Method for regulating pillow surface temperature

Abstract

The application provides a method for adjusting the surface temperature of a pillow, comprising: determining initial parameters of a temperature hot spot arrangement scheme according to the material and size of the pillow, including hot spot spacing and hot spot quantity, simulating the temperature hot spot arrangement scheme to obtain a temperature distribution cloud diagram of the pillow surface; optimizing the hot spot arrangement scheme under the premise of meeting the temperature distribution uniformity index, reducing the number of hot spots while ensuring the temperature adjustment effect; setting temperature sensors and heating elements at corresponding positions on the pillow surface according to the optimized hot spot arrangement scheme, the temperature sensors being used to collect temperature data at the hot spots in real time, and the heating elements being adjusted in power according to the temperature data; establishing a mathematical model between the power of the heating elements and the temperature of the hot spots, calculating and obtaining the control method of each heating element according to different target temperatures; and analyzing the curved surface of the pillow surface by using an ergonomics simulation software to finally adjust the hot spot arrangement scheme.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a method for regulating the surface temperature of a pillow. Background Art

[0002] When designing a temperature-adjustable pillow, multiple hotspots need to be arranged on the pillow surface, each equipped with a temperature sensor and heating element. However, determining the spacing between hotspots to ensure effective temperature regulation while minimizing the impact of excessively dense hotspots on wearing comfort is a question worthy of further exploration. Excessively large hotspot spacing can lead to significant temperature differences between adjacent hotspots, resulting in an uneven temperature distribution across the pillow surface and a negative user experience. However, too small a spacing between hotspots can lead to an excessive number of hotspots, increasing costs and potentially compromising temperature regulation accuracy due to interference between hotspots. Furthermore, overly dense hotspots can create an uneven surface, impacting wearing comfort. Therefore, balancing the conflict between hotspot density and temperature regulation effectiveness—finding an optimal hotspot placement scheme that balances temperature regulation performance and user comfort—is a pressing technical challenge. Summary of the Invention

[0003] The present invention provides a method for regulating the surface temperature of a pillow, which mainly comprises:

[0004] Based on the pillow material and size, the initial parameters of the temperature hotspot layout scheme are determined, including the hotspot spacing and the number of hotspots. The temperature hotspot layout scheme is simulated to obtain a temperature distribution cloud map on the pillow surface.

[0005] Analyze the temperature distribution cloud map and extract the temperature distribution uniformity index. If the uniformity index does not reach the preset threshold, adjust the parameters of the hotspot layout plan and repeat the simulation until the temperature distribution uniformity requirements are met;

[0006] Under the premise of meeting the temperature distribution uniformity index, the hot spot layout plan is optimized to reduce the number of hot spots while ensuring the temperature regulation effect;

[0007] According to the optimized hotspot layout plan, temperature sensors and heating elements are set at corresponding positions on the pillow surface. The temperature sensors are used to collect temperature data at the hotspots in real time, and the heating elements adjust the power according to the temperature data.

[0008] Establish a mathematical model between the power of the heating element and the hotspot temperature, and calculate and obtain the control method of each heating element according to different target temperatures;

[0009] Ergonomic simulation software was used to analyze the curved surface of the pillow and make final adjustments to the hotspot placement plan;

[0010] According to the final adjusted hotspot arrangement plan, a control circuit and program for a temperature-adjustable pillow are established at the bottom of the pillow to process the data collected by the temperature sensor and the control strategy information output by the heating element, control the on and off of each heating element, and provide users with temperature setting and mode selection functions.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The present invention discloses a method for adjusting the surface temperature of a pillow. By analyzing the material and size of the pillow, the initial layout of the temperature hot spots is determined, and the temperature distribution is optimized using simulation and image processing technology to achieve a uniform distribution of the surface temperature. On this basis, the present invention uses ergonomic simulation to make a final adjustment to the hot spot layout, and sets temperature sensors and heating elements at corresponding positions. By establishing a mathematical model, the system can automatically calculate the control strategy of the heating element according to the target temperature. The present invention also includes a bottom control circuit and a human-computer interaction interface, which realizes real-time monitoring, intelligent adjustment and user-defined settings of the temperature. This method can not only provide a comfortable and uniform temperature experience, but also can be precisely adjusted according to personal needs. It can also achieve efficient use of energy while ensuring comfort, thereby improving the user's sleep quality and experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flow chart of a method for adjusting the surface temperature of a pillow.

[0014] Figure 2 This is a schematic diagram of a method for regulating the surface temperature of a pillow according to the present invention.

[0015] Figure 3 This is another schematic diagram of a method for regulating the surface temperature of a pillow according to the present invention. DETAILED DESCRIPTION

[0016] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0017] like Figure 1-3 In this embodiment, a method for adjusting the surface temperature of a pillow may specifically include:

[0018] Step S101: Determine the initial parameters of the temperature hotspot arrangement scheme based on the pillow material and size, including the hotspot spacing and the number of hotspots, simulate the temperature hotspot arrangement scheme, and obtain a temperature distribution cloud map on the pillow surface.

[0019] Obtain pillow material parameters and size data, select pillow material from the preset material list, and obtain corresponding thermal conductivity and specific heat capacity parameters based on the selected material. Perform finite element analysis meshing on the three-dimensional digital model, set boundary conditions and initial temperature distribution. Use genetic algorithm to optimize the preset hotspot layout scheme, with hotspot spacing and number as decision variables and temperature uniformity as the objective function. If the fitness function is the inverse of the pillow surface temperature variance, then add constraints on hotspot power and total energy consumption. Obtain the optimal hotspot layout parameters through multiple iterative calculations, including hotspot position coordinates and power values. Perform transient heat conduction simulation based on the optimal hotspot layout parameters, and use a two-dimensional heat conduction partial differential equation to describe the temperature field change. The two-dimensional heat conduction partial differential equation is:

[0020]

[0021] Where u represents temperature, t represents time, x and y represent spatial coordinates, and α represents the thermal diffusion coefficient. Discretize the temperature field equation and construct a differential format to obtain the time-varying temperature field. Record the temperature value of each grid node at different times to form a temperature-time data matrix. Perform bilinear interpolation on the simulation result data to generate a continuous temperature distribution function. Map the temperature values ​​to the corresponding color space. Plot a cloud map of the pillow surface temperature and generate a temperature scale, marking the temperature ranges corresponding to different colors.

[0022] Specifically, a three-dimensional digital model was constructed based on pillow material parameters and dimensions. A specific pillow material was selected from a preset list, and the corresponding thermal conductivity and specific heat capacity parameters were called. Finite element analysis was used to mesh the model, set boundary conditions, and establish an initial temperature distribution. A genetic algorithm was used to optimize the hotspot placement scheme, using the hotspot spacing and number as decision variables and temperature uniformity as the objective function. The fitness function was defined as the inverse of the pillow surface temperature variance, while incorporating constraints such as hotspot power and total energy consumption. The optimal hotspot placement parameters, including hotspot location coordinates and power values, were determined through multiple iterations. Based on the optimal hotspot placement parameters, a transient heat conduction simulation was performed, using a two-dimensional heat conduction partial differential equation to describe the temperature field. The equation was discretized using the finite difference method, a difference scheme was constructed, and the temperature field evolution over time was determined through time-stepping. The temperature values ​​of each mesh node at different times were recorded to form a temperature-time data matrix. Bilinear interpolation was performed on the simulation data to generate a continuous temperature distribution function. A rainbow color scheme was used to map the temperature values ​​to a corresponding color space. Draw a temperature cloud map on the pillow surface to enhance the visualization of temperature changes. Generate a temperature scale and mark the temperature range corresponding to different colors to facilitate intuitive understanding of the temperature distribution. When constructing a three-dimensional digital model, select memory foam from the preset material list, whose thermal conductivity is 0.04W / (m·K) and specific heat capacity is 1800J / (kg·K). The pillow size is set to 60cm×40cm×15cm, and tetrahedral mesh is used with a unit size of 0.5cm. The boundary condition is set to natural convection, and the initial temperature is 25℃. The genetic algorithm optimizes the hotspot layout scheme, with the population size set to 100 and the number of iterations to 500. The decision variables include the x, y coordinates and power values ​​of the 10 hotspots, with value ranges of [0,60], [0,40], and [0,60], [0,40], and [0,60], [0,40], and [0,60], [0,40], and [0,40], respectively.

[0023] [0.5,2]W. The fitness function is defined as the inverse of the pillow surface temperature variance, and the constraint is that the total power does not exceed 15W. After iterative optimization, the optimal hotspot arrangement parameters are obtained as follows:

[0024] [(10,5,1.2),(20,15,1.5),(30,25,1.8),(40,35,1.3),(50,10,1.1),(15,30,1.4),(25,20,1.6),(35,5,1.2),(45,25,1.7),(55,35,1.3)], the coordinate unit is cm, and the power unit is W. The transient heat conduction simulation uses the two-dimensional heat conduction equation as follows: Where α is the thermal diffusivity, T(x,y,t) is the temperature field, representing the temperature at any given time t and any location (x,y); α is the thermal diffusivity of the material, defined as α = k / ρ*cp, where k is the thermal conductivity, ρ is the density, and cp is the specific heat capacity. The solution is solved using the explicit finite difference method with a time step of 0.1s and a spatial step of 0.5cm. The total simulation time is 1800s, and the temperature field data is recorded every 10s. The simulation results are bilinearly interpolated with an interpolation grid size of 0.1cm×0.1cm. A rainbow color spectrum is used for color mapping, with blue representing the lowest temperature of 25°C and red representing the highest temperature of 40°C. A temperature cloud map is drawn and a temperature scale is generated, marking the correspondence between color and temperature with an interval of 1°C.

[0025] Step S102 : Analyze the temperature distribution cloud map and extract the temperature distribution uniformity index. If the uniformity index does not reach the preset threshold, adjust the parameters of the hotspot arrangement scheme and repeat the simulation until the temperature distribution uniformity requirement is met.

[0026] Acquire a temperature distribution cloud map, and filter the temperature distribution cloud map using a Gaussian filter with a preset kernel size and standard deviation to obtain a filtered temperature distribution image; segment the filtered temperature distribution image into multiple temperature regions based on the filtered temperature distribution image; calculate the area and average temperature value of each temperature region for the multiple temperature regions; calculate the temperature distribution uniformity index using a weighted standard deviation method based on the area and average temperature value of each temperature region; if the temperature distribution uniformity index is greater than a preset threshold, adjust the hotspot arrangement scheme parameters; update the heat source boundary conditions based on the adjusted hotspot arrangement scheme parameters, solve the heat conduction equation, and obtain a new temperature distribution cloud map; repeat the filtering process, temperature region segmentation, uniformity index calculation, parameter adjustment, and heat conduction equation solution until the temperature distribution uniformity index is less than or equal to the preset threshold; output a hotspot arrangement scheme that meets the preset threshold conditions, wherein the hotspot arrangement scheme includes heat source coordinates and corresponding power values.

[0027] Specifically, the temperature distribution cloud map undergoes image preprocessing. A Gaussian filter with a kernel size of 5x5 and a standard deviation of 1.0 is used to eliminate noise. The image is then segmented into different temperature regions using the Otsu adaptive threshold segmentation algorithm, and the area and average temperature of each region are extracted. The temperature distribution uniformity index is calculated using the weighted standard deviation method, with the weights being the proportion of each region's area to the total area. The weighted temperature standard deviation is then used as the uniformity index. If the uniformity index exceeds a preset threshold, the hotspot placement scheme parameters are adjusted using a gradient descent algorithm. The hotspot location and power are optimized using the uniformity index as the objective function. The initial learning rate is set to 0.01, and the maximum number of iterations is set to 1000. Convergence is determined when the rate of change of the objective function is less than 0.001, and a new hotspot placement scheme is generated, including updated heat source coordinates and power values. A new simulation is then performed based on the adjusted hotspot placement scheme, with updated heat source boundary conditions, new heat source locations and powers, and the heat conduction equation solved to generate a new temperature distribution cloud map. The process of image preprocessing, uniformity calculation, parameter adjustment, and simulation was repeated until the uniformity index was less than or equal to the preset threshold. The final hotspot placement plan, including the heat source coordinates and corresponding power values, was then output. The temperature distribution cloud map was preprocessed using a Gaussian filter with a kernel size of 5x5 and a standard deviation of 1.0 to remove noise, resulting in a filtered image. The Otsu adaptive threshold segmentation algorithm was applied to segment the image into 10 temperature zones, each with a temperature range of 3°C. The area pixel count and average temperature value of each zone were extracted. For example, zone 1 had an area of ​​12,500 pixels and an average temperature of 26.5°C; zone 2 had an area of ​​18,000 pixels and an average temperature of 29.3°C. The temperature distribution uniformity index was calculated using the weighted standard deviation method, with the weight being the proportion of each zone's area to the total area. Assuming a total area of ​​100,000 pixels, zone 1 had a weight of 0.125 and zone 2 had a weight of 0.18. The calculated weighted temperature standard deviation was 2.8°C, which was used as the uniformity index. If the uniformity index is greater than the preset threshold of 2.5°C, the gradient descent algorithm is used to adjust the hotspot layout plan. Set the initial learning rate to 0.01 and the maximum number of iterations to 1000. Optimize the positions and powers of the 10 hotspots, such as adjusting hotspot 1 from (10cm, 15cm, 1.5W) to (12cm, 16cm, 1.3W). When the rate of change of the objective function is less than 0.001, convergence is determined and a new hotspot layout plan is generated. Based on the adjusted plan, the simulation is repeated, the heat source boundary conditions are updated, the new heat source position and power are set, the heat conduction equation is solved for 60 seconds, and a new temperature distribution cloud map is generated. Repeat the above process until the uniformity index drops to 2.4°C, which is less than the preset threshold, and output the final hotspot layout plan, which contains the coordinates and power values ​​of the 10 hotspots.

[0028] Step S103, under the premise of meeting the temperature distribution uniformity index, the hotspot arrangement scheme is optimized to reduce the number of hotspots while ensuring the temperature regulation effect.

[0029] According to the existing hotspot arrangement scheme, a two-dimensional Gaussian diffusion model is used to calculate the temperature distribution around the hotspots to obtain the effective coverage radius of the hotspots, wherein the effective coverage radius is defined as the distance at which the temperature drops to 1 / e of the initial value. A clustering algorithm is used to group the hotspots, taking the hotspot coordinates as the feature, and hotspots with close distances and overlapping temperature influence ranges are merged, wherein the number of clusters is gradually reduced from the current number of hotspots. The clustering center is calculated as the new hotspot position, and the sum of the powers of all hotspots in the cluster is taken as the new hotspot power to obtain the updated hotspot arrangement scheme. For the updated hotspot arrangement scheme, a finite element method is used to solve the heat conduction equation, the grid size is set, the boundary condition is natural convection, and the time step is solved to obtain a new temperature distribution cloud chart. If the temperature distribution uniformity index meets the preset threshold, the current hotspot arrangement scheme is saved, the hotspot reduction operation is continued, and the clustering and updating operations are repeated until the temperature distribution uniformity index does not meet the requirements or the number of hotspots reaches the preset minimum value.

[0030] Specifically, based on the existing hotspot layout plan, the temperature influence range of each hotspot is calculated. The temperature distribution around the hotspot is calculated using a two-dimensional Gaussian diffusion model, and the effective coverage radius of each hotspot is obtained, defined as the distance at which the temperature drops to 1 / e of the initial value. The hotspots are grouped using the K-means clustering algorithm, characterized by the hotspot coordinates. The number of clusters is gradually reduced from the current number of hotspots, and hotspots with close proximity and overlapping temperature influence ranges are merged. The new position and power of the merged hotspot are calculated, and the cluster center is taken as the new hotspot position. The sum of the powers of all hotspots in the cluster is taken as the new hotspot power. The hotspot layout plan is then updated. The temperature distribution of the updated hotspot layout plan is simulated, and the heat conduction equation is solved using the finite element method. The grid size is set to 1 cm × 1 cm, the boundary condition is natural convection, and the solution time step is 0.1 s. A new temperature distribution cloud map is obtained, and the temperature distribution uniformity index is calculated. If the temperature distribution uniformity index meets the preset threshold, the current hotspot placement plan is saved and the hotspot reduction operation is continued. The clustering, updating, and simulation steps are repeated until the temperature distribution uniformity index falls below the threshold or the number of hotspots reaches the preset minimum. A lower limit of 50% of the initial number of hotspots is taken as the lower limit, and the final optimized hotspot placement plan is output. For the 10 hotspots in the initial hotspot placement plan, a two-dimensional Gaussian diffusion model is used to calculate the temperature influence range of each hotspot. Taking hotspot 1 as an example, with an initial temperature of 40°C and an ambient temperature of 25°C, its effective coverage radius is calculated to be 8 cm. The hotspots are grouped using the K-means clustering algorithm, with an initial number of 9 clusters. Clustering is performed based on the hotspot coordinates. The clustering results show that hotspots 2 and 5 can be merged. The new hotspot location is (15 cm, 20 cm). The power of the new hotspot is 3.5 W, which is the power of the original hotspot 2 (1.8 W) plus the power of hotspot 5 (1.7 W). The temperature distribution of the updated 9 hotspot layout schemes was simulated, and the heat conduction equation was solved using the finite element method. The grid size was set to 1cm×1cm, and the boundary condition was a natural convection coefficient of 5W / (m 2 ·K), with a solution time step of 0.1s and a total simulation time of 600s. After obtaining the new temperature distribution contour map, the calculated temperature uniformity index is 2.3°C, meeting the preset threshold of 2.5°C. The current layout plan of 9 hotspots is saved, and the hotspot reduction operation is continued. The above process is repeated until the number of hotspots is reduced to 5, which is 50% of the initial number of 10, or the temperature uniformity index exceeds 2.5°C. The final optimized hotspot layout plan contains 7 hotspots with positions and powers of (12cm, 18cm, 3.2W), (25cm, 30cm, 2.8W), (38cm, 15cm, 3.5W), (50cm, 25cm, 3.0W), (15cm, 40cm, 2.5W), (35cm, 45cm, 2.7W), and (55cm, 35cm, 2.3W), with a temperature uniformity index of 2.4°C.

[0031] Step S104: According to the optimized hotspot arrangement scheme, temperature sensors and heating elements are set at corresponding positions on the pillow surface. The temperature sensors are used to collect temperature data at the hotspots in real time, and the heating elements adjust power according to the temperature data.

[0032] Hot spot distribution information on the pillow surface is obtained, and the placement of the temperature sensor and heating element is determined based on this information. The temperature sensor is a thermocouple, and the heating element is a flexible thin-film heater. Temperature data from the temperature sensor is collected at a preset frequency. This temperature data is converted to a digital signal using an analog-to-digital converter and stored in a circular queue buffer. A proportional-integral-differential control method is used to calculate the output power of the heating element based on the deviation between the current temperature and the target temperature. The controller output of this proportional-integral-differential control method consists of a proportional term, an integral term, and a differential term. The PID parameters are determined using the Ziegler-Nichols tuning method. The proportional gain is gradually increased until the system experiences amplitude oscillation. The critical gain and oscillation period are recorded to calculate the PID parameters. The calculated output power is converted into a voltage signal. The power output of the heating element is controlled using pulse width modulation, with the modulation frequency and resolution set to preset values.

[0033] Specifically, according to the optimized hotspot placement scheme, temperature sensors and heating elements were placed at corresponding locations on the pillow surface, equal to the number of hotspots. The temperature sensors were thermocouples with a measurement range of 0-100°C and an accuracy of ±0.1°C. The heating elements were flexible thin-film heaters with a power range of 0-5W. Temperature data from the temperature sensors was collected at a 1Hz frequency, converted to digital signals via an analog-to-digital converter, and stored in a circular queue buffer with a capacity of 60 data points. A proportional-integral-derivative (PID) control algorithm was employed to calculate the output power of the heating element based on the deviation between the current and target temperatures. The PID parameters were determined using the Ziegler-Nichols tuning method. The specific process involved gradually increasing the proportional gain Kp until the system exhibited constant-amplitude oscillations. The critical gain Ku and oscillation period Tu were recorded, and the PID parameters were then calculated using empirical formulas. The PID controller output u(t) consists of the proportional term Kpe(t), the integral term Ki∫e(t)dt, and the differential term Kdde(t) / dt, where e(t) is the temperature deviation. The calculated output power is converted into a corresponding voltage signal, and pulse-width modulation (PWM) is used to control the power output of the heating element at a frequency of 1kHz and an 8-bit resolution to achieve precise temperature regulation. Seven pairs of temperature sensors and heating elements are placed at the seven optimized hotspot locations, such as a K-type thermocouple and a 5W flexible thin-film heater at (12cm, 18cm). Temperature data is acquired at a frequency of 1Hz, and a 12-bit analog-to-digital converter converts the analog signal into a digital signal with a resolution of 0.024°C. A circular queue buffer stores the last 60 seconds of temperature data. In the PID control algorithm, parameters are determined using the Ziegler-Nichols method. Kp is set starting from 0 and incremented by 0.1 until constant-amplitude oscillations occur around the target temperature of 35°C, at which point Ku = 1.5 and Tu = 10s. According to the Ziegler-Nichols empirical formula, Kp = Ku / 2 = 0.75, Ki = Ku / 1.5*Tu = 0.1, and Kd = Ku*0.125 = 0.1875. During the PID controller calculation process, if the current temperature is 33.5°C, the target temperature is 35°C, the temperature deviation e(t) = 1.5°C, the integral term ∫e(t)dt = 5°C·s, and the differential term de(t) / dt = -0.1°C / s, then the output u(t) = 1.35 + 0.9 - 0.1125 = 2.1375W. This 2.1375W is converted to a 2.677V voltage signal, assuming that 5W corresponds to 6.25V. The PWM control uses a 1kHz frequency, 8-bit resolution (256 levels), and a duty cycle of (2.677 / 6.25)*255≈109. That is, a PWM waveform with a duty cycle of 109 / 256 is output to control the power of the heating element and achieve precise temperature adjustment to 35°C.

[0034] Step S105, a mathematical model between the heating element power and the hotspot temperature is established, and the control method of each heating element is calculated and obtained according to different target temperatures.

[0035] Obtain historical data of heating element power and hotspot temperature, the historical data including temperature change process under different power inputs; according to the historical data, adopt least square method to fit temperature change curve, obtain mathematical model between heating element power and hotspot temperature; based on the mathematical model, use feedback controller, the feedback controller adopts model predictive control algorithm to solve optimal power sequence; input the difference between target temperature and current temperature into the feedback controller, calculate the power control sequence of each heating element; establish hotspot coupling coefficient matrix, the hotspot coupling coefficient matrix is determined according to the distance between hotspots and the thermal conductivity of material; according to the hotspot coupling coefficient matrix, adjust the power control sequence.

[0036] Specifically, historical data of heating element power and hot spot temperature are collected, and the temperature change process under different power inputs is recorded every second for 30 minutes, including initial temperature, target temperature, ambient temperature and heating time. The least squares method is used to fit the temperature change curve, and a mathematical model between the heating element power and the hot spot temperature is established, considering the factors of heat conduction and heat loss. The error function is defined as the sum of squares of the actual temperature and the predicted temperature, and by solving the parameter value that minimizes the error function, the function expression of temperature change with time T(t) = a(1-e^(-bt))+c is obtained, where a, b, c are undetermined parameters, and e is the base of natural logarithm. Parameter a represents the maximum amplitude of temperature change, parameter b determines the speed of temperature change, i.e. the speed of temperature change with time, and parameter c is the offset of the temperature change curve, which determines the starting temperature level of the temperature change curve. According to the established mathematical model, a feedback controller is used, and a model predictive control algorithm is adopted, with a 5-minute prediction horizon, and the target function is set as the weighted sum of the square sum of temperature error and power change. The optimal power sequence is solved by quadratic programming. For different target temperatures, the temperature difference is input into the feedback controller, and the power control sequence of each heating element is calculated. Considering the thermal influence between adjacent hot spots, a hot spot coupling coefficient matrix is established, and the coupling coefficient is calculated according to the distance between hot spots and the thermal conductivity of the material, and the power distribution strategy is adjusted to realize accurate temperature control. For the seven hot spot positions, power and temperature data are collected every second for 30 minutes. Taking hot spot 1 as an example, the initial temperature is 25℃, the target temperature is 35℃, the ambient temperature is 22℃, and the power starts from 0.5W and increases by 0.5W every 5 minutes until 3W. The least squares method is used to fit the temperature change curve, and the function T(t) = 10(1-e^(-0.005t))+25 is obtained, where T is the temperature and t is the time. Based on this model, a model predictive controller is designed, with a 5-minute prediction horizon, and the target function is Σ(T-Tset)^2+0.1Σ(ΔP)^2, where Tset is the set temperature and ΔP is the power change. If the target temperature is 33℃, the controller calculates the optimal power sequence for the next 5 minutes as [2.1W, 2.0W, 1.9W, 1.9W, 1.8W]. Considering the influence of adjacent hot spots, a 7x7 coupling coefficient matrix is established, such as the coupling coefficient between hot spot 1 and hot spot 2 is 0.2, which means that 20% of the heat of hot spot 1 affects hot spot 2. According to the coupling matrix, the power output of hot spot 1 is adjusted to 1.75W, which compensates for the 0.35W heat influence from hot spot 2, thus achieving accurate temperature control, and the temperatures of the seven hot spots are maintained within the range of 33±0.5℃.

[0037] In step S106, ergonomic simulation software is used to analyze the surface curve of the pillow and make final adjustments to the hot spot layout.

[0038] A three-dimensional pillow model was obtained and imported into ANSYS ergonomic simulation software. The density, elastic modulus, and thermal conductivity coefficient were set according to the pillow material properties to construct a virtual human head and neck model. The finite element analysis method was used to simulate the contact process between the human head and the pillow, calculate the pressure distribution and contact area, and identify the key support points and high-pressure areas. The hotspot layout plan was superimposed on the pressure distribution map to determine the degree of match between the hotspot position and the high-pressure area and key support points, and calculate the contact area covered by the hotspot and the range of influence. If the pressure distribution and curvature analysis results meet the preset conditions, the gradient descent method was used to adjust the hotspot position and power distribution to optimize the uniformity of heat distribution. The final hotspot layout plan was generated based on the optimization results, and the hotspot coordinates and power parameters were obtained.

[0039] Specifically, a 3D pillow model was imported into ANSYS ergonomic simulation software. Material properties, including density, elastic modulus, and thermal conductivity, were set. A virtual human head and neck model was constructed with a head weight of 4.5 kg and a neck length of 12 cm. Two sleeping positions, supine and side-lying, were defined. Static pressure analysis was performed, using finite element analysis to simulate the contact process between the human head and the pillow. Pressure distribution and contact area were calculated, key support points and high-pressure areas were identified, and a surface curvature distribution map of the pillow was obtained. The hotspot layout plan was superimposed on the pressure distribution map, and the degree of match between the hotspot locations and the high-pressure areas and key support points was analyzed. The contact area and influence range covered by each hotspot were calculated, and the matching degree was defined as the percentage of the overlapping area between the hotspot and the high-pressure area to the total hotspot area. Based on the pressure distribution and curvature analysis results, the hotspot locations and power distribution were adjusted using a gradient descent method to optimize heat distribution uniformity. The objective function was a weighted sum of the hotspot matching degree and temperature uniformity. The optimization was iteratively optimized until convergence or the maximum number of iterations was reached. The final hotspot layout plan was generated, and the hotspot coordinates and power parameters were output. Import a 60cm×40cm×15cm pillow 3D model into ANSYS software and set the memory foam material properties to: density 50kg / m 3, elastic modulus 4000Pa, thermal conductivity 0.04W / (m·K). A virtual human head and neck model was constructed: head weight 4.5kg, neck length 12cm, head tilt angle 15° in supine position, and neck bending angle 20° in side-lying position. Static pressure analysis was performed using tetrahedral meshing with a cell size of 0.5cm. The head gravity load was set and the maximum pressure was calculated to be 1.2kPa, located 5cm behind the center of the pillow. A curvature distribution map was generated with a maximum curvature radius of 25cm. Seven hotspot layout schemes were superimposed on the pressure distribution map, and the overlapping area between each hotspot and the high-pressure area was calculated. For example, the overlapping area of ​​hotspot 1 accounted for 75%. The gradient descent method was used to optimize the hotspot layout, with a learning rate of 0.01, a maximum number of iterations of 100, and an objective function of 0.7×matching degree + 0.3×temperature uniformity. After 63 iterations, the final hotspot placement solution was achieved: Hotspot 1 was moved to (28cm, 22cm) with a power adjustment of 2.8W; Hotspot 2 was moved to (35cm, 18cm) with a power adjustment of 2.5W. With this final solution, the average matching degree of all hotspots increased to 85%, and temperature uniformity improved by 12%.

[0040] Step S107: Based on the final adjusted hotspot arrangement plan, a control circuit and program for a temperature-adjustable pillow are established at the bottom of the pillow to process the data collected by the temperature sensor and the control strategy information output by the heating element, control the on and off of each heating element, and provide the user with temperature setting and mode selection functions.

[0041] The invention obtains an input signal of a microcontroller, wherein the input signal is generated by collecting temperature sensor data by an analog-to-digital converter; executes a control program according to the input signal, wherein the control program includes a filtering processing module, a PID control algorithm module and a PWM output control module; uses a weighted average data fusion algorithm to process data from multiple temperature sensors, and calculates the optimal control output in combination with control strategy information output by the PID controller; receives user input information sent by a human-computer interaction interface, wherein the user input information includes a temperature setting value and an operating mode selection; transmits the user input information to the microcontroller, and the microcontroller updates the control parameters according to the received user input information.

[0042] Specifically, according to the hot spot arrangement scheme, a control circuit is designed, an STM32F103 series microcontroller is selected as a main controller, a 12-bit analog-to-digital converter, a 5A rated current power drive module and a UART communication interface are equipped, and the temperature sensor and the heating element are connected to the corresponding circuit port. A control program is written to realize temperature data acquisition, filtering processing, PID control algorithm and PWM output control, a 1ms timing interrupt is set for periodic sampling and control update, and the Ziegler-Nichols tuning method is used to set the PID parameters. A weighted average data fusion algorithm is designed to comprehensively process the data of multiple temperature sensors and the control strategy information output by the PID controller, calculate the optimal control output, and judge the on-off state of each heating element according to the preset threshold 0.5℃. A man-machine interface is developed, including a 3.5-inch touch screen display module and four function keys, to realize temperature setting (adjustable from 18-32℃), mode selection (normal, energy saving, sleep) and state display functions, the interface layout includes a current temperature display area, a target temperature adjustment slider, mode selection buttons and a setting menu, user input is transmitted to the main controller through UART serial communication, and the refresh rate is 10Hz. Based on the arrangement scheme of 7 hot spots, the control circuit uses an STM32F103RBT6 microcontroller with a clock frequency of 72MHz, and 7 12-bit ADC channels are used to collect temperature data, and 7 5A rated current MOS tube drive circuits are used to control the heating elements. In the control program, 1ms timing interrupt triggers temperature sampling, 20-point moving average filtering is used, and the filtered data enters the PID controller. The PID parameters are tuned by the Ziegler-Nichols method, and Kp=2.5, Ki=0.8, Kd=0.4 are obtained. The PID output is processed by the weighted average data fusion algorithm, and the weights are allocated according to the importance of the hot spot position, such as the central hot spot weight 0.3 and the edge hot spot weight 0.1. The fused control output is modulated by the heating element power through PWM, the PWM frequency is 20kHz, and the resolution is 8 bits. The man-machine interface uses a 3.5-inch 320x240 resolution touch screen, displays the current temperature 33.5℃, the target temperature adjustment range 18-32℃, and the slider single step 0.5℃. The interface layout includes a temperature display area, which occupies 60% of the area, a temperature adjustment slider, which occupies 20% of the area, mode selection buttons, which occupy 15% of the area, including normal, energy saving, sleep three modes, and a setting menu, which occupies 5% of the area. The UART communication baud rate is set to 115200bps, and user input data is sent to the main controller every 100ms.

[0043] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. Any modifications, equivalent replacements and improvements made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for regulating the surface temperature of a pillow, characterized in that: The method comprises: Based on the pillow material and size, the initial parameters of the temperature hotspot layout scheme are determined, including the hotspot spacing and the number of hotspots. The temperature hotspot layout scheme is simulated to obtain a temperature distribution cloud map on the pillow surface. Analyze the temperature distribution cloud map and extract the temperature distribution uniformity index. If the uniformity index does not reach the preset threshold, adjust the parameters of the hotspot layout plan and repeat the simulation until the temperature distribution uniformity requirements are met; Under the premise of meeting the temperature distribution uniformity index, the hot spot layout plan is optimized to reduce the number of hot spots while ensuring the temperature regulation effect; According to the optimized hotspot layout plan, temperature sensors and heating elements are set at corresponding positions on the pillow surface. The temperature sensors are used to collect temperature data at the hotspots in real time, and the heating elements adjust the power according to the temperature data. Establish a mathematical model between the power of the heating element and the hotspot temperature, and calculate and obtain the control method of each heating element according to different target temperatures; Ergonomic simulation software was used to analyze the curved surface of the pillow and make final adjustments to the hotspot placement plan; According to the final adjusted hotspot arrangement plan, a control circuit and program for a temperature-adjustable pillow are established at the bottom of the pillow to process the data collected by the temperature sensor and the control strategy information output by the heating element, control the on and off of each heating element, and provide users with temperature setting and mode selection functions.

2. The method according to claim 1, characterized in that The initial parameters of the temperature hotspot arrangement scheme are determined based on the pillow material and size, including the hotspot spacing and the number of hotspots. The temperature hotspot arrangement scheme is simulated to obtain a temperature distribution cloud map on the pillow surface, including: Obtain pillow material parameters and size data and build a three-dimensional digital model, select the pillow material from the preset material list, and obtain the corresponding thermal conductivity and specific heat capacity parameters based on the selected material; Perform finite element analysis meshing on the three-dimensional digital model, set boundary conditions and initial temperature distribution; A genetic algorithm is used to optimize the preset hotspot arrangement scheme, with the hotspot spacing and number as decision variables and temperature uniformity as the objective function; If the fitness function is the inverse of the pillow surface temperature variance, then constraints on hotspot power and total energy consumption are added; The optimal hotspot layout parameters, including hotspot position coordinates and power values, are obtained through multiple iterative calculations; According to the optimal hotspot arrangement parameters, transient heat conduction simulation is performed, and the temperature field change is described using a two-dimensional heat conduction partial differential equation. The two-dimensional heat conduction partial differential equation is: u represents temperature, t represents time, x and y represent spatial coordinates, and α represents the thermal diffusivity; Discretize the temperature field equation, construct a differential format, and obtain the change process of the temperature field over time; Record the temperature value of each grid node at different times to form a temperature and time data matrix; Perform bilinear interpolation on the simulation result data to generate a continuous temperature distribution function; Map the temperature value to the corresponding color space; Draw a temperature cloud map of the pillow surface, generate a temperature scale, and mark the temperature ranges corresponding to different colors.

3. The method according to claim 1, characterized in that The temperature distribution cloud map is analyzed to extract the uniformity index of the temperature distribution. If the uniformity index does not reach a preset threshold, the parameters of the hotspot arrangement scheme are adjusted and the simulation is repeated until the temperature distribution uniformity requirement is met, including: Obtain a temperature distribution cloud map, and filter the temperature distribution cloud map using a Gaussian filter with a preset kernel size and standard deviation to obtain a filtered temperature distribution image; According to the filtered temperature distribution image, segmenting the filtered temperature distribution image into a plurality of temperature regions; For the multiple temperature regions, calculating the area and average temperature value of each temperature region; According to the area and average temperature value of each temperature zone, the temperature distribution uniformity index is calculated using the weighted standard deviation method; If the temperature distribution uniformity index is greater than a preset threshold, the hotspot arrangement scheme parameters are adjusted; According to the adjusted hotspot arrangement parameters, the heat source boundary conditions are updated, the heat conduction equation is solved, and a new temperature distribution cloud map is obtained; Repeat the filtering process, temperature region segmentation, uniformity index calculation, parameter adjustment, and heat conduction equation solution until the temperature distribution uniformity index is less than or equal to a preset threshold; A hotspot arrangement scheme that meets the preset threshold condition is output, where the hotspot arrangement scheme includes heat source coordinates and corresponding power values.

4. The method according to claim 1, wherein Under the premise of meeting the temperature distribution uniformity index, the hot spot arrangement scheme is optimized to reduce the number of hot spots while ensuring the temperature regulation effect, including: According to the existing hotspot arrangement scheme, a two-dimensional Gaussian diffusion model is used to calculate the temperature distribution around the hotspot to obtain the effective coverage radius of the hotspot, where the effective coverage radius is defined as the distance where the temperature drops to 1 / e of the initial value; A clustering algorithm is used to group hotspots. Hotspots with close proximity and overlapping temperature influence ranges are merged based on their coordinates, with the number of clusters gradually decreasing from the current number of hotspots. The updated hotspot layout plan is obtained by calculating the cluster center as the new hotspot location and taking the sum of all hotspot powers in the cluster as the new hotspot power. For the updated hotspot arrangement scheme, the finite element method is used to solve the heat conduction equation, the grid size is set, the boundary condition is natural convection, the time step is solved, and a new temperature distribution cloud map is obtained; If the temperature distribution uniformity index meets the preset threshold, the current hotspot arrangement plan is saved, the hotspot reduction operation is continued, and the clustering and updating operations are repeated until the temperature distribution uniformity index does not meet the requirements or the number of hotspots reaches the preset minimum value.

5. The method according to claim 1, wherein According to the optimized hotspot arrangement scheme, temperature sensors and heating elements are arranged at corresponding positions on the pillow surface. The temperature sensors are used to collect temperature data at the hotspots in real time, and the heating elements adjust power according to the temperature data, including: Obtaining hot spot distribution information on the pillow surface, and determining the arrangement positions of the temperature sensor and the heating element based on the hot spot distribution information; Wherein, the temperature sensor is a thermocouple type, and the heating element is a flexible thin film heater; Collect temperature data from the temperature sensor at a preset frequency; Converting the temperature data into a digital signal through an analog-to-digital converter and storing the digital signal in a circular queue buffer; The proportional-integral-differential control method is used to calculate the output power of the heating element based on the deviation between the current temperature and the target temperature; Wherein, the controller output of the proportional-integral-differential control method consists of a proportional term, an integral term and a differential term; The Ziegler-Nichols tuning method is used to determine the PID parameters, and the proportional gain is gradually increased until the system experiences amplitude oscillation. Record critical gain and oscillation period, and calculate PID parameters; Convert the calculated output power into a voltage signal; The power output of the heating element is controlled by a pulse width modulation method, wherein the modulation frequency and resolution of the pulse width modulation method are preset values.

6. The method according to claim 1, characterized in that The method of establishing a mathematical model between the power of the heating element and the hot spot temperature and calculating and obtaining the control method of each heating element according to different target temperatures includes: Obtaining historical data of heating element power and hot spot temperature, wherein the historical data includes temperature variation processes under different power inputs; Based on the historical data, the temperature variation curve is fitted using the least square method to obtain a mathematical model between the heating element power and the hot spot temperature; Based on the mathematical model, a feedback controller is used, wherein the feedback controller adopts a model predictive control algorithm to solve the optimal power sequence; The difference between the target temperature and the current temperature is input into the feedback controller to calculate the power control sequence of each heating element; Establishing a hotspot coupling coefficient matrix, wherein the hotspot coupling coefficient matrix is ​​determined according to the distance between hotspots and the thermal conductivity of the material; The power control sequence is adjusted according to the hotspot coupling coefficient matrix.

7. The method according to claim 1, characterized in that The ergonomic simulation software is used to analyze the curved surface of the pillow and make final adjustments to the hotspot arrangement plan, including: Obtain a 3D pillow model and import it into ANSYS ergonomic simulation software. Set the density, elastic modulus, and thermal conductivity coefficient based on the pillow material properties to construct a virtual human head and neck model. Finite element analysis is used to simulate the contact process between the human head and the pillow, calculate the pressure distribution and contact area, and determine the key support points and high-pressure areas; Overlay the hotspot layout plan onto the pressure distribution map to determine the degree of match between the hotspot location and the high-pressure area and key support points, and calculate the contact area and impact range covered by the hotspot; If the pressure distribution and curvature analysis results meet the preset conditions, the gradient descent method is used to adjust the hotspot position and power distribution to optimize the uniformity of heat distribution; The final hotspot layout plan is generated based on the optimization results, and the hotspot coordinates and power parameters are obtained.

8. The method according to claim 1, characterized in that According to the final adjusted hotspot arrangement plan, a control circuit and program for the temperature-adjustable pillow are established at the bottom of the pillow to process the data collected by the temperature sensor and the control strategy information output by the heating element, control the on and off of each heating element, and provide the user with temperature setting and mode selection functions, including: Obtaining an input signal from a microcontroller, where the input signal is generated by an analog-to-digital converter collecting temperature sensor data; Execute a control program according to the input signal, the control program including a filtering processing module, a PID control algorithm module and a PWM output control module; A weighted average data fusion algorithm is used to process data from multiple temperature sensors and the optimal control output is calculated based on the control strategy information output by the PID controller. Receiving user input information sent by the human-computer interaction interface, the user input information including a temperature setting value and an operation mode selection; The user input information is transmitted to the microcontroller, and the microcontroller updates the control parameters according to the received user input information.

Images