Mattress control method and device and intelligent mattress

Abstract

The invention provides a mattress control method and device and an intelligent mattress, and belongs to the technical field of intelligent homes. The method comprises the steps that sleeping posture information of a user is obtained, the sleeping posture information comprises position information and temperature information when the user is located on the intelligent mattress, and the temperature information is used for indicating the temperature in an area, corresponding to the position information, of the intelligent mattress; determining a real-time sleeping posture according to the sleeping posture information; according to the real-time sleeping posture, adjusting the air pressure value of each air bag in the intelligent mattress so as to adjust the hardness of the intelligent mattress. The intelligent mattress can be automatically adjusted according to the sleeping posture of the user, so that effective sleep assistance can be provided for the user.

Description

Technical Field

[0001] The present application relates to the field of smart home technology, and in particular to a mattress control method, device and smart mattress. Background Art

[0002] With the rapid development of science and technology, various intelligent household appliances have been popularized in thousands of households, such as smart mattresses, etc. These smart mattresses are equipped with multiple adjustable air bags, which can adjust the air pressure value of the air bags in the mattresses, thereby changing the hardness of the mattresses.

[0003] In the related art, generally, corresponding communication devices or input devices are also provided in these smart mattresses, and users can trigger corresponding control instructions by operating the input devices, or send corresponding control instructions to the control unit in the smart mattress through the communication device. When the control unit in the smart mattress receives these control instructions, it can control the airbags in the smart mattress, and then adjust the smart mattress to the softness or hardness desired by the user.

[0004] However, the solutions in the related art require the user to manually adjust the control. Once the user falls asleep, the control cannot be adjusted accordingly according to the user's sleeping posture, making it difficult to provide the user with effective sleep assistance. Summary of the invention

[0005] In view of this, the mattress control method, device and smart mattress provided in the embodiments of the present application can automatically adjust the smart mattress according to the user's sleeping posture, thereby providing the user with effective sleep assistance. The mattress control method, device and smart mattress provided in the embodiments of the present application are implemented as follows:

[0006] In one aspect of an embodiment of the present application, a mattress control method is provided, which is applied to a smart mattress, wherein the smart mattress includes at least one airbag; the method includes:

[0007] Acquire sleeping posture information of the user, the sleeping posture information including position information and temperature information of the user when the user is located on the smart mattress, the temperature information being used to indicate the temperature in an area on the smart mattress corresponding to the position information;

[0008] determining a real-time sleeping posture according to the sleeping posture information;

[0009] According to the real-time sleeping posture, the air pressure value of each air bag in the smart mattress is adjusted to adjust the softness and hardness of the smart mattress.

[0010] Optionally, obtaining the sleeping posture information of the user includes:

[0011] Acquire a plurality of coordinate data corresponding to the user to obtain the position information;

[0012] The target temperature value at the position corresponding to each coordinate data is obtained to obtain temperature information.

[0013] Optionally, a plurality of acceleration sensors are installed in the smart mattress; and the acquiring of a plurality of coordinate data corresponding to the user includes:

[0014] Acquiring the three-dimensional acceleration detected by each of the acceleration sensors;

[0015] Based on the three-dimensional accelerations, determining the angle between any two adjacent acceleration sensors in the acceleration sensors and the chord length of the curve;

[0016] The position information is determined based on the angle and the chord length of the curve, and the position information is used to indicate the three-dimensional coordinates of each of the acceleration sensors in the area corresponding to the smart mattress and the user.

[0017] Optionally, a plurality of temperature sensors are installed in the smart mattress; the step of obtaining the target temperature value at the position corresponding to each coordinate data includes:

[0018] Acquire the temperature value detected by each of the temperature sensors;

[0019] The temperature value detected by the temperature sensor corresponding to the three-dimensional coordinates of each of the acceleration sensors is used as the target temperature value.

[0020] Optionally, determining the real-time sleeping posture according to the sleeping posture information includes:

[0021] Obtaining the real-time sleeping posture according to the sleeping posture information and the physiological information of the user;

[0022] The physiological information includes at least one of the following: height, weight, and waist-to-hip ratio.

[0023] Optionally, determining the real-time sleeping posture according to the sleeping posture information includes:

[0024] The sleeping posture information is input into a pre-trained target sleeping posture recognition model to obtain a real-time sleeping posture.

[0025] Optionally, inputting the sleeping posture information into a pre-trained target sleeping posture recognition model to obtain the real-time sleeping posture includes:

[0026] The sleeping posture information and the physiological information of the user are input into the target sleeping posture recognition model to obtain the real-time sleeping posture.

[0027] Optionally, the method further comprises:

[0028] Acquire a sample data set; the sample data set includes multiple sets of sample sleeping posture information and sample labels corresponding to each set of the sample sleeping posture information; the sample sleeping posture information includes sample position information of the user on the smart mattress and sample temperature information of the smart mattress, and the sample label is used to indicate the sleeping posture of the user;

[0029] Inputting the sample data set into an initial sleeping posture recognition model, and the initial sleeping posture recognition model outputs first recognition information and second recognition information; the first recognition information and the second recognition information both include a predicted label and a probability value corresponding to the predicted label;

[0030] determining loss information of the initial sleeping posture recognition model based on the first recognition information and the second recognition information;

[0031] According to the loss information of the initial sleeping position recognition model, the initial sleeping position recognition model is iteratively corrected until the initial sleeping position recognition model reaches a preset condition and the training is terminated, and the initial sleeping position recognition model that reaches the preset condition is used as the target sleeping position recognition model.

[0032] Optionally, determining the loss information of the initial sleeping posture recognition model based on the first recognition information and the second recognition information includes:

[0033] Determine a first loss value between the first identification information and the sample label, and determine a second loss value between the second identification information and the sample label;

[0034] Obtain a first loss result based on the first loss value and the second loss value, where the first loss result is a cross entropy loss;

[0035] Determine a second loss result between the first identification information and the second identification information, the second loss result being a predicted probability divergence loss;

[0036] Determine a third loss result between the first identification information and the second identification information, the third loss result being the deepest feature vector loss;

[0037] The loss information of the initial sleeping posture recognition model is obtained according to the first loss result, the second loss result, and the third loss result.

[0038] In another aspect of the embodiment of the present application, a mattress control device is provided, which is applied to a smart mattress, wherein the smart mattress includes at least one air bag; the device includes:

[0039] an acquisition module, used to acquire sleeping posture information of the user, the sleeping posture information including position information and temperature information of the user when the user is located on the smart mattress, the temperature information being used to indicate the temperature in an area on the smart mattress corresponding to the position information;

[0040] a determination module, configured to determine a real-time sleeping posture according to the sleeping posture information;

[0041] The regulating module is used to adjust the air pressure value of each air bag in the smart mattress according to the real-time sleeping posture, so as to adjust the softness and hardness of the smart mattress.

[0042] The smart mattress provided in the embodiment of the present application includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the method described in the embodiment of the present application is implemented when the computer program is executed.

[0043] The computer-readable storage medium provided in the embodiment of the present application has a computer program stored thereon, and when the computer program is executed by a processor, the method provided in the embodiment of the present application is implemented.

[0044] The mattress control method, device and smart mattress provided in the embodiments of the present application obtain the sleeping posture information of the user, determine the real-time sleeping posture according to the sleeping posture information, and adjust the air pressure value of each air bag in the smart mattress according to the real-time sleeping posture to adjust the softness and hardness of the smart mattress.

[0045] The sleeping posture information includes location information and temperature information of the user when the user is located on the smart mattress, and the temperature information indicates the temperature in an area on the smart mattress corresponding to the location information.

[0046] As can be seen from the above, in the process of adjusting the air pressure value of each airbag in the smart mattress, the user does not need to operate other input devices or other communication devices to trigger the corresponding control instructions, that is, the method provided by the present application can realize automatic adjustment of the air pressure value of the airbag in the smart mattress. In this way, even if the user enters the sleep state, the air pressure value of each airbag in the smart mattress can be adjusted in a timely and accurate manner according to the real-time sleeping posture of the user in the sleep state.

[0047] Moreover, when adjusting the smart mattress, the user's real-time sleeping posture is determined based on the position information and temperature information of the user when the user is on the smart mattress, and the air bags in the smart mattress are adjusted. That is to say, the mattress control method provided by the present application can adjust the air bags in the smart mattress to the air pressure value that best matches or is most suitable for the user's current sleeping posture, thereby improving the comfort of the user lying on the smart mattress.

[0048] In this way, effective sleep assistance can be provided to users, solving the technical problems raised in the background technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 A flowchart of a first mattress control method provided in an embodiment of the present application;

[0051] Figure 2 A flowchart of a second mattress control method provided in an embodiment of the present application;

[0052] Figure 3 A schematic diagram of the arrangement of the acceleration sensor provided in the embodiment of the present application;

[0053] Figure 4 A flowchart of a third mattress control method provided in an embodiment of the present application;

[0054] Figure 5 A schematic diagram of the arrangement of the temperature sensors provided in the embodiment of the present application;

[0055] Figure 6 A flowchart of a fourth mattress control method provided in an embodiment of the present application;

[0056] Figure 7 A flowchart of a fifth mattress control method provided in an embodiment of the present application;

[0057] Figure 8 A schematic diagram of a sleeping position provided in an embodiment of the present application;

[0058] Fig. 9 A schematic diagram of the structure of a sleeping posture recognition model provided in an embodiment of the present application;

[0059] Fig.10 A schematic diagram of the structure of a mattress control device provided in an embodiment of the present application;

[0060] Fig.11 A schematic diagram of the structure of a smart mattress provided in an embodiment of the present application. DETAILED DESCRIPTION

[0061] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the specific technical solution of the present application will be further described in detail below in conjunction with the drawings in the embodiments of the present application. The following embodiments are used to illustrate the present application, but are not used to limit the scope of the present application.

[0062] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0063] In the following description, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it will be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0064] It should be pointed out that the terms "first\second\third" involved in the embodiments of the present application are used to distinguish similar or different objects, and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

[0065] In the related art, generally, corresponding communication devices or input devices are also provided in these smart mattresses, and users can trigger corresponding control instructions by operating the input devices, or send corresponding control instructions to the control unit in the smart mattress through the communication device. When the control unit in the smart mattress receives these control instructions, it can control each airbag in the smart mattress, and then adjust the smart mattress to the softness or hardness desired by the user.

[0066] However, the solutions in the related art require the user to manually adjust the control. Once the user falls asleep, the control cannot be adjusted accordingly according to the user's sleeping posture, making it difficult to provide the user with effective sleep assistance.

[0067] To this end, the embodiment of the present application provides a mattress control method, which obtains the user's sleeping position information, determines the real-time sleeping position according to the sleeping position information, and adjusts the air pressure value of each airbag in the smart mattress according to the real-time sleeping position to adjust the softness and hardness of the smart mattress. The sleeping position information includes the position information and temperature information of the user when the user is located on the smart mattress, and the temperature information is used to indicate the temperature in the area corresponding to the position information on the smart mattress. In this way, the smart mattress can be automatically adjusted according to the user's sleeping position, thereby providing the user with effective sleep assistance.

[0068] The embodiment of the present application is described by taking the mattress control method applied in a smart mattress as an example, but it does not mean that the embodiment of the present application can only be applied to a smart mattress for mattress control.

[0069] The mattress control method provided in the embodiment of the present application is explained in detail below.

[0070] Figure 1 This is a flowchart of a mattress control method provided by the present application. The method can be applied to an electronic mattress, which may include any control unit and / or processor with processing, control, identification, calculation and other functions. Figure 1 , an embodiment of the present application provides a mattress control method, the method comprising:

[0071] Step 1001: Obtain the user's sleeping posture information.

[0072] In this embodiment, the sleeping posture information includes the position information and temperature information of the user when the user is located on the smart mattress.

[0073] Generally, the user may be lying on the smart mattress in any possible posture, such as lying flat, lying on the side, lying prone, etc. When the user is lying on the smart mattress in different postures, different pressures will be generated on the smart mattress, thereby causing the smart mattress to have different deformations.

[0074] That is to say, the position information is associated with the sleeping posture of the user, and the position information can be used to indicate the relative position relationship between the user and the smart mattress, and can also be used to indicate the position where the user exerts pressure and / or deformation on the smart mattress.

[0075] Optionally, the temperature information is used to indicate the temperature in an area on the smart mattress corresponding to the position information.

[0076] The area corresponding to the location information may refer to an area where the user contacts the smart mattress when the user is on the smart mattress, or may refer to an area where the smart mattress is deformed when the user is on the smart mattress.

[0077] The temperature information can represent the heat exchange between the user and the smart mattress, and can also be used to indicate the approximate contact range between the user and the smart mattress.

[0078] In this embodiment, the electronic mattress may include a plurality of airbags, and each airbag may be controlled by a corresponding control unit and / or processor to adjust the air pressure value of each airbag.

[0079] In one possible manner, the smart mattress may also be provided with any possible detection device to obtain the sleeping posture information, for example, a corresponding position sensor, pressure sensor or camera device may be provided to detect the position information, and a corresponding temperature sensor or infrared temperature measuring device may be provided to detect the temperature information, but the embodiment of the present application does not limit this.

[0080] It is understandable that if the user is not on the smart mattress, the temperature of the smart mattress is generally the same (for example, it may be room temperature); when the user is on the smart mattress, since the user's body temperature is generally different from the temperature of the smart mattress, heat exchange occurs between the smart mattress and the user's body, thereby causing the temperature of the area of ​​the smart mattress corresponding to the location information to change.

[0081] It is worth noting that since the position information in the sleeping posture information can indicate the relative position relationship between the user and the smart mattress, and can also indicate the pressure and / or deformation of the smart mattress; and the temperature information in the sleeping posture information can indicate the heat exchange between the user and the smart mattress, and / or the approximate contact range. Therefore, obtaining the sleeping posture information of the user facilitates the subsequent determination of the user's posture and other information.

[0082] Step 1002: Determine the real-time sleeping position according to the sleeping position information.

[0083] Specifically, the real-time sleeping posture may be determined according to the position information and the temperature information.

[0084] Optionally, the real-time sleeping posture may refer to the posture of the user on the smart mattress at the current moment, and specifically may be the posture of the user on the smart mattress when the sleeping posture information is acquired.

[0085] Exemplarily, the real-time sleeping posture may include any possible posture, for example, the real-time sleeping posture may include lying on the left side, lying on the right side, lying prone, lying supine, and any other possible postures, which are not limited in the embodiments of the present application.

[0086] In one possible manner, relevant technical personnel may collect in advance the pressure data and / or deformation data generated by the smart mattress when the user lies on the smart mattress in various sample sleeping positions, and collect the temperature data of the user lying on the smart mattress in various sleeping positions, and then organize the pressure data and / or the deformation data into corresponding sample position information, and organize the temperature data into corresponding sample temperature information, and then establish a corresponding retrieval table based on each sample sleeping position, the sample position information and the sample temperature information.

[0087] Then, the sample position information and the sample temperature information corresponding to the position information and the temperature information indicated by the sleeping position information can be searched in the search table based on the table lookup method, and then the sample sleeping position corresponding to the sleeping position information can be determined. In this case, the sample sleeping position corresponding to the sleeping position information is the real-time sleeping position.

[0088] In one possible manner, after relevant technical personnel have collected the above-mentioned sample sleeping positions, the sample position information and the sample temperature information, they can also input each sample sleeping position, the sample position information and the sample temperature information into a corresponding neural network model, and iteratively train the neural network model in any possible manner to obtain a model that can accurately identify or calculate the corresponding sleeping position based on the position information and temperature information.

[0089] Then, the position information and temperature information indicated by the sleeping posture information can be input into the trained neural network model, and then a sample sleeping posture corresponding to the sleeping posture information can be obtained. In this case, the sample sleeping posture corresponding to the sleeping posture information is the real-time sleeping posture.

[0090] It should be understood that the above is only to illustrate two possible examples of determining the real-time sleeping position based on the sleeping position information in the embodiment of the present application, and does not mean that the embodiment of the present application can only determine the real-time sleeping position by table lookup method or neural network model. In actual application, any possible method can be used to determine the real-time sleeping position, and the embodiment of the present application is not limited to this.

[0091] It is worth noting that when the user lies on the electric air mattress in different sleeping positions, the smart mattress will deform differently, and the smart mattress will also provide different support to the user. However, since the sleeping position information includes the position information and temperature information of the user lying on the smart mattress, the real-time sleeping position can be accurately determined based on the sleeping position information. In this way, it is convenient to adjust the smart mattress according to the real-time sleeping position.

[0092] Step 1003: According to the real-time sleeping posture, the air pressure value of each air bag in the smart mattress is adjusted to adjust the softness and hardness of the smart mattress.

[0093] Optionally, the greater the air pressure value of each air bag, the harder the smart mattress is; conversely, the smaller the air pressure value of each air bag, the softer the smart mattress is; this embodiment of the present application is not limited to this.

[0094] For example, the smart mattress may include a control unit, multiple airbags, and multiple electric air pumps, and each airbag corresponds to at least one electric air pump. In this way, the control unit can control the corresponding electric air pump to inflate or deflate each airbag to increase or decrease the air pressure value of each airbag.

[0095] In one possible manner, air bags corresponding to different body parts of a human being can be respectively arranged at corresponding positions in the smart mattress, so that the smart mattress can provide adaptive support for different body parts of the user.

[0096] Furthermore, airbags can be provided on the smart mattress for the upper back, lower back, buttocks, and legs, respectively, and each part corresponds to at least one airbag and at least one electric air pump. When the air pressure value of each airbag in the smart mattress is adjusted according to the real-time sleeping posture, the airbags corresponding to different parts can be adjusted according to the real-time sleeping posture, so that the airbags corresponding to different parts can have different air pressure values, thereby making the softness and hardness of different positions of the smart mattress different. In this way, the practicality of the smart mattress and the mattress control method can be improved.

[0097] In another possible manner, the smart mattress may also be provided with an air pressure sensor for detecting the air pressure value of each air bag, and each air pressure sensor may be used to send an electrical signal indicating the air pressure value of each air bag to the processor or the control unit, so that the control unit can accurately adjust the air pressure value of each air bag. The embodiments of the present application are not limited to this.

[0098] It is worth noting that when the user lies on the electric air cushion in different sleeping positions, the smart mattress will produce different deformations. Therefore, the smart mattress needs to be adjusted accordingly according to the real-time sleeping position so that the smart mattress can be adjusted to the softness and hardness corresponding to the real-time sleeping position as much as possible, thereby improving the comfort of the user lying on the smart mattress and the user experience.

[0099] It is worth noting that in this embodiment, the user's sleeping position information can be obtained and the user's real-time sleeping position can be determined based on the sleeping position information, and then the air pressure value of each airbag in the smart mattress can be adjusted according to the real-time sleeping position to adjust the softness and hardness of the smart mattress. In this process, the user does not need to operate other input devices or other communication devices to trigger the corresponding control instructions, that is, the method provided by this embodiment can realize the automatic adjustment of the air pressure value of the airbag in the smart mattress. In this way, even if the user enters the sleep state, the air pressure value of each airbag in the smart mattress can be adjusted in a timely and accurate manner according to the real-time sleeping position of the user in the sleep state.

[0100] In an embodiment of the present application, by obtaining the sleeping position information of the user, the real-time sleeping position is determined according to the sleeping position information, and according to the real-time sleeping position, the air pressure value of each airbag in the smart mattress is adjusted to adjust the softness and hardness of the smart mattress.

[0101] The sleeping posture information includes the position information and temperature information of the user when the user is located on the smart mattress, and the temperature information is used to indicate the temperature in the area on the smart mattress corresponding to the position information.

[0102] As can be seen from the above, in the process of adjusting the air pressure value of each airbag in the smart mattress, the user does not need to operate other input devices or other communication devices to trigger the corresponding control instructions, that is, the method provided by the present application can realize automatic adjustment of the air pressure value of the airbag in the smart mattress. In this way, even if the user enters the sleep state, the air pressure value of each airbag in the smart mattress can be adjusted in a timely and accurate manner according to the real-time sleeping posture of the user in the sleep state.

[0103] Moreover, when adjusting the smart mattress, the user's real-time sleeping posture is determined based on the position information and temperature information of the user when the user is on the smart mattress, and the air bags in the smart mattress are adjusted. That is to say, the mattress control method provided by the present application can adjust the air bags in the smart mattress to the air pressure value that best matches or is most suitable for the user's current sleeping posture, thereby improving the comfort of the user lying on the smart mattress.

[0104] In this way, effective sleep assistance can be provided to users.

[0105] For a possible implementation, see Figure 2 , get the user's sleeping position information, including:

[0106] Step 1004: Acquire multiple coordinate data corresponding to the user to obtain the location information.

[0107] Optionally, the coordinate data may include coordinate values ​​of various points on the contact surface between the user and the smart mattress.

[0108] Specifically, the coordinate data refers to coordinate values ​​in a preset coordinate system. The preset coordinate system may be any possible three-dimensional coordinate system, and the coordinate origin and each coordinate axis of the preset coordinate system may be set according to actual needs.

[0109] Generally, in order to simplify calculations and reduce the processing pressure of the control unit and / or the processor, a first coordinate axis (z axis) can be established in the vertical direction, and any two directions perpendicular to the vertical direction can be used as the second coordinate axis and the third coordinate axis, respectively. Moreover, the second coordinate axis is perpendicular to the third coordinate axis.

[0110] In addition, if the smart mattress is rectangular, then one vertex of the smart mattress can be used as the coordinate origin, and the length direction of the smart mattress can be used as the second coordinate axis (x-axis) and the width direction can be used as the third coordinate axis (y-axis), which can further simplify the calculation.

[0111] Step 1005: Obtain the target temperature value at the position corresponding to each coordinate data to obtain temperature information.

[0112] Optionally, the target temperature value refers to the temperature value of each point on the contact surface between the user and the smart mattress.

[0113] Generally, the target temperature value corresponds to each coordinate data, that is, the point corresponding to each coordinate data includes the point corresponding to each target temperature value.

[0114] It is worth noting that through steps 1004 and 1005, the coordinate data and target temperature value of each point on the contact surface between the user and the smart mattress can be accurately obtained, and then the position information and the temperature information can be accurately obtained, which can improve the accuracy of determining the above-mentioned real-time sleeping position based on the position information and the temperature information.

[0115] In order to explain the specific method of obtaining the coordinate data and the target temperature value, several possible implementation methods are provided below for detailed description:

[0116] For a possible implementation, see Figure 3 The smart mattress is equipped with multiple acceleration sensors, each of which is arranged on the top surface of the smart mattress (the side in contact with the user's body and facing away from the ground). Generally, the acceleration sensors can be distributed in the smart mattress in a matrix shape, specifically along Figure 3 The "horizontal" and "vertical" positions are equidistant, which is not limited in the present application.

[0117] exist Figure 3 Based on Figure 4 , obtain multiple coordinate data corresponding to the user, including:

[0118] Step 1006: Obtain the three-dimensional acceleration detected by each acceleration sensor.

[0119] Specifically, obtaining the three-dimensional acceleration detected by each acceleration sensor may refer to obtaining the three-dimensional acceleration detected by some acceleration sensors whose detection values ​​have changed among all the acceleration sensors installed on the smart mattress. That is, if the detection value of a certain acceleration sensor has not changed, it means that the user lying on the smart mattress has not affected this acceleration sensor, and therefore, the detection value of this acceleration sensor has nothing to do with the coordinate data of the user.

[0120] Generally speaking, the acceleration sensors installed on the smart mattress can detect a non-zero value only when the user turns over or moves. If the user is not on the smart mattress or lies still on the smart mattress, the detection values ​​of the acceleration sensors are all 0.

[0121] For example, when determining whether the detection value of a certain acceleration sensor is irrelevant to the coordinate data of the user, the determination may be made according to the following situations:

[0122] In the first case, if the user is not lying on the smart mattress, the detection values ​​of each acceleration sensor may be 0, and the detection values ​​of each acceleration sensor are invalid (ie, irrelevant to the coordinate data).

[0123] In the second case, if the user lies on the smart mattress but is stationary and does not turn over or move, the detection values ​​of each acceleration sensor are also 0, so the detection values ​​of each acceleration sensor are invalid (ie, irrelevant to the coordinate data).

[0124] In the third case, if the user lies on the smart mattress and turns over or moves, pressure and / or deformation will be generated on the smart mattress when the user turns over or moves. In this case, the detection values ​​of each acceleration sensor that can detect the acceleration change are valid (that is, related to the coordinate data), and the detection values ​​of the acceleration sensors that do not detect the acceleration change (that is, each acceleration sensor with a detection value of 0) are also invalid.

[0125] In simple terms, when the user lies on the smart mattress and performs actions such as turning over or moving, the unchanged detection value has nothing to do with the user's coordinate data.

[0126] Optionally, each acceleration sensor may be a three-dimensional acceleration sensor for detecting acceleration in the first coordinate axis direction, the second coordinate axis direction, and the third coordinate axis direction.

[0127] The three-dimensional acceleration may include the acceleration G in the direction of the first coordinate axis. z , the acceleration G in the direction of the second coordinate axis x , and the acceleration G in the direction of the third coordinate axisy .

[0128] Step 1007: Based on each three-dimensional acceleration, determine the angle between any two adjacent acceleration sensors in each acceleration sensor and the chord length of the curve.

[0129] Specifically, determining the angle and the chord length of the curve between any two adjacent acceleration sensors in each acceleration sensor may specifically refer to determining the angle and the chord length of the curve between any two adjacent acceleration sensors in the direction of the second coordinate axis, and the angle and the chord length of the curve between any two adjacent acceleration sensors in the direction of the third coordinate axis.

[0130] Optionally, the angle refers to the angle between the first straight line and the second straight line.

[0131] The first straight line is a straight line perpendicular to the bottom surface of one of the two adjacent acceleration sensors, and the second straight line is a straight line perpendicular to the bottom surface of the other of the two adjacent acceleration sensors. The bottom surface of the acceleration sensor is the contact surface between the acceleration sensor and the smart mattress.

[0132] Optionally, the chord length of the curve may refer to the length of any two adjacent acceleration sensors on the connecting curve when the smart mattress is deformed by pressure from a user.

[0133] For example, see Figure 3 , Figure 3 Sensor S n 、Sensor S n+1 、Sensor S n+2 There are three acceleration sensors distributed along the y-axis. n 、Sensor S n+1 As an example, we will introduce the calculation of the angle θ and the chord length L of the curve. n way.

[0134] When calculating the angle θ, first obtain Sensor S n 、Sensor S n+1 Detected three-dimensional acceleration G x , G y , G z , and respectively Sensor S n 、Sensor S n+1 The detected three-dimensional acceleration is substituted into the following formula to obtain the pitch angle pitch of each acceleration sensor.

[0135]

[0136] Sensor S n 、Sensor S n+1 The pitch angle is recorded as pitch n and pitch n+1 , and substitute it into the following formula to get Sensor S n 、Sensor S n+1 The angle θ between n , that is, the angle between the first straight line L1 and the second straight line L2. It should be noted that the angle θ n Usually with Sensor S n+1 The corresponding value is used to calculate Sensor S n+1 The curve chord length and coordinate data.

[0137] θ n =pitch n -pitch n+1

[0138] Similarly, the angle between the acceleration sensors distributed along the x-axis can be calculated in a similar way to the above two formulas, but when calculating the pitch angle, the numerator in the formula needs to be replaced with G x This embodiment of the present application does not limit this.

[0139] In addition, when calculating the curve chord length L n When Sensor S is obtained n 、Sensor S n+1 The curve length Arc between them is a fixed value determined according to the installation position of each acceleration sensor.

[0140] Then Sensor S n 、Sensor S n+1 Substitute the curve length Arc and the angle θ between them into the following formula to obtain the curve chord length L n .

[0141]

[0142] Step 1008: Determine the position information based on the included angle and the chord length of the curve.

[0143] Optionally, the location information is used to indicate the three-dimensional coordinates of each acceleration sensor in the area corresponding to the smart mattress and the user, and the three-dimensional coordinates also refer to the coordinate values ​​in the above-mentioned preset coordinate system.

[0144] The area of ​​the smart mattress corresponding to the user refers to the area where the user contacts the smart mattress when the user is on the smart mattress, or the area where the smart mattress generates pressure and / or deformation.

[0145] Specifically, the three-dimensional coordinates of one of the accelerations may be calculated according to the included angle between any two adjacent acceleration sensors and the chord length of the curve.

[0146] For example, see Figure 3 , continue with Sensor S n 、Sensor S n+1 As an example, we introduce the calculation of SensorS n+1 The three-dimensional coordinates (x n ,y n ,z n ) method.

[0147] The pitch angle pitch, angle θ, and curve chord length L calculated above are n Substitute into the following formulas respectively:

[0148]

[0149] In this way, Sensor S can be calculated n+1 The two-dimensional coordinates (x n ,z n ).

[0150] Similarly, the two-dimensional coordinates (y) of the acceleration sensors distributed along the x-axis can also be calculated in a similar manner as above. n ,z n ), but it is necessary to substitute the pitch angle, angle, curve chord length and other parameters between the acceleration sensors distributed along the x-axis. The present application embodiment does not limit this.

[0151] Among them, calculate y n The formula for the coordinate value can be as follows:

[0152]

[0153] It is worth noting that after calculating the two-dimensional coordinates (x n ,z n ) and two-dimensional coordinates (y n ,z n ) then we can get the three-dimensional coordinates (x n ,y n ,z n ).

[0154] For a possible implementation, see Figure 5The smart mattress is also equipped with a plurality of temperature sensors W, each temperature sensor W may be arranged in one-to-one correspondence with each acceleration sensor Sensor, or each temperature sensor W may be arranged in one-to-one correspondence with at least some acceleration sensors Sensor, which is not limited in the embodiment of the present application.

[0155] exist Figure 5 Based on Figure 6 , obtain the target temperature value at the corresponding position of each three-dimensional coordinate, including:

[0156] Step 1009: Acquire the temperature value detected by each temperature sensor.

[0157] Optionally, when executing step 1009, specifically, the temperature values ​​detected by all the temperature sensors W on the smart mattress are obtained.

[0158] Step 1010: The temperature value detected by the temperature sensor corresponding to the three-dimensional coordinates of each acceleration sensor is used as the target temperature value.

[0159] Optionally, the step of taking the temperature value detected by the temperature sensor corresponding to the three-dimensional coordinates of each acceleration sensor as the target temperature value may specifically refer to: first determining some acceleration sensors whose detection values ​​have changed among all the acceleration sensors of the smart mattress, and taking the temperature values ​​detected by the temperature sensors corresponding to these acceleration sensors as the target temperature value.

[0160] For example, see Figure 5 If the detection values ​​of the two acceleration sensors Sensor at the 5th row and 4th column and the 6th row and 2nd column in each acceleration sensor Sensor change, then the temperature values ​​detected by the two temperature sensors W at the 5th row and 4th column and the 6th row and 2nd column in each temperature sensor W will be used as the target temperature value. This embodiment of the application will not be elaborated on this.

[0161] Moreover, after obtaining the target temperature value, a temperature change diagram of the user on the smart mattress can be generated based on the target temperature value (see Figure 5 As shown in the diagram on the right of the figure, since the temperature values ​​of different parts of the body are generally different, for example, the chest cavity has the highest temperature and the lower limbs have the lowest temperature, the contours of different parts of the user's body can be distinguished by different colors. Furthermore, the approximate contour of the user lying on the smart mattress can also be determined based on the temperature information and the position information, so as to more accurately identify the user's real-time sleeping position.

[0162] It is worth noting that, since the target temperature value only includes the temperature value corresponding to the three-dimensional coordinates of each acceleration sensor, it can be ensured that the target temperature value and / or the temperature information can indicate the approximate contact range between the user and the smart mattress.

[0163] In a possible manner, adjusting the air pressure value of each airbag in the smart mattress according to the real-time sleeping posture also includes:

[0164] Get the preset gear position and the maximum air pressure value of each airbag.

[0165] The air pressure value of each air bag is adjusted according to the real-time sleeping posture, the preset gear position, and the maximum air pressure value of each air bag.

[0166] Optionally, the preset gear is a soft or hard gear set by the user according to actual needs, for example, the preset gear may include soft, hard, and medium, a total of three gears. This embodiment of the application does not limit this.

[0167] The preset gear can be used to indicate the degree of softness or hardness desired by the user.

[0168] Exemplarily, when the real-time sleeping posture and the maximum air pressure value of each airbag are the same, the air pressure value of each airbag when the preset gear is "hard" will be greater than the air pressure value of each airbag in the smart mattress when the preset gear is "moderate", and the air pressure value of each airbag when the preset gear is "moderate" will be greater than the air pressure value of each airbag in the smart mattress when the preset gear is "soft".

[0169] In this way, the air bags in the smart mattress can be adjusted based on the softness or hardness desired by the user.

[0170] In one possible manner, the air pressure value of each airbag can also be adjusted according to the user's medical history, the real-time sleeping posture, the preset gear position, and the maximum air pressure value of each airbag.

[0171] The medical history may specifically refer to whether the user has a history of spinal disease. Generally, users with a history of spinal disease and those without a history of spinal disease require different support strengths for different parts of the body from the smart mattress. Therefore, the airbags of various parts of the smart mattress can be adaptively adjusted according to the user's medical history.

[0172] In a possible implementation, determining the real-time sleeping posture according to the sleeping posture information includes:

[0173] The real-time sleeping posture is obtained according to the sleeping posture information and the physiological information of the user.

[0174] Optionally, the physiological information may be input in advance by the user, and the physiological information includes at least one of the following: height, weight, and waist-to-hip ratio.

[0175] It is worth noting that because different users have different physiological information, it is possible that the physiological information of the user will affect the process of determining the real-time sleeping position, resulting in the possibility of errors in determining the real-time sleeping position based only on the sleeping position information. Therefore, obtaining the real-time sleeping position based on the sleeping position information and the physiological information can further improve the accuracy of the real-time sleeping position, and further improve the accuracy of adjusting each airbag based on the real-time sleeping position.

[0176] As can be seen from the above, the mattress control method provided in the embodiment of the present application can also adaptively adjust each airbag according to the user's medical history and other physiological information, which can also improve the flexibility and practicality of the mattress control method.

[0177] In order to better explain the above-mentioned way of adjusting each airbag according to the real-time sleeping posture, a possible example is provided below: the real-time sleeping posture includes lying flat, lying on the left side, lying on the right side, and lying prone, the preset gears include soft, hard, and moderate, the physiological information includes the user's spinal history, and the smart mattress is respectively provided with airbags in four parts of the upper back, lower back, buttocks, and legs, and the peak air pressure of each airbag is P.

[0178] After identifying the real-time sleeping posture, the air pressure value of each airbag can be adjusted according to the different real-time sleeping postures, the preset gear selected by the user, and the maximum air pressure value of each airbag. The specific scheme can be seen in the table below. When adjusting the air pressure value of each airbag, select the corresponding value in the table below, multiply it by %, and then multiply it by the maximum air pressure value P, and you can get the air pressure value of each airbag.

[0179]

[0180] It can be understood that, taking the adjustment parameters [60, 70, 80, 85] when the preset gear is soft, there is no history of spinal disease, and the real-time sleeping posture is lying flat as an example, that is, the air pressure values ​​of the airbags in the upper back, lower back, buttocks, and legs can be adjusted to [60%*P, 70%*P, 80%*P, 85%*P] respectively.

[0181] In a possible implementation, determining the real-time sleeping posture according to the sleeping posture information includes:

[0182] The sleeping posture information is input into a pre-trained target sleeping posture recognition model to obtain the real-time sleeping posture.

[0183] Optionally, the target sleeping posture recognition model may be the above-mentioned neural network model for determining the real-time sleeping posture according to the position information and temperature information.

[0184] The target sleeping posture recognition model can be any possible type of model, such as a 3D-CNN model trained based on a self-distillation training algorithm, or any other possible model, which is not limited in the embodiments of the present application.

[0185] It is worth noting that since the neural network model can be trained multiple times with a larger amount of data, the sleeping position information is processed by the target sleeping position recognition model to identify the real-time sleeping position, which can further improve the accuracy of determining the real-time sleeping position.

[0186] In a possible implementation, the sleeping posture information is input into a pre-trained target sleeping posture recognition model to obtain the real-time sleeping posture, including:

[0187] The sleeping posture information and the physiological information of the user are input into the target sleeping posture recognition model to obtain the real-time sleeping posture.

[0188] It is worth noting that obtaining the real-time sleeping position based on the sleeping position information and the physiological information can minimize the impact of the user's physiological information on the process of determining the real-time sleeping position, thereby improving the accuracy of adjusting each airbag according to the real-time sleeping position.

[0189] In order to illustrate the process of training the target sleeping posture recognition model, the present application embodiment also provides a possible implementation method, see Figure 7 , the method further comprises:

[0190] Step 1011: Obtain a sample data set.

[0191] Optionally, the sample data set includes multiple groups of sample sleeping posture information and sample labels corresponding to each group of sample sleeping posture information.

[0192] Optionally, the sample sleeping posture information includes sample position information of the user on the smart mattress and sample temperature information of the smart mattress. The sample sleeping posture information may be collected and sorted by relevant technical personnel before training the model.

[0193] Optionally, the sample label is used to indicate the sleeping posture of the user, and each sample label is associated with a set of sample sleeping posture information. The sample label can be obtained by an image acquisition device (such as a camera or a webcam), see Figure 8 The sample label can specifically indicate a corresponding set of sample sleeping posture information, specifically lying flat, lying on the left side, lying on the right side, or lying prone, so that the neural network model can learn.

[0194] Step 1012: input the sample data set into an initial sleeping posture recognition model, and the initial sleeping posture recognition model outputs first recognition information and second recognition information.

[0195] Optionally, the initial sleeping posture recognition model is a neural network model that has not yet started training or has not been trained to completion.

[0196] Optionally, the first identification information and the second identification information respectively include a predicted label and a probability value corresponding to the predicted label.

[0197] The predicted label is the sleeping position output by the initial sleeping position recognition model according to the sample sleeping position information, and the probability value corresponding to the predicted label is the possibility of various sleeping positions output by the initial sleeping position recognition model.

[0198] For example, the first identification information may be [lying flat 80%, lying on the left side 10%, lying on the right side 10%, lying prone 0%], and the first identification information may be [lying flat 78%, lying on the left side 9%, lying on the right side 11%, lying prone 1%].

[0199] Step 1013: Determine loss information of the initial sleeping posture recognition model based on the first recognition information and the second recognition information.

[0200] Optionally, the loss information may be used to indicate the accuracy of the initial sleeping posture recognition model.

[0201] Exemplarily, the loss information can be obtained based on any possible loss function, such as cross entropy loss, probability divergence loss, etc., which is not limited in the embodiments of the present application.

[0202] Step 1014: iteratively correct the initial sleeping position recognition model according to the loss information of the initial sleeping position recognition model until the initial sleeping position recognition model reaches a preset condition and the training is terminated, and the initial sleeping position recognition model that reaches the preset condition is used as the target sleeping position recognition model.

[0203] Optionally, the preset condition may be set by relevant technical personnel according to actual needs.

[0204] For example, the preset condition may be that the loss information is less than a preset threshold, or that the first identification information and the second identification information output by the initial sleeping posture recognition model are sufficiently accurate, and the error between the first identification information and the second identification information is sufficiently small. This embodiment of the application does not limit this.

[0205] It is worth noting that after iterative training and correction of the initial sleeping position recognition model, the accuracy of the initial sleeping position recognition model in predicting the user's sleeping position can be improved. Due to a large amount of iterative training of the sample sleeping position information, the performance of the target sleeping position recognition model in predicting the real-time sleeping position is greatly improved compared with the initial sleeping position recognition model.

[0206] It should be noted that, after the above steps, a target sleeping position recognition model that can accurately predict the real-time sleeping position corresponding to different sleeping position information can be obtained. After the user or relevant technical personnel input any sleeping position information into the target sleeping position recognition model, the target sleeping position recognition model can output the real-time sleeping position corresponding to the any sleeping position information, without spending a lot of time and manual work to build and look up tables for each sleeping position information, thus, the efficiency of determining the real-time sleeping position can be improved.

[0207] For a possible implementation, see Fig. 9 The initial sleeping posture recognition model includes a first sub-model and at least one second sub-model, the first sub-model is used to output the first recognition information, and the second sub-model is used to output the second recognition information.

[0208] In this embodiment, the initial sleeping posture recognition model may be a 3D-CNN model, and the initial sleeping posture recognition model may be trained based on a self-distillation training algorithm. Therefore, the first sub-model may be a teacher model, and each second sub-model may be a student model.

[0209] from Fig. 9 As can be seen from the structural diagram of the sleeping posture recognition model shown, in order to implement the self-distillation algorithm, each convolution block is followed by a BottleNeck module, a fully connected layer, and a softmax layer, that is, a classifier. Each shallow classifier is a student model, and the 3D-CNN model and its deepest output are used together as a teacher model. This classifier is only used to implement the distillation algorithm. After the training is completed, the teacher model can be used directly, or a student model with better performance can be used to output the real-time sleeping posture. This embodiment of the present application does not limit this.

[0210] Moreover, in addition to integrating the sleeping position information into the initial sleeping position recognition model of the present application, the present solution adds three physiological data, namely height, weight, and waist-to-hip ratio, to the fully connected layer to improve the recognition accuracy.

[0211] Moreover, the convolution block ConvBlock and the BottleNeck module in the 3D-CN model may respectively include multiple convolutional layers, batch normalization layers, ReLU layers, maximum pooling layers and / or any other possible processing layers, which is not limited in the embodiments of the present application.

[0212] Furthermore, the sample data set is input into an initial sleeping posture recognition model, and the initial sleeping posture recognition model outputs first recognition information and second recognition information, further comprising:

[0213] The sample data set is input into the first sub-model.

[0214] The first sub-model performs feature extraction on the sample sleeping posture information in the sample data set to obtain at least one feature information and a first feature vector.

[0215] Optionally, the feature information is the feature information obtained after the first sub-model performs at least one convolution on the sample sleeping posture information. The first feature vector is the feature vector output to the full connection after the first sub-model completes the convolution processing on the sample sleeping posture information, and the first feature vector can generally be a 1-dimensional vector.

[0216] The first sub-model outputs the first recognition information based on the first feature vector.

[0217] The feature information obtained by the first sub-model is output to each second sub-model respectively.

[0218] Each second sub-model performs feature extraction on each feature information to obtain a second feature vector.

[0219] Optionally, the second feature vector may be obtained after the BottleNeck module performs convolution and normalization processing on each feature information.

[0220] The second sub-models respectively output the second identification information based on the second feature vector.

[0221] In this way, each second sub-model can be trained based on the feature information output by the first sub-model. It can be seen that the initial sleeping posture recognition model requires fewer computing resources during the training and reasoning process, avoids the use of more complex models, and improves the training efficiency.

[0222] In a possible implementation, the first sub-model outputs the first identification information based on the first feature vector, including:

[0223] The first sub-model performs feature integration on the first feature vector and the physiological information of the user to obtain the first identification information.

[0224] Optionally, the first feature vector (eg, a 1*N vector) and the physiological information (eg, a 1*4 vector) may be processed based on the fully connected layer T1, and finally the first identification information may be output.

[0225] In a possible implementation, each second sub-model outputs the second identification information based on the second feature vector, including.

[0226] The second sub-models respectively perform feature integration on the second feature vector and the physiological information of the user to obtain the second identification information.

[0227] Optionally, the second feature vector and the physiological information may be processed based on the fully connected layer S1 or S2 to finally output the second identification information.

[0228] In this way, physiological information can be added to the training process to improve the accuracy of the target sleeping posture recognition model in performing sleeping posture recognition based on physiological information.

[0229] In a possible implementation, determining the loss information of the initial sleeping posture recognition model based on the first recognition information and the second recognition information includes:

[0230] A first loss value between the first identification information and the sample label is determined, and a second loss value between the second identification information and the sample label is determined.

[0231] A first loss result is obtained based on the first loss value and the second loss value.

[0232] Optionally, the first loss result is a cross entropy loss.

[0233] A second loss result between the first identification information and the second identification information is determined.

[0234] Optionally, the second loss result is a predicted probability divergence loss.

[0235] A third loss result between the first identification information and the second identification information is determined.

[0236] Optionally, the third loss result is the deepest feature vector loss.

[0237] The loss information of the initial sleeping posture recognition model is obtained according to the first loss result, the second loss result, and the third loss result.

[0238] Exemplarily, the loss information can be calculated by the following loss function:

[0239]

[0240] Wherein, N represents the total number of the first sub-model and the second sub-model, which is 3 in this embodiment. In addition, N can also represent the number of sample data sets. y and q represent the sample label and the probability corresponding to the predicted label, respectively. α i and λ i Represents the weight hyperparameters of different models in the loss.

[0241] In general, for the first submodel, α i and λ iAll are set to 0 as the benchmark of the loss function. It can be observed from the loss function that L1 is the first loss result, which is used to represent the cross entropy loss between the output of each sub-model and the sample label, L2 is the second loss result, which is used to represent the predicted probability divergence loss between each student model and the teacher model, and L3 is the third loss result, which is used to represent the deepest feature vector loss between each second sub-model and the first model.

[0242] It can be understood that in this embodiment, two second sub-models are used as an example for explanation, but it does not mean that only two can be set in the initial sleeping position recognition model and / or the template sleeping position recognition model provided in the embodiment of the present application. In actual application, 1, 3, or more second sub-models can also be set, and the embodiment of the present application does not limit this.

[0243] It should be understood that, although the steps in the above-mentioned flowcharts are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a part of the steps in the above-mentioned flowcharts may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0244] The following describes a device, a smart mattress, and a computer-readable storage medium for executing the mattress control method provided in the present application. The specific implementation process and technical effects are described above and will not be repeated below.

[0245] The embodiment of the present application provides a mattress control device, which includes various modules and various units included in each module, which can be implemented by a processor; of course, it can also be implemented by a specific logic circuit; in the implementation process, the processor can be a central processing unit (CPU), a microprocessor (MPU), a digital signal processor (DSP) or a field programmable gate array (FPGA), etc.

[0246] The mattress control device is applied to a smart mattress, which includes at least one air bag; see Fig.10 , the device comprises:

[0247] An acquisition module 201 is used to acquire sleeping posture information of a user, where the sleeping posture information includes location information and temperature information of the user when the user is located on the smart mattress, where the temperature information is used to indicate the temperature in an area on the smart mattress corresponding to the location information;

[0248] A determination module 202, configured to determine a real-time sleeping posture according to the sleeping posture information;

[0249] The adjustment module 203 is used to adjust the air pressure value of each air bag in the smart mattress according to the real-time sleeping posture, so as to adjust the softness and hardness of the smart mattress.

[0250] Optionally, the device may further include:

[0251] A training module is used to obtain a sample data set, input the sample data set into an initial sleeping position recognition model, and have the initial sleeping position recognition model output a first recognition information and a second recognition information, determine the loss information of the initial sleeping position recognition model based on the first recognition information and the second recognition information, and iteratively correct the initial sleeping position recognition model according to the loss information of the initial sleeping position recognition model until the initial sleeping position recognition model meets the preset conditions and the training is terminated, and the initial sleeping position recognition model that meets the preset conditions is used as the target sleeping position recognition model.

[0252] The description of the above device embodiment is similar to the description of the above method embodiment, and has similar beneficial effects as the method embodiment. For technical details not disclosed in the device embodiment of the present application, please refer to the description of the method embodiment of the present application for understanding.

[0253] It should be noted that in the embodiments of this application Fig.10 The division of modules in the mattress control device shown is schematic and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional unit in each embodiment of the present application may be integrated into a processing unit, or may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional unit. It may also be implemented in the form of a combination of software and hardware.

[0254] It should be noted that in the embodiments of the present application, if the above method is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the embodiment of the present application can be essentially or partly embodied in the form of a software product that contributes to the relevant technology. The computer software product is stored in a storage medium, including several instructions to enable an electronic device to execute all or part of the method of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM), a disk or an optical disk. In this way, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0255] The embodiment of the present application provides a smart mattress. The internal structure diagram of the smart mattress can be as follows: Fig.11 As shown. The smart mattress includes a processor and a memory connected via a system bus. The processor of the smart mattress is used to provide computing and control capabilities. The memory of the smart mattress includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the smart mattress is used to store data. When the computer program is executed by the processor, the above method is implemented.

[0256] In addition, the smart mattress may also include a network interface for communicating with an external terminal via a network connection.

[0257] An embodiment of the present application provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the method provided in the above embodiment are implemented.

[0258] An embodiment of the present application provides a computer program product including instructions, which, when executed on a computer, enables the computer to execute the steps of the method provided in the above method embodiment.

[0259] Those skilled in the art will understand that Fig.11 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the smart mattress to which the solution of the present application is applied. The specific smart mattress may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0260] In one embodiment, the mattress control device provided by the present application can be implemented in the form of a computer program. The computer program can be Fig.11 The memory of the smart mattress can store various program modules constituting the above-mentioned device. The computer program composed of various program modules enables the processor to execute the steps of the method of each embodiment of the present application described in this specification.

[0261] It should be noted here that the description of the above storage medium and device embodiments is similar to the description of the above method embodiments, and has similar beneficial effects as the method embodiments. For technical details not disclosed in the storage medium, storage medium and device embodiments of this application, please refer to the description of the method embodiments of this application for understanding.

[0262] It should be understood that "one embodiment" or "an embodiment" or "some embodiments" mentioned throughout the specification means that specific features, structures or characteristics related to the embodiment are included in at least one embodiment of the present application. Therefore, "in one embodiment" or "in one embodiment" or "in some embodiments" appearing throughout the specification may not necessarily refer to the same embodiment. In addition, these specific features, structures or characteristics can be combined in one or more embodiments in any suitable manner. It should be understood that in various embodiments of the present application, the size of the sequence number of the above-mentioned processes does not mean the order of execution, and the execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present application. The above-mentioned sequence numbers of the embodiments of the present application are only for description and do not represent the advantages and disadvantages of the embodiments. The above description of each embodiment tends to emphasize the differences between the various embodiments, and the same or similar aspects can be referenced to each other. For the sake of brevity, this article will not repeat them.

[0263] The term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there may be three relationships. For example, object A and / or object B can represent three situations: object A exists alone, object A and object B exist at the same time, and object B exists alone.

[0264] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0265] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The embodiments described above are only schematic. For example, the division of the module is only a logical function division. There may be other division methods in actual implementation, such as: multiple modules or components can be combined, or can be integrated into another system, or some features can be ignored, or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the device or module can be electrical, mechanical or other forms.

[0266] The modules described above as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical modules; they may be located in one place or distributed on multiple network units; some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.

[0267] In addition, all functional modules in the embodiments of the present application may be integrated into one processing unit, or each module may be a separate unit, or two or more modules may be integrated into one unit; the above-mentioned integrated modules may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0268] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: mobile storage devices, read-only memories (ROM), magnetic disks or optical disks, etc., various media that can store program codes.

[0269] Alternatively, if the above-mentioned integrated unit of the present application is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present application can essentially or in other words, the part that contributes to the relevant technology can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for an electronic device to execute all or part of the method of each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROMs, magnetic disks or optical disks.

[0270] The methods disclosed in several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

[0271] The features disclosed in several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

[0272] The features disclosed in several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

[0273] The above is only an implementation method of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0274] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A mattress control method, characterized in that: Applied to a smart mattress, the smart mattress includes at least one airbag; the method includes: Acquire sleeping posture information of the user, the sleeping posture information including position information and temperature information of the user when the user is located on the smart mattress, the temperature information being used to indicate the temperature in an area on the smart mattress corresponding to the position information; determining a real-time sleeping posture according to the sleeping posture information; According to the real-time sleeping posture, the air pressure value of each air bag in the smart mattress is adjusted to adjust the softness and hardness of the smart mattress.

2. The mattress control method according to claim 1, characterized in that: The step of obtaining the sleeping posture information of the user includes: Acquire a plurality of coordinate data corresponding to the user to obtain the position information; The target temperature value at the position corresponding to each coordinate data is obtained to obtain temperature information.

3. The mattress control method according to claim 2, characterized in that: The smart mattress is equipped with a plurality of acceleration sensors; the step of obtaining a plurality of coordinate data corresponding to the user includes: Acquiring the three-dimensional acceleration detected by each of the acceleration sensors; Based on the three-dimensional accelerations, determining the angle between any two adjacent acceleration sensors in the acceleration sensors and the chord length of the curve; The position information is determined based on the angle and the chord length of the curve, and the position information is used to indicate the three-dimensional coordinates of each of the acceleration sensors in the area corresponding to the smart mattress and the user.

4. The mattress control method according to claim 3, characterized in that: The smart mattress is also equipped with a plurality of temperature sensors; the step of obtaining the target temperature value at the position corresponding to each coordinate data includes: Acquire the temperature value detected by each of the temperature sensors; The temperature value detected by the temperature sensor corresponding to the three-dimensional coordinates of each of the acceleration sensors is used as the target temperature value.

5. The mattress control method according to claim 1, characterized in that: The determining the real-time sleeping posture according to the sleeping posture information includes: Obtaining the real-time sleeping posture according to the sleeping posture information and the physiological information of the user; The physiological information includes at least one of the following: height, weight, and waist-to-hip ratio.

6. The mattress control method according to claim 1, characterized in that: The determining the real-time sleeping posture according to the sleeping posture information includes: The sleeping posture information is input into a pre-trained target sleeping posture recognition model to obtain a real-time sleeping posture.

7. The mattress control method according to claim 6, characterized in that: The step of inputting the sleeping posture information into a pre-trained target sleeping posture recognition model to obtain a real-time sleeping posture includes: The sleeping posture information and the physiological information of the user are input into the target sleeping posture recognition model to obtain the real-time sleeping posture, wherein the physiological information includes at least one of the following: height, weight, and waist-to-hip ratio.

8. The mattress control method according to claim 6, characterized in that: The method further comprises: Acquire a sample data set; the sample data set includes multiple sets of sample sleeping posture information and sample labels corresponding to each set of the sample sleeping posture information; the sample sleeping posture information includes sample position information of the user on the smart mattress and sample temperature information of the smart mattress, and the sample label is used to indicate the sleeping posture of the user; Inputting the sample data set into an initial sleeping posture recognition model, and the initial sleeping posture recognition model outputs first recognition information and second recognition information; the first recognition information and the second recognition information both include a predicted label and a probability value corresponding to the predicted label; determining loss information of the initial sleeping posture recognition model based on the first recognition information and the second recognition information; According to the loss information of the initial sleeping position recognition model, the initial sleeping position recognition model is iteratively corrected until the initial sleeping position recognition model reaches a preset condition and the training is terminated, and the initial sleeping position recognition model that reaches the preset condition is used as the target sleeping position recognition model.

9. The mattress control method according to claim 8, characterized in that: The determining, based on the first recognition information and the second recognition information, loss information of the initial sleeping posture recognition model includes: Determine a first loss value between the first identification information and the sample label, and determine a second loss value between the second identification information and the sample label; Obtain a first loss result based on the first loss value and the second loss value, where the first loss result is a cross entropy loss; Determine a second loss result between the first identification information and the second identification information, the second loss result being a predicted probability divergence loss; Determine a third loss result between the first identification information and the second identification information, the third loss result being the deepest feature vector loss; The loss information of the initial sleeping posture recognition model is obtained according to the first loss result, the second loss result, and the third loss result.

10. A smart mattress, characterized in that: include: A memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, the steps of the method described in any one of claims 1 to 9 are implemented.

Images