Massage mattress pressure adjusting method and system based on air bag control

Abstract

The invention provides a massage mattress pressure adjusting method and system based on air bag control, and relates to the technical field of massage appliances, and the massage mattress pressure adjusting method comprises the steps that a flexible sensing strain gauge module is used for detecting the state of a massage mattress, and the massage mattress is a bendable massage mattress; if the massage mattress is in a first state, a bent mattress area is determined, and the first state is a bent state; and connecting the first group of air bag control modules corresponding to the bent mattress area, identifying the first group of initial air bag control parameter sets, inputting the first group of initial air bag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting an enhanced air bag control parameter set, and controlling the air bag pressure of the bent mattress area according to the enhanced air bag control parameter set. The technical problem that in the prior art, pressure adjustment mostly depends on a fixed adjustment mode, an intelligent adjustment mechanism is lacked to conduct timely and effective pressure adjustment according to different forms of the mattress, and consequently the comfort of the massage mattress is insufficient is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of massage appliances, and particularly to a method and system for adjusting the pressure of a massage mattress based on airbag control. Background Art

[0002] The massage mattress provides a comfortable massage experience for users through an in-built airbag adjustment technology. By changing the pressure distribution of the airbags, it helps to relieve the physical fatigue and stress of users. However, most of the existing massage mattresses use simple pressure sensors to monitor the pressure distribution of the mattress, and the pressure adjustment mostly relies on fixed adjustment modes. When the mattress is in a bent state, there will be a gap between the human body and the bent part of the mattress, resulting in a reduction in the contact area and a decrease in local pressure. The fixed adjustment mode will affect the adjustment effect of the mattress, and there is a lack of an intelligent adjustment mechanism to perform timely pressure adjustment improvement according to different forms of the mattress. Therefore, the airbag pressure of the mattress may be too high or too low in different states, resulting in insufficient comfort. Summary of the Invention

[0003] This application provides a method and system for adjusting the pressure of a massage mattress based on airbag control, aiming to solve the technical problem that the pressure adjustment in the prior art mostly relies on fixed adjustment modes and lacks an intelligent adjustment mechanism to perform timely and effective pressure adjustment according to different forms of the mattress, resulting in insufficient comfort of the massage mattress.

[0004] In the first aspect disclosed in this application, a method for adjusting the pressure of a massage mattress based on airbag control is provided. The method includes: using a flexible strain gauge module to detect the state of the massage mattress, where the massage mattress is a bendable massage mattress; if the massage mattress is in a first state, determining the bent mattress area, where the first state is a bent state; connecting the first group of airbag control modules corresponding to the bent mattress area, identifying the first group of initial airbag control parameter sets of the first group of airbag control modules, inputting the first group of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting an enhanced airbag control parameter set, and controlling the airbag pressure of the bent mattress area according to the enhanced airbag control parameter set.

[0005] The second aspect disclosed in this application provides a pressure regulation system for a massage mattress based on airbag control. The system is used for the above-mentioned pressure regulation method of the massage mattress based on airbag control. The system includes: a massage mattress state detection module for detecting the state of the massage mattress using a flexible strain gauge module, where the massage mattress is a bendable massage mattress; a bent mattress area determination module for determining the bent mattress area if the massage mattress is in a first state, where the first state is a bent state; an airbag pressure control module for connecting to the first group of airbag control modules corresponding to the bent mattress area, identifying the first group of initial airbag control parameter sets of the first group of airbag control modules, inputting the first group of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting an enhanced airbag control parameter set, and controlling the airbag pressure in the bent mattress area according to the enhanced airbag control parameter set.

[0006] One or more technical solutions provided in this application have at least the following beneficial effects:

[0007] Using a flexible strain gauge module to detect the state of the massage mattress can real-time sense the deformation and pressure changes of the mattress. Through this process, it can accurately determine whether the mattress is currently in a bent state, thus providing data support for subsequent airbag control; when the massage mattress is in a bent state, accurately identify and determine the bent mattress area. By effectively partitioning the bent area of the mattress, targeted airbag control can be carried out, making the airbag adjustment more accurate and efficient. The accurate identification of the bent area helps to avoid unnecessary airbag intervention in the subsequent control process, thus realizing local optimization adjustment; when the mattress is in a bent state, there will be gaps between the human body and the bent part of the mattress, resulting in a decrease in the pressure felt by the human body. By using a control enhancement model to enhance the airbag pressure adjustment in the bent area, these gap areas can be effectively compensated when in the bent state, enabling the mattress to still provide appropriate support and comfort when bent, enhancing the output of the enhanced airbag control parameter set, so that the airbag can accurately control the pressure in this area, thereby optimizing the user's massage experience and ensuring the uniformity and comfort of the massage effect.

[0008] The above description is only an overview of the technical solutions of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically illustrates the specific embodiments of this application. Specific Embodiments

[0009] By providing a pressure adjustment method and system for a massage mattress based on airbag control, embodiments of the present application solve the technical problem in the prior art that pressure adjustment mostly relies on fixed adjustment modes and lacks an intelligent adjustment mechanism to perform timely and effective pressure adjustment according to different forms of the mattress, resulting in insufficient comfort of the massage mattress.

[0010] After introducing the basic principle of the present application, various non-limiting implementation manners of the present application will be specifically introduced below. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0011] Embodiment 1, embodiments of the present application provide a pressure adjustment method for a massage mattress based on airbag control, and the method includes:

[0012] Use a flexible sensing strain gauge module to detect the state of the massage mattress, and the massage mattress is a bendable massage mattress.

[0013] The flexible sensing strain gauge module is composed of an optical fiber grating array and a piezoelectric thin film network. Among them, the optical fiber grating array senses the deformation of the surface of the massage mattress by monitoring the deformation of the optical fiber, and the piezoelectric thin film network monitors the pressure at different positions on the surface of the massage mattress. There is a close relationship between the optical fiber deformation and the pressure data, and these data can be used to judge whether the massage mattress is in a bent state. Specifically, through the optical fiber grating array and the piezoelectric thin film network of the sensing module, optical fiber deformation data and pressure data are collected, and the collected optical fiber deformation data is demodulated to obtain wavelength shift data. At the same time, according to the piezoelectric pressure data, the regional pressure mean value and the regional pressure standard deviation are calculated. By analyzing the wavelength shift data, the regional pressure mean value and the regional pressure standard deviation, the state of the massage mattress is determined, including whether the mattress is in a bent state or a flat state.

[0014] If the massage mattress is in a first state, determine the bent mattress area, where the first state is the bent state.

[0015] According to the obtained data, judge whether the massage mattress is in a bent state through a preset threshold. For example, if the wavelength shift data is greater than a certain preset value, it means that the mattress is in a bent state. At this time, mark that the massage mattress is in the first state. When the mattress is confirmed to be in a bent state, determine the specific bent area, that is, which parts of the mattress are in a bent state. The core of this process is to locate and delimit the bent area of the mattress. Usually, these bent areas are the parts of the mattress that are bent or deformed under external forces. The deformation data obtained by the sensor is used to identify which areas are subjected to bending or pressure concentration, so as to delimit the bent mattress area for the configuration of the airbag control module and pressure adjustment.

[0016] Define the state space and the action space, and use the PPO algorithm to train the reward function based on the state space and the action space to obtain a trained control enhancement model. Among them, the state space includes pressure point distribution and wavelength shift data, and the action space includes airbag pressure adjustment step size and amplitude.

[0017] The state space is a concept in reinforcement learning, referring to the set of all possible states that the system may be in at a certain moment, including pressure point distribution and wavelength shift data. Among them, the pressure point distribution refers to the pressure distribution of different regions on the mattress surface. The pressure distribution of the mattress can be monitored in real time through a piezoelectric film network, which can reflect the pressure state of each point on the mattress; the wavelength shift data is the deformation data obtained through a fiber Bragg grating array, which is closely related to the deformation of the mattress surface. The wavelength shift data can be used to further evaluate the pressure distribution and deformation of the mattress surface, so as to judge the state of the mattress, such as bending, flatness, etc. The state space is jointly composed of the pressure point distribution and wavelength shift data of the mattress, representing the pressure and deformation state of the mattress at a specific moment.

[0018] The action space is another concept in reinforcement learning, referring to the set of all actions that the system can choose, including airbag pressure adjustment step size and amplitude. The step size determines the speed of airbag adjustment, and the amplitude determines the range of pressure change. The inflation volume of the airbag will change with the change of the pressure adjustment step size, and precise pressure adjustment can be achieved by adjusting this step size. The airbag pressure adjustment step size controls the change speed and range of the airbag pressure, and it determines how the system adjusts the airbag pressure in a given state to achieve the required mattress state.

[0019] The PPO algorithm is a reinforcement learning algorithm used to train an agent to optimize its behavioral strategy. In the PPO algorithm, the agent continuously adjusts its strategy through interaction with the environment to maximize the obtained reward. Specifically, the PPO algorithm takes the above-defined state space (pressure point distribution and wavelength shift data) and action space (airbag pressure adjustment step size) as inputs. Through training under these inputs, the PPO algorithm will learn how to select the most appropriate airbag pressure adjustment step size in different mattress states. The reward function is used to evaluate the behavior of the agent. The reward function is usually designed as the feedback obtained by the system after selecting a certain action, that is, adjusting the airbag pressure step size, in the current state. This feedback includes the optimization of the mattress pressure distribution, the improvement of comfort, or the improvement of wavelength shift data, etc. Ideally, the PPO algorithm maximizes the reward function, that is, selects the action that can best adjust the airbag pressure to improve the overall comfort and stability.

[0020] Through multiple trainings, the PPO algorithm gradually optimizes the strategy to learn the best airbag pressure regulation scheme. In each training iteration, the agent selects an action according to the current state space, that is, adjusts the step size of the airbag, and obtains a reward. As the training progresses, the PPO algorithm continuously adjusts the airbag control strategy to maximize the reward, and finally obtains a trained control enhancement model. This model can automatically calculate the appropriate airbag pressure adjustment step size based on the current state of the mattress (pressure point distribution and wavelength offset data), thereby optimizing the comfort and pressure distribution of the mattress.

[0021] Connect the first set of airbag control modules corresponding to the bent mattress area, identify the first set of initial airbag control parameter sets of the first set of airbag control modules, input the first set of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, output the enhanced airbag control parameter sets, and control the airbag pressure of the bent mattress area according to the enhanced airbag control parameter sets.

[0022] The main function of the airbag control module is to adjust the pressure of the mattress by changing the inflation volume of the airbag, thereby affecting the hardness and comfort of the massage mattress. The inflation volume of the airbag directly affects the bending degree and comfort of the mattress. After the bent mattress area is identified, the first set of airbag control modules corresponding to the bent mattress area is recognized. Each airbag control module has its own control parameter set, and these parameters determine the pressure and adjustment method of the airbag, which are used to adjust the pressure of the airbag. Identify the first set of initial airbag control parameter sets of the first set of airbag control modules, including a series of parameters such as airbag inflation volume, pressure increment, and response speed of the airbag. These initial control parameter sets are set according to the initial state of the mattress, such as the degree of bending and pressure distribution.

[0023] Input the first set of initial airbag control parameter sets into the control enhancement model. These initial parameter sets provide a starting point for the model. The model uses the control enhancement learning algorithm to continuously adjust and feedback optimize these parameters. Since once the mattress is bent, there will be a gap between the human body and the mattress at the bent place, and the human feeling will decrease. The pressure can be increased to maintain the pressure feeling at the gap. To achieve this goal, the enhancement learning model continuously adjusts the control parameter set of the airbag through interaction with the simulation environment, so that the airbag pressure in the bent area is optimized.

[0024] In the reinforcement learning model, by continuously adjusting the control parameter set, the model finally outputs an optimized set of enhanced airbag control parameters. This set of parameters is more adaptable to the actual mattress state than the initial parameter set and can adjust the airbag pressure more precisely, thereby providing the best comfort and massage effect. According to the output enhanced airbag control parameter set, the airbag pressure is precisely adjusted. The pressure change of the airbag will directly affect the bending degree and massage effect of the mattress. Specifically, once the mattress is bent, there will be a gap between the human body and the mattress at the bent place, and the human body's feeling will be reduced. The pressure can be increased to maintain the pressure feeling at the gap.

[0025] Furthermore, the flexible strain gauge module includes a fiber Bragg grating array and a piezoelectric film network; wherein, the fiber Bragg grating array is longitudinally arranged on the surface of the massage mattress at a preset interval, and the piezoelectric film network is arranged on the surface of the massage mattress in a 5 cm × 5 cm grid distribution.

[0026] The flexible strain gauge module is a component used to detect the state of the massage mattress, especially the deformation and pressure distribution of the mattress. The module includes a fiber Bragg grating array and a piezoelectric film network. Among them, the fiber Bragg grating array is composed of multiple fiber Bragg gratings. A fiber Bragg grating is a fiber structure composed of reflectors at specific intervals and is used to detect the wavelength shift of light waves. Whenever the fiber deforms, the propagation characteristics of light change, resulting in a shift in the reflection wavelength of the fiber. This change is directly related to the deformation of the mattress; the piezoelectric film network uses the piezoelectric effect to detect the pressure change on the surface of the mattress. When a piezoelectric material is subjected to an external force, it generates an electric charge, thereby generating a corresponding signal, reflecting the pressure distribution on the surface of the mattress.

[0027] The fiber Bragg grating array is arranged along the longitudinal direction of the mattress, that is, the length direction of the mattress. There is a preset interval between each fiber Bragg grating. In this way, a series of fiber Bragg gratings can be covered on the surface of the mattress. Each fiber Bragg grating can independently monitor the deformation of a certain point on the surface of the mattress. The preset interval helps to achieve precise deformation detection and monitoring of pressure fluctuations.

[0028] The piezoelectric film network is evenly arranged on the surface of the mattress, and the grid spacing is 5 cm × 5 cm, that is, the sensing area of each piezoelectric film is 5 cm × 5 cm. This grid layout can ensure that the pressure in each area of the mattress surface can be independently monitored. If the mattress is subjected to a large pressure at certain positions, for example, the user's weight is concentrated at a certain position, the piezoelectric film network can detect this change and provide feedback data.

[0029] Furthermore, using the flexible strain gauge module to detect the state of the massage mattress, the method includes:

[0030] Based on the fiber grating array and piezoelectric film network of the flexible sensing transducer module, optical fiber deformation data and piezoelectric pressure data are respectively obtained; a functional relationship between the Bragg grating wavelength offset and deformation is established, and the optical fiber deformation data is demodulated according to the functional relationship to obtain wavelength offset data Δλ; the regional pressure mean μ and the regional pressure standard deviation σ are calculated based on the piezoelectric pressure data; and the state of the massage mattress is determined based on the wavelength offset data Δλ, the regional pressure mean μ, and the regional pressure standard deviation σ.

[0031] When the mattress surface deforms, the reflected wavelength of the fiber Bragg grating (FBG) shifts. These wavelength shifts are measured to determine the magnitude of the deformation. The FBG array monitors the fiber deformation data at every location on the mattress surface in real time. This data reflects the degree of deformation when the mattress is under pressure, helping to determine the pressure distribution in different areas of the mattress. The piezoelectric film network provides pressure data by monitoring pressure changes in each area of the mattress surface. When pressure is applied to a specific area of the mattress, the piezoelectric film generates an electrical signal reflecting the pressure value in that area. Piezoelectric pressure data represents the real-time pressure values for each area of the mattress. By acquiring this data through the piezoelectric film network, pressure changes on the mattress surface can be monitored.

[0032] A Bragg grating (FBG) is composed of an optical fiber with a periodic structure. It can change the reflected wavelength of light waves based on the fiber's deformation. When the fiber deforms, the period of the FBG changes, causing the reflected wavelength to shift. A functional relationship between the FBG wavelength shift and deformation has been established. This relationship describes the mathematical relationship between fiber deformation and the shift in the reflected wavelength. Based on this functional relationship, the collected fiber deformation data is demodulated and converted into wavelength shift data, Δλ. This data is useful for assessing mattress deformation and pressure distribution.

[0033] The regional pressure mean μ refers to the average pressure in a specific area of the mattress. This is calculated by averaging the pressure values detected by all piezoelectric film networks in the area. The mean provides the overall level of pressure in the area. The average pressure value of the area can be obtained by adding the pressure values of all measurement points in a certain area and dividing it by the number of measurement points.

[0034] The regional pressure standard deviation, σ, reflects the degree of pressure fluctuation within a region. A small standard deviation indicates relatively uniform pressure within the region; a large standard deviation indicates drastic pressure fluctuations within the region. The regional pressure standard deviation is calculated by summing the squares of the differences between the pressure at each measurement point within the region and the regional mean pressure, dividing this sum by the number of measurement points, and taking the square root. This helps assess the pressure uniformity of a mattress.

[0035] Based on the obtained wavelength shift data Δλ, the regional pressure mean μ, and the regional pressure standard deviation σ, comprehensively judge the state of the mattress. For example, a large wavelength shift indicates that the mattress is in a bent state, while a more uniform regional pressure indicates that the mattress is in a flat state.

[0036] Furthermore, judging the state of the massage mattress according to the wavelength shift data Δλ, the regional pressure mean μ, and the regional pressure standard deviation σ, the method includes:

[0037] If the wavelength shift data Δλ is greater than or equal to a preset wavelength shift threshold, and the point pressure P i < μ - 3σ, it is judged that the state of the massage mattress is in the first state, and the first state is the bent state; if the wavelength shift data Δλ is less than the preset wavelength shift threshold or the point pressure P i ≥ μ - 3σ, it is judged that the state of the massage mattress is in the second state, and the second state is the non-bent state.

[0038] The preset wavelength shift threshold is a pre-set standard for judging whether the mattress has undergone a large deformation. If the wavelength shift data is greater than or equal to this threshold, it indicates that the mattress has undergone an obvious deformation or bending. The larger the wavelength shift value, the more obvious the deformation of the mattress; the point pressure P i refers to the pressure value at a certain point on the surface of the mattress. By comparing it with the regional pressure mean and standard deviation, it can be judged whether the pressure at this point is at a relatively low level. The standard deviation reflects the degree of pressure fluctuation. Therefore, μ - 3σ represents a relatively low pressure value area. If the point pressure P i < μ - 3σ, it indicates that the pressure at this point is significantly lower than the average pressure, indicating that there is almost no pressure in some areas of the mattress, or the force on the mattress is uneven. When the above two conditions are met, it is judged that the state of the mattress is in the first state, and the first state is that the mattress is in a bent state.

[0039] If the wavelength shift data is less than the preset wavelength shift threshold, it indicates that the deformation or bending of the mattress is small and it is in a flat state; if the point pressure P i ≥ μ - 3σ, it indicates that the pressure in some areas of the mattress is relatively uniform and there will be no excessively low pressure values. Opposite to the low pressure area in the first state, this situation indicates that the pressure distribution of the mattress is relatively stable and uniform. When any of the above conditions is met, it is judged that the state of the mattress is in the second state, and the second state indicates that the mattress is in a flat state with a relatively uniform pressure distribution.

[0040] Furthermore, the functional relationship includes:

[0041]

[0042] Among them, Δλ is the wavelength shift data, n en is the effective refractive index, ∈ is the strain data, v is the Poisson's ratio, and λ is the input optical fiber deformation data.

[0043] Specifically, the functional relationship includes:

[0044]

[0045] n e n is the effective refractive index, which is a parameter reflecting the propagation speed of light in the optical fiber. The effective refractive index determines the propagation characteristics of light waves in the optical fiber. ∈ is the strain data, representing the deformation amount generated on the surface of the mattress due to compression or bending. The degree of compression or deformation of the mattress will affect this value. v is the Poisson's ratio, which describes the deformation degree in the vertical direction when the material is compressed in one direction. For the mattress, the Poisson's ratio is related to the elastic characteristics of the mattress material.

[0046] This functional relationship illustrates the relationship between the wavelength shift of the fiber grating array and the strain on the surface of the mattress. Specifically, the wavelength shift is proportional to the strain and is adjusted by the refractive index and the Poisson's ratio. When the mattress is compressed, the deformation on the surface will cause the wavelength of the light wave reflected by the fiber grating to shift. This shift can be used as an indication of the surface pressure and deformation of the mattress. The effective refractive index and the Poisson's ratio adjust the conversion relationship between the strain and the wavelength shift. The effective refractive index affects the propagation speed of light waves, and the Poisson's ratio affects the deformation degree of the mattress in different directions. These two factors comprehensively affect the degree of wavelength shift. The strain data is directly related to the change in the surface pressure of the mattress. Therefore, through this functional relationship, the wavelength shift data can be used to reflect the deformation degree of the mattress surface. For example, when the mattress is subjected to the weight or pressure of the user, some areas of the mattress will deform. By monitoring the wavelength shift of the fiber grating array, the magnitude of these deformations can be deduced, thus helping to achieve precise adjustment of the mattress state.

[0047] Furthermore, the method further includes:

[0048] If the massage mattress is in the second state, connect the M airbag control modules of the massage mattress, and identify the M initial airbag control parameter sets corresponding to the M airbag control modules through PID steady-state regulation, where M is the total number of airbag control modules set in the massage mattress; control the airbag pressure of the massage mattress according to the M initial airbag control parameter sets.

[0049] Judge the state of the massage mattress according to the sensor data. If the mattress is not in the first state, i.e., the bent state, then mark that the mattress is in the second state, i.e., the flat state. In this case, connect the M airbag control modules of the massage mattress, where M represents the total number of airbag control modules set on the massage mattress. In the second state, all the airbag control modules in the entire mattress area need to participate in the adjustment because the mattress may not have significant bending, and the airbag pressure adjustment needs to evenly cover the entire mattress. Connect all these airbag control modules, which are distributed in different areas of the mattress, such as the head, back, legs, etc. Each airbag control module is responsible for adjusting the airbag pressure in a certain part of the area.

[0050] PID (Proportional-Integral-Derivative) control is a common feedback control method used to maintain the stability and accuracy of a system. Here, PID steady-state regulation is used to adjust the pressure of each airbag control module. The PID control algorithm will target each airbag control module and gradually adjust to the target pressure by adjusting the inflation volume of the airbag until the steady state is finally reached. Identify the initial airbag control parameter set of each airbag control module through PID steady-state regulation. These initial parameter sets are set according to the current state (second state) of the mattress. The initial parameter sets include parameters such as the target pressure of the airbag, inflation and deflation rates.

[0051] Control the airbag pressure in the entire mattress area according to the identified M initial airbag control parameter sets. The purpose of this process is to ensure that the pressure distribution of the mattress is uniform to adapt to the second state of the mattress. Specifically, according to the initial airbag control parameter set of each airbag control module, accurately adjust the inflation volume of the airbag. If the airbag pressure in some areas is too high or too low, it will be automatically adjusted to reach the target pressure. Since the mattress is in the second state, the airbag pressure adjustment is mainly to ensure that the mattress still provides comfortable support in the non-bent state.

[0052] Furthermore, adopt the PPO algorithm to train the reward function based on the state space and the action space to obtain a trained control enhancement model. The method includes:

[0053] Construct a reward function, which is a function constructed by weighting the pressure fluctuation amplitude, wavelength offset stability, and control action energy consumption, and is used to measure the degree of balance of human contact pressure; collect the state space simulation data of the flexible strain gauge module in multiple simulated bent states, and define adjustable action space simulation data in the action space; call the reward function to repeatedly iterate and train the PPO control network with the state space simulation data and the action space simulation data until the reward of the PPO control network converges to obtain a trained control enhancement model.

[0054] The objective of the reward function is to measure the balance degree of the human body contact pressure. It is constructed by weighting the pressure fluctuation amplitude, wavelength shift stability, and control action energy consumption. Among them, the pressure fluctuation amplitude reflects the pressure change range at the human body contact points of the mattress. Ideally, the pressure fluctuation amplitude should be as small as possible to ensure the comfort of the human body during the massage process. A smaller fluctuation represents a more uniform mattress pressure; the wavelength shift stability detects the change of the wavelength through the fiber Bragg grating array and the piezoelectric film network, and judges whether the pressure distribution is uniform according to this change. The stability of the wavelength shift means that the deformation degree of the massage mattress is well controlled, which helps to improve the comfort. A more stable wavelength shift indicates a more balanced state of the mattress; the control action energy consumption reflects the efficiency of adjusting the airbag pressure. A lower energy consumption means that the system can complete the control task without wasting energy and achieve the required pressure balance target. Among them, the weight factor is used to balance the importance of these three factors and can be adjusted according to actual needs.

[0055] Under multiple simulated bending states, a flexible sensing strain gauge module is used to collect state space simulation data. The data in the state space includes the pressure point distribution on the surface of the massage mattress and the wavelength shift data. Among them, through the fiber Bragg grating array and the piezoelectric film network on the flexible sensing strain gauge module, the pressure distribution data in the human body contact area of the massage mattress is obtained. These data can help identify the pressure fluctuation of the mattress and the pressure change in the local area; using the change of the fiber deformation in the fiber Bragg grating array, the wavelength shift data is obtained through the Bragg grating principle. These data reflect the degree of bending deformation of the mattress surface and help to analyze the stability of the mattress. By simulating multiple bending states, the sensor data can reflect the performance of the mattress under different human body contact positions and states.

[0056] Adjustable action space simulation data is defined in the action space. Specifically, the action space includes the airbag pressure adjustment step size and the airbag pressure adjustment amplitude. Among them, the airbag pressure adjustment step size is the change amount each time the airbag pressure is adjusted, and its range can be determined through experiments or simulations; the airbag pressure adjustment amplitude is the maximum change amount each time the airbag pressure is adjusted, and the optimal pressure adjustment scheme can be selected within this range. These action data are used for simulation adjustment during the training process, that is, to achieve the control of pressure balance through different adjustment step sizes and amplitudes.

[0057] Apply the above state space simulation data and action space simulation data to the PPO control network, that is, the reinforcement learning algorithm, to construct a training model. The PPO algorithm will iteratively train the model according to the reward function so that the model can select appropriate airbag control actions according to the given state (such as pressure fluctuation amplitude, wavelength shift stability, etc.), and finally achieve pressure balance.

[0058] During the training process, the PPO control network is iteratively trained using state space simulation data and action space simulation data until the value of the reward function converges. Specifically, the state space simulation data and action space simulation data collected by the flexible strain gauge module are input into the PPO control network. In each training session, based on the current state space data, the PPO control network selects a corresponding action, namely the pressure adjustment step size and amplitude for airbag control. These actions may not be optimal, but the PPO algorithm will attempt to find the optimal solution from a large number of different actions. According to the execution effect of the current action, a reward value is calculated using the reward function. The higher the reward value, the more appropriate the selected action is, and a lower reward value indicates a poorer current strategy. During the training process, the PPO algorithm will perform multiple iterations. In each round of iteration, the PPO control network selects actions according to the current strategy and adjusts the strategy based on the feedback of the reward function. Each training is optimized based on simulation data. When the reward value in the PPO algorithm stabilizes at a relatively high level, the training stops, and a trained control enhancement model is obtained. This model will be able to automatically adjust the airbag pressure according to the real-time mattress state in actual applications to ensure the comfort and pressure balance of the massage mattress.

[0059] Furthermore, an enhanced airbag control parameter set is output, and the method includes:

[0060] Collect the real-time state data of the flexible strain gauge module; input the first set of initial airbag control parameter sets into the control enhancement model. The control enhancement model determines the equilibrium target by calling the reward function according to the real-time state data, and performs pressure equilibrium adjustment on the first set of initial airbag control parameter sets in the action space with a preset step size, and outputs the real-time action data that meets the equilibrium target, including the pressure adjustment step size and amplitude for airbag control; output the pressure adjustment step size and amplitude for airbag control as the enhanced airbag control parameter set.

[0061] Collect the real-time state data of the current massage mattress through the flexible strain gauge module, including pressure distribution and wavelength shift data, etc. The collected data needs to be real-time so that the control system can make the most accurate response at each moment to adjust the pressure of the mattress.

[0062] Input the first set of initial airbag control parameter sets into the control enhancement model. These initial parameter sets are the initial airbag pressure configurations. According to the real-time state data, the control enhancement model calls the reward function. The role of the reward function is to evaluate the pressure distribution in the current state and output the reward value regarding pressure equilibrium. This reward value is the target that the model hopes to maximize during the learning process. According to the output of the reward function, the model determines the equilibrium target, that is, the adjusted ideal airbag pressure value, so as to make the pressure distribution of the mattress balanced and avoid any local pressure being too high or too low.

[0063] In the action space, each action of adjusting the airbag pressure has a preset step size. The preset step size refers to the amount by which the airbag pressure increases or decreases each time, and it needs to be reasonably set within the control range. The action space includes all possible airbag control actions, such as the pressure adjustment range of each airbag. The model selects appropriate pressure adjustment step sizes and amplitudes in the action space to achieve the target pressure balance. The first set of initial airbag control parameter sets is adjusted within the action space using the preset step size, so that the pressure distribution of the mattress gradually approaches the balance target. The execution of each action will cause a slight adjustment of the airbag pressure, aiming to reduce pressure fluctuations and improve the stability of wavelength shift.

[0064] Through adjustment, the action real-time data that meets the balance target is finally output, including the pressure adjustment step size and amplitude for airbag control. Among them, the pressure adjustment step size is the amount of each airbag pressure adjustment. For example, the step size indicates that the change value when adjusting the airbag pressure each time is +0.5 kPa or -0.5 kPa; the amplitude is the maximum range or amplitude of the airbag pressure adjustment, representing the maximum range within which the airbag pressure can be adjusted. For example, the amplitude is the pressure range from 0 to 100 kPa.

[0065] According to the output of the control enhancement model, including the pressure adjustment step size and amplitude, as the enhanced airbag control parameter set, these parameters will be used to adjust the airbag pressure to ensure that the pressure distribution of the mattress meets the comfort requirements of the human body.

[0066] Furthermore, if the massage mattress is in the first state, the method further includes:

[0067] Determine the non-bent mattress area of the massage mattress in the first state, connect the second set of airbag control modules corresponding to the non-bent mattress area, and identify the second set of initial airbag control parameter sets of the second set of airbag control modules; input the second set of initial airbag control parameter sets into the control weakening model for parameter weakening adjustment, output the weakened airbag control parameter set, and control the airbag pressure of the non-bent mattress area according to the weakened airbag control parameter set.

[0068] When the mattress is in the first state, there are gaps in the bent part, but there are no gaps in other parts. These other parts are called non-bent areas, and it is possible to determine which areas belong to the non-bent areas by real-time monitoring of the deformation and pressure on the mattress surface.

[0069] The non-bending region corresponds to the second set of airbag control modules, which are used to adjust the airbag pressure in the non-bending region. Similar to the airbag control modules in the bending region, the second set of airbag control modules also has an initial set of second initial airbag control parameter sets. These parameter sets include the target value of the airbag pressure, the adjustment rate, the adjustment step size, etc. These parameter sets represent the preliminary pressure adjustment settings required for the mattress in the non-bending state.

[0070] The control weakening model is also a model used to adjust the airbag pressure. Its goal is to make the airbag control more precise by weakening the airbag pressure to adapt to the non-bending state of the mattress. Different from the control enhancement model that needs to increase the airbag pressure in the bending state, the control weakening model is used to reduce the airbag pressure when the mattress remains flat. The second initial airbag control parameter set is input into the control weakening model, and the control weakening model fine-tunes the airbag pressure based on the initial airbag control parameter set to reduce unnecessary pressure fluctuations or excessive pressure values. Specifically, the change range of the airbag pressure is reduced according to the state of the mattress, the needs of the user, and other environmental factors to ensure that the mattress still provides comfortable support in the non-bending state. After the weakening adjustment, the control weakening model outputs an optimized weakened airbag control parameter set, which is adjusted to be more suitable for the airbag pressure adjustment in the non-bending state.

[0071] Use the weakened airbag control parameter set to adjust the airbag pressure in the non-bending mattress region to ensure that the airbag pressure meets the expected comfort and stability requirements. This process will maintain a suitable pressure distribution on the mattress surface in the non-bending state, thus providing a more comfortable use experience for the user.

[0072] In summary, the pressure adjustment method of the massage mattress based on airbag control provided by the embodiments of the present application has the following technical effects:

[0073] Using a flexible sensor strain gauge module to detect the state of a massage mattress can sense the deformation and pressure changes of the mattress in real time. Through this process, it can accurately determine whether the mattress is in a bent state currently, providing data support for subsequent airbag control. When the massage mattress is in a bent state, it can accurately identify and determine the bent mattress area. By effectively partitioning the bent area of the mattress, targeted airbag control can be carried out, making the airbag adjustment more precise and efficient. The accurate identification of the bent area helps to avoid unnecessary airbag intervention in the subsequent control process, thus achieving local optimization adjustment. When the mattress is in a bent state, there will be a gap between the human body and the bent part of the mattress, resulting in a decrease in the pressure felt by the human body. By using a control enhancement model in the bent area to perform airbag pressure enhancement adjustment, these gap areas can be effectively compensated, enabling the mattress to still provide appropriate support and comfort when bent, enhancing the output of the airbag control parameter set, so that the airbag can accurately control the pressure in this area, thereby optimizing the user's massage experience and ensuring the uniformity and comfort of the massage effect.

[0074] Embodiment 2, based on the same inventive concept as the method for adjusting the pressure of a massage mattress based on airbag control in the foregoing embodiment, the embodiment of the present application provides a system for adjusting the pressure of a massage mattress based on airbag control. The system includes:

[0075] A massage mattress state detection module for using a flexible sensor strain gauge module to detect the state of a massage mattress, and the massage mattress is a bendable massage mattress.

[0076] A bent mattress area determination module for determining the bent mattress area if the massage mattress is in a first state, where the first state is a bent state.

[0077] A control enhancement model training module for defining a state space and an action space, and training a reward function based on the state space and the action space using the PPO algorithm to obtain a trained control enhancement model, where the state space includes pressure point distribution and wavelength offset data, and the action space includes airbag pressure adjustment step size and amplitude.

[0078] An airbag pressure control module for connecting to the first group of airbag control modules corresponding to the bent mattress area, identifying the first group of initial airbag control parameter sets of the first group of airbag control modules, inputting the first group of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting an enhanced airbag control parameter set, and controlling the airbag pressure in the bent mattress area according to the enhanced airbag control parameter set.

[0079] Furthermore, the flexible strain sensing patch module includes a fiber Bragg grating array and a piezoelectric film network; wherein, the fiber Bragg grating array is longitudinally arranged on the surface of the massage mattress at a preset interval, and the piezoelectric film network is arranged on the surface of the massage mattress in a 5 cm × 5 cm grid distribution.

[0080] Furthermore, the massage mattress state detection module includes the following operation steps:

[0081] According to the fiber Bragg grating array and the piezoelectric film network of the flexible strain sensing patch module, fiber deformation data and piezoelectric pressure data are respectively obtained; a functional relationship between the wavelength shift of the Bragg grating and the deformation is established, and the fiber deformation data is signal demodulated according to the functional relationship to obtain wavelength shift data Δλ; the regional pressure mean μ and the regional pressure standard deviation σ are calculated according to the piezoelectric pressure data; the state of the massage mattress is judged according to the wavelength shift data Δλ, the regional pressure mean μ and the regional pressure standard deviation σ.

[0082] Furthermore, the massage mattress state detection module includes the following operation steps:

[0083] If the wavelength shift data Δλ is greater than or equal to a preset wavelength shift threshold, and the point pressure P i < μ - 3σ, it is judged that the state of the massage mattress is in the first state, and the first state is the bent state; if the wavelength shift data Δλ is less than the preset wavelength shift threshold or the point pressure P i ≥ μ - 3σ, it is judged that the state of the massage mattress is in the second state, and the second state is the non-bent state.

[0084] Furthermore, the functional relationship includes:

[0085]

[0086] wherein, Δλ is the wavelength shift data, n e is the effective refractive index, v is the Poisson's ratio, and λ is the input fiber deformation data.

[0087] Furthermore, the massage mattress state detection module includes the following operation steps:

[0088] If the massage mattress is in the second state, connect the M airbag control modules of the massage mattress, and identify the M initial airbag control parameter sets corresponding to the M airbag control modules through PID steady-state regulation, where M is the total number of airbag control modules set in the massage mattress; control the airbag pressure of the massage mattress according to the M initial airbag control parameter sets.

[0089] Furthermore, the control enhancement model training module includes the following operation steps:

[0090] Construct a reward function, which is a function constructed by weighting the pressure fluctuation amplitude, wavelength offset stability, and control action energy consumption, and is used to measure the balance degree of human contact pressure; under multiple simulated bending states, collect the state space simulation data of the flexible strain gauge module, and define adjustable action space simulation data in the action space; call the reward function to perform iterative training on the PPO control network with the state space simulation data and the action space simulation data until the reward of the PPO control network converges, and obtain a trained control enhancement model.

[0091] Furthermore, the airbag pressure control module includes the following operation steps:

[0092] Collect the real-time state data of the flexible strain gauge module; input the first set of initial airbag control parameter sets into the control enhancement model, and the control enhancement model calls the reward function according to the real-time state data to determine the balance target, and perform pressure balance adjustment on the first set of initial airbag control parameter sets in the action space with a preset step size, and output the real-time action data that meets the balance target, including the pressure adjustment step size and amplitude for airbag control; output the pressure adjustment step size and amplitude for airbag control as the enhanced airbag control parameter set.

[0093] Furthermore, the bending mattress area determination module includes the following operation steps:

[0094] Determine the non-bending mattress area of the massage mattress in the first state, connect the second set of airbag control modules corresponding to the non-bending mattress area, and identify the second set of initial airbag control parameter sets of the second set of airbag control modules; input the second set of initial airbag control parameter sets into the control weakening model for parameter weakening adjustment, output the weakened airbag control parameter set, and control the airbag pressure of the non-bending mattress area according to the weakened airbag control parameter set.

[0095] Through the foregoing detailed description of the method for adjusting the pressure of the massage mattress based on airbag control in this specification, those skilled in the art can clearly know the pressure adjustment system of the massage mattress based on airbag control in this embodiment. Since it corresponds to the method disclosed in the embodiment, it is described relatively simply, and for the related parts, refer to the description in the method part.

[0096] The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A pressure adjustment method for a massage mattress based on airbag control, characterized in that The method includes: Using a flexible sensor strain gauge module to detect the state of a massage mattress, where the massage mattress is a bendable massage mattress; If the massage mattress is in a first state, determining the bent mattress area, where the first state is a bent state; Defining a state space and an action space, and training a reward function based on the state space and the action space using the PPO algorithm to obtain a trained control enhancement model, where the state space includes pressure point distribution and wavelength shift data, and the action space includes airbag pressure adjustment step size and amplitude; Connecting the first group of airbag control modules corresponding to the bent mattress area, identifying the first group of initial airbag control parameter sets of the first group of airbag control modules, inputting the first group of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting an enhanced airbag control parameter set, and controlling the airbag pressure of the bent mattress area according to the enhanced airbag control parameter set.

2. The method according to claim 1, wherein The flexible sensor strain gauge module includes an optical fiber grating array and a piezoelectric thin film network; Among them, the optical fiber grating array is longitudinally arranged on the surface of the massage mattress at a preset interval, and the piezoelectric thin film network is arranged on the surface of the massage mattress in a 5 cm × 5 cm grid distribution.

3. The method according to claim 2, wherein Using a flexible sensor strain gauge module to detect the state of a massage mattress, the method includes: According to the optical fiber grating array and the piezoelectric thin film network of the flexible sensor strain gauge module, obtaining optical fiber deformation data and piezoelectric pressure data respectively; Establishing a functional relationship between the wavelength shift of the Bragg grating and the deformation, and demodulating the optical fiber deformation data according to the functional relationship to obtain wavelength shift data Δλ; Calculating the regional pressure mean μ and the regional pressure standard deviation σ according to the piezoelectric pressure data; Judging the state of the massage mattress according to the wavelength shift data Δλ, the regional pressure mean μ and the regional pressure standard deviation σ.

4. The method according to claim 3, wherein Judging the state of the massage mattress according to the wavelength shift data Δλ, the regional pressure mean μ and the regional pressure standard deviation σ, the method includes: If the wavelength offset data Δλ is greater than or equal to a preset wavelength offset threshold, and the point pressure P i < μ - 3σ, it is determined that the state of the massage mattress is in a first state, and the first state is a bent state; If the wavelength shift data Δλ is less than the preset wavelength shift threshold or the point pressure P i ≥ μ - 3σ, it is determined that the state of the massage mattress is in the second state, and the second state is a non-bending state.

5. The method according to claim 3, characterized in that, s where Δλ is the wavelength offset data, n e is the effective refractive index, v is the Poisson's ratio, and λ is the input fiber deformation data.

6. The method according to claim 1, wherein Demodulating the optical fiber deformation data according to the functional relationship, and the functional relationship includes: Using a flexible sensor strain gauge module to detect the state of a massage mattress, the method further includes: If the massage mattress is in a second state, connecting the M airbag control modules of the massage mattress, and identifying the M initial airbag control parameter sets corresponding to the M airbag control modules through PID steady-state regulation, where M is the total number of airbag control modules set in the massage mattress; 7. The method according to claim 1, characterized in that, Controlling the airbag pressure of the massage mattress according to the M initial airbag control parameter sets. Training a reward function based on the state space and the action space using the PPO algorithm to obtain a trained control enhancement model, the method includes: Constructing a reward function, which is a function weighted by pressure fluctuation amplitude, wavelength shift stability, and control action energy consumption, and is used to measure the balance degree of human contact pressure; Under multiple simulated bending states, collecting the state space simulation data of the flexible sensor strain gauge module, and defining adjustable action space simulation data in the action space; The trained control enhancement model is obtained by repeatedly iteratively training the PPO control network with the state space simulation data and the action space simulation data by invoking the reward function until the reward of the PPO control network converges.

8. The method according to claim 7, characterized in that Output an enhanced airbag control parameter set, the method comprising: Collect the real-time state data of the flexible strain gauge module; Input the first set of initial airbag control parameter sets into the control enhancement model. The control enhancement model determines an equilibrium target by invoking the reward function according to the real-time state data, and performs pressure equilibrium adjustment on the first set of initial airbag control parameter sets in the action space with a preset step size, and outputs real-time action data that meets the equilibrium target, including the pressure adjustment step size and amplitude for airbag control; Output the pressure adjustment step size and amplitude for airbag control as the enhanced airbag control parameter set.

9. The method according to claim 1, wherein If the massage mattress is in the first state, the method further comprises: Determine the non-bent mattress area of the massage mattress in the first state, connect the second set of airbag control modules corresponding to the non-bent mattress area, and identify the second set of initial airbag control parameter sets of the second set of airbag control modules; Input the second set of initial airbag control parameter sets into the control weakening model for parameter weakening adjustment, output the weakened airbag control parameter set, and control the airbag pressure of the non-bent mattress area according to the weakened airbag control parameter set.

10. A pressure adjustment system for a massage mattress based on airbag control, characterized in that, For implementing the airbag control-based massage mattress pressure adjustment method according to any one of claims 1-9, the system comprises: A massage mattress state detection module for detecting the state of the massage mattress using a flexible strain gauge module, the massage mattress being a bendable massage mattress; A bent mattress area determination module for determining the bent mattress area if the massage mattress is in the first state, wherein the first state is a bent state; A control enhancement model training module for defining a state space and an action space, training a reward function based on the state space and the action space using the PPO algorithm to obtain a trained control enhancement model, wherein the state space includes pressure point distribution and wavelength offset data, and the action space includes airbag pressure adjustment step size and amplitude; An airbag pressure control module for connecting the first set of airbag control modules corresponding to the bent mattress area, identifying the first set of initial airbag control parameter sets of the first set of airbag control modules, inputting the first set of initial airbag control parameter sets into the control enhancement model for parameter enhancement adjustment, outputting the enhanced airbag control parameter set, and controlling the airbag pressure of the bent mattress area according to the enhanced airbag control parameter set.

Images