Self-adaptive height adjusting intelligent pillow based on recognition of six sleeping postures and control method of self-adaptive height adjusting intelligent pillow

Abstract

The invention discloses a self-adaptive height adjusting intelligent pillow based on recognition of six sleeping postures and a control method of the self-adaptive height adjusting intelligent pillow. The self-adaptive height adjusting intelligent pillow comprises a pillow body, a pressure sensing array, an upper computer, a communication module, an STM32 single-chip microcomputer control module, an air pump, an electromagnetic valve set and a multi-area air bag assembly below the pillow body. The pressure sensing array collects pressure distribution data of the pillow surface in real time, the upper computer carries out baseline calibration on the pressure data, calculates the characteristics of a pressure center, a local pressure center and the like, carries out secondary sleeping posture recognition in combination with pressure data of a left partition, a middle partition and a right partition of the pillow surface, and outputs six sleeping posture types. The problem that accurate sleeping posture recognition fails due to the fact that a traditional method depends on rough judgment of the head position is solved. The upper computer transmits an identification result to the STM32 single-chip microcomputer, the single-chip microcomputer controls the two air bags below each area to be inflated and deflated according to sleeping posture types, pressure feedback indexes are maintained within a health range defined by a configurable threshold parameter set, real-time adjustment of the height of the pillow is achieved, and self-adaptive sleeping posture adjustment and comfortable supporting are guaranteed.

Description

Technical Field

[0001] This invention relates to the field of smart home and healthy sleep aids technology, and in particular to a smart pillow and its control method that recognizes six sleeping positions through pressure distribution, sends the recognition results to an STM32 microcontroller to control the inflation and deflation of multi-area airbags, so that the pressure feedback index enters the health threshold range defined by a set of configurable threshold parameters, thereby achieving adaptive pillow height adjustment. Background Technology

[0002] Current smart pillows on the market typically use a fixed height or manual adjustment, which is difficult to adapt to the frequent changes in posture during sleep. As sleeping positions change, the support requirements for the neck and shoulders also vary. Because changes in sleeping position require different pillow heights, fixed-height pillows cannot provide continuous comfortable support, easily leading to health problems such as neck discomfort, muscle fatigue, or uneven spinal pressure.

[0003] Most existing smart pillows rely on data from a single pressure sensor to determine the user's sleeping posture and adjust the pillow height. However, many existing solutions only rely on the pillow's position to roughly determine sleeping posture, such as dividing the head position into left, right, and middle areas, and then assuming that the "middle area" is a supine sleeping position, while the left and right sides are a side sleeping position. This simple method cannot accurately reflect the complex changes in posture during actual sleep, especially ignoring the possibility of "middle-zone side sleeping" or unilateral side sleeping. This leads to a mismatch between the sleeping posture determination and height adjustment strategies of traditional smart pillows, failing to meet personalized needs.

[0004] Furthermore, current pillow height adjustment technologies mostly rely on a single pressure distribution or sleeping position zone division, lacking adjustment mechanisms tailored to individual user differences. Everyone's shoulder width, neck length, and sleeping habits differ, thus requiring personalized pressure feedback adjustment solutions. Traditional pillow height adjustment methods cannot provide real-time, precise height adjustments based on user differences, resulting in excessively strong or weak support, which in turn affects sleep quality and neck and shoulder health.

[0005] To address these issues, this invention proposes a smart pillow based on precise recognition of six sleeping postures. By combining secondary precise sleeping posture recognition with personalized pressure adjustment, it can accurately identify various postures of the user during sleep, including complex situations such as mid-side sleeping and partial side sleeping. Furthermore, this invention, taking into account user differences, employs a configurable threshold parameter set to adjust the pressure feedback index to a clinically defined healthy pressure range, achieving highly adjustable personalized and adaptive closed-loop control.

[0006] Based on this, the smart pillow of the present invention can adjust the pillow height in real time according to the user's body shape differences (such as shoulder width, neck length, head weight, etc.) to maintain it within a healthy pressure range. Through this technical solution, the present invention not only improves the accuracy of sleeping posture recognition, but also adjusts the pressure feedback according to the specific needs of each user, achieving a personalized, comfortable and healthy sleep experience. Summary of the Invention

[0007] The purpose of this invention is to provide a smart pillow and its control method, realizing a closed-loop process of "first recognition - then feedback - finally control": six sleeping positions are identified through Python algorithm, the recognition results are sent to STM32 microcontroller, the microcontroller controls the inflation and deflation of two airbags in each of the left / middle / right areas, and the adaptive height adjustment is completed with the pressure feedback index entering the healthy threshold range as the goal.

[0008] Technical solution The smart pillow comprises a pillow body, a pressure sensor array, a host computer, a communication module, an STM32 microcontroller, an air pump and solenoid valve assembly, and an airbag assembly. The airbag assembly is divided into three regions: left, middle, and right, with at least two independently inflatable and deflated airbags in each region. The host computer performs baseline calibration on the pressure data. First, it determines the main stress areas based on the aggregated pressure of the left / middle / right regions. Then, within the main stress areas, it performs secondary identification based on regional characteristics such as the local center of pressure (local COP), the left-right pressure asymmetry within the region, pressure concentration, and peak pressure distribution pattern to distinguish between supine and side sleeping positions, thus outputting six sleeping position categories. The host computer sends the sleeping position categories to the STM32 microcontroller via the communication module. The STM32 microcontroller selects the target region according to the sleeping position category and controls the solenoid valve and air pump to inflate, deflate, or maintain pressure in the two airbags of that region. The control target is defined by a configurable threshold parameter set T. The threshold parameter set T contains one or more pressure feedback indicators and their upper and lower thresholds. Selectable indicators include average pressure, peak pressure, pressure concentration, pressure center offset, and asymmetry, and can be configured according to sleeping position type and region. The system performs closed-loop adjustment based on the comparison results between the feedback indicators and the threshold parameter set T, ensuring that the indicators enter and remain within the healthy threshold range.

[0009] Beneficial effects (1) “Regional judgment + secondary recognition within the region” covers the central area for side sleeping, avoiding misadjustment caused by only coarse judgment based on position; (2) The recognition result is sent to STM32 to drive the valve pump to control the three-area dual airbags, realizing automatic height adjustment; (3) The health threshold is expressed by a configurable parameter set, which does not expose the specific value, but realizes quantification, reproducibility and individualized calibration; (4) The dual airbags in each of the three areas improve the support precision and stability, and can be expanded to more partitions and more index combinations; (5) Safety pressure limit, abnormal degradation and shock suppression strategies can be set to improve safety and comfort. Attached Figure Description

[0010] Figure 1 This is a schematic diagram of the overall system structure; Figure 2 A schematic diagram of the arrangement of dual airbags in three areas below the pillow body; Figure 3 A schematic diagram showing the pressure sensor array locations and its left / middle / right partitions; Figure 4 Flowchart for recognizing six sleeping positions; Figure 5 The flowchart shows the communication and closed-loop control process. Figure 6 This is a schematic diagram showing the mapping between six types of sleeping positions and control strategies. Detailed Implementation

[0011] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. Those skilled in the art can make equivalent substitutions or improvements to the structural form, number of sensors, index combinations, and threshold parameters without departing from the concept of the present invention, and all such substitutions or improvements should fall within the protection scope of the present invention.

[0012] Example 1: Hardware Structure. A pressure sensor array is installed inside the pillow to collect the pressure distribution on the pillow surface. An airbag assembly is located below the pillow, divided into three areas: left, middle, and right, with two airbags in each area. An air pump and a solenoid valve assembly constitute the inflation / deflation actuator, used to inflate, deflate, and maintain pressure in each airbag. The host computer is connected to an STM32 microcontroller via a communication module, and the STM32 outputs valve and pump control signals.

[0013] Example 2: Baseline Calibration and Data Preprocessing. When the system starts up or detects an unloaded state, pressure array data for a preset duration is collected and the average value is used as the baseline; real-time pressure data is calibrated according to "real-time value - baseline value" to suppress sensor drift and environmental noise.

[0014] Example 3: Six Sleeping Position Recognition. First, the pressure point set is divided into left, middle, and right zones. The aggregated pressure across these three zones is calculated to determine the primary stress areas. Then, within the primary stress areas, the local pressure center (local-COP) and related regional features (asymmetry, concentration, peak shape, etc.) are calculated. A second-level recognition process distinguishes between supine and side sleeping positions, resulting in six sleeping position categories. This recognition method does not rely solely on "position = sleeping position" and can handle situations where the user is in the middle zone but still sleeping on their side.

[0015] Example 4: Communication and Control. The host computer encodes the six sleeping posture categories into instructions and sends them to the STM32 via serial port or other means. After receiving the instructions, the STM32 maps the sleeping posture categories to valve pump control strategies, selecting two airbags in the target area for inflation, deflation, or pressure maintenance.

[0016] Example 5: Closed-loop adjustment of health thresholds based on a configurable threshold parameter set. The threshold parameter set T can be represented as: T = {(mj, Lj, Uj)}. The index mj can be average pressure, peak pressure, pressure concentration, pressure center offset, asymmetry, etc. The system calculates the index on the host computer and compares it with the threshold set: when an index is below the lower limit Lj, inflation is performed; when an index is above the upper limit Uj, deflation is performed; when all indexes are within the corresponding range, pressure is maintained or fine-tuned in small steps, and the current round of adjustment is determined to be complete. The threshold parameter set can be stored separately according to sleeping posture category and region, and can be configured, updated through the host computer interface or generated in the individualized calibration process.

[0017] Example 6: Suppression of jitter and safety protection. The control strategy can be set with minimum adjustment interval, minimum effective deviation threshold, and pressure holding window; at the same time, safety limits such as maximum allowable pressure and maximum allowable inflation time can be set. When abnormal pressure data, valve / pump malfunction, or over-limit is detected, the system enters protection mode: inflation stops and pressure is released or the pressure drops back to the safe default height.

[0018] Figure Labels 1 Pillowcase; 2 Pressure sensor array; 3 Pillow body; 4 Airbag assembly; 41 Left zone first airbag; 42 Left zone second airbag; 43 Central zone first airbag; 44 Central zone second airbag; 45 Right zone first airbag; 46 Right zone second airbag; 5 Tracheal assembly; 51 Right zone tracheal tube; 52 Central zone tracheal tube; 53 Left zone tracheal tube; 6 Inflation / depression actuator; 61 Air pump; 62 Solenoid valve assembly; 7 STM32 microcontroller.

Claims

1. A smart pillow with adaptive height adjustment based on six sleeping postures, characterized in that, The device includes a pillow body; a pressure sensor array, located on the upper surface or inside the pillow body, for collecting pressure distribution data on the pillow surface; a host computer, including a Python algorithm processing module and a display module; a communication module; an STM32 microcontroller control module; an inflation / deflation actuator, including an air pump and a solenoid valve assembly; and an airbag assembly, located below the pillow body and divided into three areas: left, middle, and right, with at least two independently inflatable / deflated airbags in each area. The host computer is configured to: perform baseline acquisition and calibration of the pressure distribution data from the pressure sensor array; determine the current main pressure area based on the aggregated pressure in the left, middle, and right areas; and within the main pressure area, without relying on a single rule of "position = posture," perform secondary recognition based on at least one feature within the area to distinguish between supine and side sleeping positions. This feature includes at least one or more of the following: local center of pressure (local-COP), left-right pressure asymmetry within the area, pressure concentration, and peak pressure distribution pattern, thereby outputting six sleeping position categories. The six sleeping position categories include at least left supine sleeping, left side sleeping, middle supine sleeping, middle side sleeping, right supine sleeping, and right side sleeping; the host computer sends the six sleeping position categories to the STM32 microcontroller control module; the STM32 microcontroller control module selects two airbags in the corresponding area according to the six sleeping position categories and controls the inflation / deflation actuator to inflate, deflate, or maintain pressure on them; wherein, the control target of inflation / deflation is defined by a configurable threshold parameter set, and each pressure feedback index (such as average pressure, peak pressure, pressure concentration, etc.) in the parameter set has upper and lower limit thresholds. When the pressure feedback index exceeds the preset threshold range, the STM32 microcontroller controls the airbags in the corresponding area to inflate or deflate, ensuring that the pressure is always kept within a healthy range. The configurable threshold parameter set includes one or more pressure feedback indices and their upper and lower thresholds. The pressure feedback indices include at least one or more of the following: average pressure, peak pressure, pressure concentration, pressure center offset, and pressure asymmetry. Based on the pressure feedback indices calculated and fed back by the host computer, the STM32 microcontroller control module ensures that the pressure feedback index of the target area enters and remains within the healthy threshold range corresponding to the configurable threshold parameter set, thereby achieving closed-loop adaptive height adjustment.

2. The smart pillow according to claim 1, characterized in that: The pressure sensing array is a multi-point array sensor. The collected data is used to calculate the center of pressure (COP) and local-COP after baseline calibration.

3. The smart pillow according to claim 1, characterized in that: The main stress area is determined by the largest of the total pressure in the left zone, the total pressure in the middle zone, and the total pressure in the right zone.

4. The smart pillow according to claim 3, characterized in that: When the difference between the maximum region aggregation pressure and the second largest region aggregation pressure is less than a preset proportional threshold, the host computer outputs the mixed region status, and the STM32 uses a weighted control strategy to coordinate the adjustment of the airbags in the two adjacent regions.

5. The smart pillow according to claim 1, characterized in that: The configurable threshold parameter set is a parameter set T, represented as: T = {(m1, L1, U1), (m2, L2, U2), …, (mk, Lk, Uk)}, where k is a positive integer; mj is the j-th pressure feedback index, and Lj and Uj are the lower and upper thresholds of the index, respectively; and the parameter set T is stored according to sleeping posture category and / or region.

6. The smart pillow according to claim 5, characterized in that: The pressure feedback index mj includes at least one of the following: peak pressure, average pressure, pressure concentration, high pressure point percentage, pressure distribution entropy, pressure center offset, and left-right pressure asymmetry; wherein pressure concentration is any one of the following: the ratio of peak pressure to average pressure, the high pressure point percentage, or the entropy value.

7. The smart pillow according to claim 5, characterized in that: The parameter set T can be configured and updated through the human-machine interface of the host computer, or automatically generated and written to the memory through an individualized calibration process; the individualized calibration process considers at least one or more of the user's shoulder width, neck length, and mattress firmness parameters.

8. The smart pillow according to claim 1, characterized in that: The two airbags in each region of the airbag assembly are either a "support airbag + fine-tuning airbag" structure or a "front and rear distributed airbag" structure, and different sleeping positions correspond to different inflation and deflation priorities and valve combinations.

9. The smart pillow according to claim 1, characterized in that: The host computer and the STM32 microcontroller control module communicate via serial port. The six sleeping posture categories are sent in the form of coded instructions. The STM32 maps the coded instructions to the opening and closing combinations of the solenoid valve group and the start and stop strategies of the air pump.

10. The smart pillow according to claim 1, characterized in that: The closed-loop regulation includes a jitter suppression strategy, which includes at least one or more of the following: minimum adjustment interval, minimum effective deviation threshold, and pressure holding window, to reduce frequent adjustments during sleep.

Images