is achieved by adjusting the sitting position of the human body and the backrest inclination angle. The control strategy is decided based on the inclination angle and pressure.

[0029] It should be noted that through the initial adjustment of the seat, the fore and aft position of the seat and the inclination of the backrest are adjusted, so that the seat as a whole is in close contact with the human body, creating conditions for the subsequent collection of body pressure distribution, thereby improving the practicality of seat adjustment. Accordingly, when the backrest pressure is high, the position of the seat and the angle of the backrest will also be adjusted adaptively.

[0030] The fore-and-aft position of the seat after the initial adjustment is obtained, and according to the fore-and-aft movement ratio of the seat setting and the correlation between the seat height and the fore-and-aft position of the seat, the seat height is automatically adjusted to complete the adaptation with the fore-and-aft adjustment.

[0031] It should be noted that, through the relationship between the height and the front and back positions, the seat height adjustment is directly completed according to the adjusted front and back positions, so as to achieve quick adaptation of different physiological parameters of the human body, complete the front and back and height adjustment of different users' seats, and ensure the appropriateness of the user's field of vision. Among them, the acquisition of the relationship between the seat height and the front and back includes the following steps: S31. Data collection: Select a group of representative users (different genders, heights, weights, etc.) for testing to ensure the wide applicability of the data. In each test, record the user's basic physiological parameters, including: height, leg length (length from hips to soles of feet), back length (length from hips to shoulders); During the driving process of the car, data is collected: The fore-aft position of the seat (e.g., forward, rearward displacement); The height of the seat (e.g., from the ground to the seat surface); The user's field of view (determine whether it meets driving safety standards through actual observation or the field of view captured by the camera).

[0032] It should be noted that the purpose of collecting data during driving is to ensure that the collected data is the final adjustment data. Here, the data is mainly collected for establishing associations. The driving process is the final parameters of the seat adjustment made by the user within the set time in the test. Subsequent driving monitoring is the automatic collection of data in the application scenario. The parameters obtained by the user's manual adjustment are used to update the sitting posture recognition model and improve the personalized adjustment of the seat.

[0033] S32. Association analysis: Use statistical methods (such as regression analysis) to determine the relationship between seat height and fore-aft position, including: Calculate the correlation between seat height and fore-aft position and plot a scatter plot to visualize data trends. Use a linear regression model or a multiple regression model to build a mathematical model to find the best fit line between height and fore-aft position.

[0034] S33, Fit verification: Verify the accuracy and applicability of the model by testing with a new group of users. Observe the users’ comfort and field of view under the new adjustment to ensure that the model can effectively adapt to different users.

[0035] By using the correlation fitted by the above method, the seat height that meets the user's optimal field of view is adapted according to the front and rear adjustment position of the seat, the initial adjustment of the seat is completed, and the coarse adjustment response speed is improved.

[0036] S4, collecting and identifying data: obtaining human body pressure data and backrest inclination angle data through a data collection module, wherein the body pressure data includes buttocks body pressure data and back body pressure data.

[0037] It is understandable that in the automatic adjustment of the car seat, when the human body fits the seat (the ratio of the buttocks to the seat reaches a predetermined value) and sits down, due to the different physiological parameters such as height and weight of different human bodies, in order to meet the vision requirements of driving, the human body will automatically adjust the sitting posture. At this time, the body pressure data of the back and buttocks will change. Therefore, this embodiment automatically generates the seat adjustment parameters according to the distribution of body pressure to complete the seat adjustment of the human body's sitting comfort and driving requirements. To ensure driving safety, the control process of this embodiment is only performed in static mode, and follow-up control is performed according to the adjustment of the human body posture.

[0038] S5, obtaining adjustment parameters: according to the human body pressure data and backrest inclination angle data, the user posture is obtained through the established sitting posture recognition model, and the seat adjustment parameters are generated according to the user's sitting posture, such as Figure 3 As shown, the following steps are included: S51, establishing a sitting posture recognition model: collecting the user's body pressure data and back inclination data through the pressure sensor installed on the seat and the monitoring vest worn by the user as training data for the model, and extracting features and establishing labels for the data by selecting statistics such as mean, standard deviation, skewness, and kurtosis; Model construction: includes a one-dimensional convolutional neural network layer, a dropout layer to prevent overfitting, a pooling layer for extracting features, a flattening layer for connecting the output layer, and an output layer for outputting the final classification results.

[0039] It should be noted that the Tensorflow framework is used in combination with a third-party library to build a neural network model. Tensorflow2.0 provides the definition and connection of the network layer. A one-dimensional convolutional layer (Conv1D) is used for feature learning. In order to reduce the extremely fast learning speed of CNN, a dropout layer (Dropout) is designed to randomly remove neurons in the network to slow down the learning process to control the risk of overfitting; the commonly used pooling layers Pooling have a maximum pooling layer and an average pooling layer, and the maximum pooling layer MaxPooling1D selected in this embodiment can reduce the deviation of the estimated mean caused by the convolution layer parameter error; after the transition of the flattening layer Flatten, the output layer is finally connected to output the classification result. The convolution layer uses the ReLU function to speed up the training speed. This function converges quickly and can overcome the "gradient disappearance" problem.

[0040] The extracted features and established labels are input into the constructed convolutional neural network model for training to complete the sitting posture recognition.

[0041] It is understandable that in order to improve the accuracy and training efficiency of the model, the hyperparameters can be optimized through the validation set to ensure that the model performs best on both training data and test data, avoid overfitting or underfitting, and find a model configuration suitable for a specific task with reasonable time and resource consumption.

[0042] S52, identifying the user's current sitting posture: judging the state of the user's sitting posture and the seat through the user's back pressure data and backrest inclination data. When the fit state is maintained, inputting the collected user body pressure data and backrest inclination data into the sitting posture recognition model to classify the user's sitting posture.

[0043] It should be noted that, in this embodiment, the inclination angle of the backrest is used as the inclination data of the human back to input the model for classification. Based on the initial adjustment step of the seat of this method, the following of the seat and the human body is completed. Therefore, the inclination angle of the seat backrest replaces the inclination data of the human back. Firstly, in order to improve the accuracy of the model in recognizing the sitting posture, and secondly, in order to associate the inclination angle of the backrest with the pressure distribution of the human back, the initial adjustment of the seat ensures the validity of the collected data.

[0044] The judgment of the fitting state is consistent with the initial adjustment of the seat, and is achieved through the set threshold. The reason for fitting again here is to ensure the interchangeability of the backrest inclination data during the sitting posture recognition process and to improve the practicality of the recognition results.

[0045] S53, generating seat adjustment parameters: generating seat adjustment parameters according to the identified different sitting postures (such as leaning forward, leaning back, bending over, etc.), and performing fine adjustments on the seat.

[0046] S6. Implement control strategy: automatically adjust the seat angle according to the seat adjustment parameters, re-collect body pressure data for feedback, and adjust the support degree of the corresponding part through the seat control module.

[0047] When the car is driving, the data of the user's manual adjustment of the seat is monitored, and the adjustment parameters are obtained based on the collected data to determine whether the user's posture is abnormal. When it is in a normal state, the adjustment data is recorded to update the sitting posture recognition model. When it is in an abnormal state, a voice reminder of the user's sitting posture warning is issued.

[0048] It is understandable that in this embodiment, the posture adjustment of the seat is completed through coarse adjustment and fine adjustment, and finally the support degree of the seat to different parts of the human body is adjusted through the adjusted data feedback, so as to optimize the sitting comfort of the human body.

[0049] This embodiment also provides an intelligent seat control system based on sitting posture recognition, such as Figure 4 As shown, it includes a data acquisition module, a human body positioning module, a parameter acquisition module and a seat control module, and the data acquisition module, the human body positioning module, the parameter acquisition module and the seat control module are communicatively connected in sequence.

[0050] The data acquisition module includes an inertial sensor and a piezoelectric film sensor, which are used to collect human body pressure data and seat back inclination angle data.

[0051] It is understandable that piezoelectric film sensors have the advantages of high sensitivity, good real-time signal, less environmental impact on performance, strong durability, flexibility and the ability to be cut into various shapes and sizes to fit different types of seats.

[0052] The human body positioning module collects the pressure distribution of the human body's buttocks, calculates the center coordinates of the human body's buttocks, and obtains the relative distance between the human body's sitting position and the seat.

[0053] It is understandable that the human body positioning module is mainly used to obtain the distance difference between the human body and the seat after sitting down, that is, the distance between the human back and the bottom of the seat back, so as to ensure the angle adjustment of the seat and the support of the airbag to the human waist and legs, so as to maximize the seat adjustment capacity.

[0054] The parameter acquisition module is used to generate seat adjustment parameters and change the body pressure distribution by identifying the sitting posture of the human body.

[0055] It should be noted that the parameter acquisition module collects the user's body pressure data and backrest inclination data based on the initial adjustment of the seat, and identifies the user's sitting posture through the sitting posture recognition model. The user's sitting posture here is not just the conventional postures such as leaning forward and leaning back, but a collection of different posture parameter groups defined by serial number codes. Through research and evaluation of ergonomic comfort, refined set members are obtained.

[0056] The seat control module includes an electric system and a pneumatic system, wherein the electric system adjusts the front and rear position, height and backrest of the seat by controlling the motor, and the pneumatic system adjusts the inflation degree of the airbag by controlling the air pump.

[0057] The seat control module sets seat adjustment rules and controls the seat posture according to the different recognized sitting postures. The actual sitting posture can be finely set and the fine-tuning scheme can be set for its encoding, such as defining the seat adjustment rules for different sitting postures, including: Forward tilt: the backrest angle is forward and the seat height is lowered; Recline: the backrest angle moves backward and the seat height is adjusted upward; Normal sitting position: Keep the seat position unchanged or make slight adjustments.

[0058] It is understandable that in order to ensure the user's seat adjustment flexibility, during driving, the user can adjust the seat comfort through manual adjustment. In this process, the user's adjustment data is collected to optimize the sitting posture recognition model and improve the personalization of automatic seat control.

[0059] The present invention firstly obtains the position information of the human body sitting, adjusts the seat fore and aft, and adapts the seat height according to the adjustment distance of the fore and aft position, and then completes the fitting of the seat and the human body by setting the backrest inclination angle and pressure threshold, thereby ensuring the validity and accuracy of the collected data; then, after the initial adjustment of the seat, the recognition data is collected to identify the sitting posture of the human body, thereby obtaining the adjustment parameters of the seat fine-tuning according to the sitting posture, and adjusting the seat support of different parts of the human body again through the feedback of the body pressure data.

[0060] During the initial adjustment of the seat, the present invention establishes data fitting of the front and rear position and height of the seat in order to ensure the user's driving field of vision, thereby adapting to different human physiological parameters and realizing rapid adjustment; during the sitting posture recognition process, the body pressure data and the seat inclination angle data are simultaneously input into the model for recognition, thereby realizing a combination of the sitting posture recognition method based on a wearable device and the sitting posture recognition method based on a pressure sensor, thereby improving the accuracy of recognition, and at the same time correlating the seat back inclination angle with the human back pressure data, thereby obtaining a more refined basis for adjusting the support degree of human parts and improving human comfort.

[0061] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An intelligent seat control method based on sitting posture recognition, characterized in that: The following steps are involved: Human body positioning: The human body position is obtained according to the initial state of the human body and the seat adjustment, and then the seat is adjusted to fit the human body; Initial seat adjustment: The seat is adjusted forward and backward according to the acquired seat adjustment parameters, and the seat back follows the backrest according to the backrest tilt angle data and backrest body pressure; Collect and identify data: obtain human body pressure data and backrest inclination angle data through the data acquisition module, where the body pressure data includes buttocks body pressure data and back body pressure data; Obtaining adjustment parameters: According to the human body pressure data and backrest inclination angle data, the user posture is obtained through the established sitting posture recognition model, and the seat adjustment parameters are generated according to the user's sitting posture; The following strategies are included: Adjust the backrest tilt angle according to the backrest pressure distribution so that the backrest body pressure reaches the initial backrest pressure threshold, and set the backrest tilt angle threshold as a restriction condition for backrest following; When the backrest inclination angle reaches the set threshold, the backrest body pressure reaches the initial backrest pressure threshold by adjusting the front and rear position of the seat again; The fore-and-aft position of the seat after the initial adjustment is obtained, and according to the fore-and-aft movement ratio of the seat setting and the correlation between the seat height and the fore-and-aft position of the seat, the seat height is automatically adjusted to complete the adaptation with the fore-and-aft adjustment.

2. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: The human body position positioning comprises the following steps: Pressure data preparation: Establish a pressure distribution matrix through the collected human buttocks pressure distribution data, and record the pressure value of each point in the pressure distribution matrix P i,j ; Solve the center coordinates: According to the collected pressure distribution data, calculate the geometric center point of the hip of the human body ( x 0, y 0); Obtain seat adjustment parameters: Obtain seat adjustment parameters for forward and backward movement based on the calculated difference between the geometric center coordinates of the seated hip and the geometric center of the sensor array.

3. The intelligent seat control method based on sitting posture recognition according to claim 2 is characterized in that: The method for solving the center coordinates comprises the following steps: Select the origin: According to the sensor matrix set on the seat, use the row and column index of the sensor in the matrix as the acquisition point coordinates, and count the point at the upper left corner of the matrix (1,1); Solve the center point: The calculation formula is: ； ； In the formula, i and j are the row and column indices of the sensor in the matrix, and 1≤ i ≤ m , 1≤ j ≤ n , P t is the total pressure, and the calculation formula is: .

4. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: The acquisition of the correlation between the seat height and the front and rear comprises the following steps: Data collection: A group of representative users were selected for testing. In each test, the basic physiological parameters of the users were recorded, including height, leg length, and back length. During the driving process, data is collected: the front and rear position of the seat, the height of the seat, and the user's field of vision; Correlation analysis: Use statistical methods to determine the relationship between seat height and fore-aft position, including: calculating the correlation between seat height and fore-aft position, drawing a scatter plot to visualize data trends, and using a multivariate regression model to build a mathematical model to find the best fit line between height and fore-aft position; Fit verification: Verify the accuracy and fit of the model by testing with a new group of users, observing the users’ comfort and field of view under the new adjustments.

5. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: The step of obtaining the adjustment parameters comprises the following steps: Establish a sitting posture recognition model: collect the user's body pressure data and back inclination data through the pressure sensor installed on the seat and the monitoring vest worn by the user as the training data of the model, and extract features and establish labels for the data by selecting the mean, standard deviation, skewness and kurtosis statistics; Model construction: including a one-dimensional convolutional neural network layer, a dropout layer to prevent overfitting, a pooling layer for extracting features, a flattening layer for connecting the output layer, and an output layer for outputting the final classification results; The extracted features and established labels are input into the constructed convolutional neural network model for training to complete the sitting posture recognition.

6. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: The step of obtaining the adjustment parameters further includes: Identify the user's current sitting posture: The user's back pressure data and backrest inclination data are used to determine the state of the user's sitting posture and the seat. When the fit is maintained, the collected user body pressure data and backrest inclination data are input into the sitting posture recognition model to classify the user's sitting posture; Generate seat adjustment parameters: Generate seat adjustment parameters based on the different recognized sitting postures and make fine adjustments to the seat.

7. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: Also includes the steps: Collecting the initial state: The data acquisition module collects the body pressure distribution data of the human buttocks and the seat back tilt angle data to obtain the initial state of the human body and seat control.

8. The intelligent seat control method based on sitting posture recognition according to claim 1, characterized in that: Also includes the steps: Implement control strategy: automatically adjust the seat angle according to the seat adjustment parameters, re-collect body pressure data for feedback, and adjust the support degree of the corresponding parts through the seat control module; When the car is driving, the data of the user's manual adjustment of the seat is monitored, and the adjustment parameters are obtained based on the collected data to determine whether the user's posture is abnormal. When it is in a normal state, the adjustment data is recorded to update the sitting posture recognition model. When it is in an abnormal state, a voice reminder of the user's sitting posture warning is issued.

9. An intelligent seat control system based on sitting posture recognition, using an intelligent seat control method based on sitting posture recognition as claimed in any one of claims 1 to 8, characterized in that: It includes a data acquisition module, a human body positioning module, a parameter acquisition module and a seat control module, wherein the data acquisition module, the human body positioning module, the parameter acquisition module and the seat control module are sequentially connected in communication; The data acquisition module includes an inertial sensor and a piezoelectric film sensor, which are used to collect seat back inclination data and human body pressure data respectively; The human body positioning module collects the pressure distribution of the human body's buttocks, calculates the center coordinates of the human body's buttocks, and obtains the relative distance between the human body's sitting position and the seat.

10. The intelligent seat control system based on sitting posture recognition according to claim 9, characterized in that: The parameter acquisition module is used to generate seat adjustment parameters and change the body pressure distribution by identifying the sitting posture of the human body; the seat control module includes an electric system and a pneumatic system, wherein the electric system controls the motor to adjust the front and back, height and backrest of the seat, and the pneumatic system controls the air pump to adjust the inflation degree of the airbag; Among them, the seat control module sets seat adjustment rules and controls the seat posture according to the different recognized sitting postures.

Images