A hidden roller type leisure elderly chair based on an intelligent control algorithm

Abstract

The present invention relates to the technical field of furniture equipment, and particularly relates to a hidden roller type leisure elderly chair based on an intelligent control algorithm. It includes: a control unit, which undertakes the real-time calculation work for the operation of the control system; a detection unit, which is used to detect various physical indicators, body postures, and states of the user; an execution unit; a human-computer interaction unit, which collects signals through touch perception, voice, and gestures and sends them to the control unit, and the control unit performs calculations and issues control instructions to the execution unit; a driving unit, when it is necessary to move the elderly chair, it drives the wheels of the motor to automatically lift to achieve movement, and when the elderly chair is placed at a certain position, the wheels are automatically hidden through the control unit. The hidden wheel structure is adopted. When it is necessary to move the elderly chair, the control unit issues control instructions to the driving unit through the human-computer interaction unit or gravity sensing, so as to achieve movement, which can greatly ensure the stability and safety of the elderly chair.

Description

Technical Field

[0001] The present invention discloses a leisure chair for the elderly, belonging to the technical field of furniture equipment, and particularly relates to a hidden roller type leisure chair for the elderly based on an intelligent control algorithm. Background Art

[0002] A leisure chair for the elderly is a piece of furniture designed for comfort and suitable for the elderly to rest. It usually has the following characteristics: comfort, the leisure chair for the elderly usually uses soft materials such as fabric or leather to provide a comfortable sitting feeling and support; ergonomic design, these chairs usually take into account the physical characteristics of the elderly, such as providing good lumbar support, comfortable backrests and armrests, etc.; easy to use, with a simple and easy-to-understand design, so that the elderly can easily use and move it; safety, the leisure chair for the elderly usually has a stable structure and a safe design to prevent the elderly from falling or getting injured during use; functionality, some leisure chairs for the elderly may also have other functions, such as adjustable backrest angles, footrests, massage functions, etc., to provide more comfortable experiences. These chairs are commonly found in nursing homes, families and public places, providing a comfortable and safe resting place for the elderly.

[0003] Chinese Patent with the publication number CN110226850A discloses an electric leisure chair for the elderly, proposing an electric leisure chair for the elderly. The electric leisure chair for the elderly includes a backrest assembly, a seat cushion assembly, and a footrest assembly. The backrest assembly is fixedly connected to the seat cushion assembly, and the footrest assembly is rotatably and foldably connected to the seat cushion assembly. The electric leisure chair for the elderly is provided with a first lifting mechanism for adjusting the rotation and lifting of the seat cushion assembly. Through the setting of the first lifting mechanism, the elderly can easily stand up from the chair. Through the setting that the backrest assembly is fixedly connected to the seat cushion assembly and the footrest assembly is rotatably and foldably connected to the seat cushion assembly, the leisure chair can realize different functions such as lying position, sitting position and standing position.

[0004] Chinese Patent with the publication number CN216494361U discloses a multifunctional fitness and leisure chair for the elderly, including a seat board. A plurality of support legs are fixedly connected to the bottom of the seat board. Armrests are fixedly connected to both sides of the top of the seat board. An extrusion groove is formed in the seat board. A plurality of extrusion springs are fixedly connected in the extrusion groove. One end of each of the plurality of extrusion springs away from the extrusion groove is fixedly connected to a first extrusion plate. An air storage tank is fixedly connected to the bottom of the seat board. An extrusion rod is fixedly connected to the bottom of the first extrusion plate. One end of the extrusion rod away from the first extrusion plate is fixedly connected to a second extrusion plate. An air inlet pipe is fixedly connected to the bottom of the air storage tank. Through the cooperation among the extrusion plate, the extrusion spring, the extrusion rod, the extrusion plate, the air storage tank, the air inlet pipe and the airbag, when the leisure chair is idle, the airbag is in a deflated state, avoiding damage to the airbag caused by sharp objects, improving the comfort of the elderly and reducing economic losses.

[0005] However, the existing leisure chairs for the elderly still have the following problems:

[0006] (1) It can only perform simple massage and manual angle adjustment functions, lacking intelligence;

[0007] (2) The mobile phone-controlled intelligent product has realized functions such as remote control of the chair, real-time data viewing, and analysis reports to a certain extent. However, the need to control it through the mobile phone APP has also hindered the acceptance and use of the product by the elderly. Especially, there are numerous mobile phone APP interfaces, and the cumbersome and complex usage process also makes the elderly shy away from the product, generating a sense of frustration and rejection towards intelligent products;

[0008] (3) The driving wheels are mostly fixed. The driving wheels are in a working state for a long time. Such exposed driving wheels will make the elderly chair unstable during daily use, and there will be small movements over time, posing a safety hazard. SUMMARY OF THE INVENTION

[0009] To achieve the object of the present invention, the present invention is realized through the following technical solutions: A hidden roller type leisure chair for the elderly based on an intelligent control algorithm, comprising:

[0010] A control unit, composed of an embedded microprocessor board, which undertakes the real-time calculation work of the control system operation;

[0011] A detection unit, used to detect various physical indicators, body postures, and states of the user, and feed the detection data back to the control unit;

[0012] An execution unit, according to the instructions of the control unit, adjusts the device state and gives reminders based on the detection results of the detection unit;

[0013] A human-computer interaction unit, which collects signals through touch perception, voice, and gestures and sends them to the control unit. The control unit calculates and issues control instructions to the execution unit;

[0014] A driving unit, when it is necessary to move the elderly chair, drives the wheels of the motor to automatically lift for movement. When the elderly chair is placed at a certain position, the wheels are automatically hidden through the control unit.

[0015] Further, the detection unit includes:

[0016] A posture detection module, which automatically detects the heights and sitting postures of different elderly people and sends them to the control unit, and automatically adjusts the inclination angle of the backrest according to the reliance and supine situations in the sitting posture state of the elderly;

[0017] A sleep detection module, which collects sleep physiological signals through a microphone sensor and sends them to the control unit for judgment, and adjusts the sitting posture of the seat according to the results;

[0018] A health detection module, through multiple sensors, automatically detects data of human body characteristics such as the sitting posture, heart rate, and respiratory rate of the elderly when they sit down, and sends the collected data to the control unit to analyze the physical condition of the elderly;

[0019] A temperature detection module detects the seat temperature and ambient temperature through a thermostat and sends the signal data to the control unit to control the seat temperature in combination with the weather conditions.

[0020] Further, the pose detection module includes: behavior pose detection and sitting posture detection;

[0021] The sitting posture detection is for the adjustment of the sitting height and the inclination angle of the backrest. It automatically adjusts the height according to the heights and sitting postures of different elderly people, and automatically adjusts the inclination angle of the backrest according to the leaning and supine situations when the elderly are in the sitting posture, ensuring the comfort of the elderly's sitting posture;

[0022] The behavior pose detection collects the actions of the elderly on the chair and conducts behavior recognition and motion analysis. First, it locates the head and shoulders of the human body; then it calculates the head pose using the planar imaging characteristics of the human head, and at the same time calculates the torso pose using the contour change characteristics of the human shoulders; finally, it estimates the pose of the human body in motion by combining the poses of the head and torso.

[0023] Further, the specific steps of the sitting posture detection are as follows:

[0024] Collect images of the user on the elderly chair through a camera and output the coordinate information of the joints of the person in the image to the control unit;

[0025] Use a convolutional neural network to predict the confidence map corresponding to the body parts, and use a geometric transformation kernel to realize the transmission of information between adjacent joints;

[0026] Again, use the bidirectional graph structure information model incorporated in the convolutional neural network to realize further information transmission;

[0027] Finally, superimpose the joint information obtained through the geometric transformation kernel and the joint point information obtained from the bidirectional graph structure information model to obtain the specific positions of all the final joints, completing the sitting posture detection.

[0028] Further, the sleep detection module includes a microphone sensor and a sleep physiological signal acquisition module;

[0029] The output end of the microphone sensor is connected to the input end of the sleep physiological signal acquisition module, and the output end of the sleep physiological signal acquisition module is wirelessly connected to the signal processing terminal; the sleep physiological signal acquisition module includes a signal filter, a signal amplifier, a processor, a memory, and a wireless transmission module. The output end of the microphone sensor is connected to the input end of the signal filter, the output end of the signal filter is connected to the input end of the signal amplifier, the output end of the signal filter is connected to the control unit, and the signal filter includes a pulse signal filter, a respiration signal filter, and a displacement signal filter.

[0030] Furthermore, the working method of the sleep detection module is as follows:

[0031] Collect sleep physiological signals through the microphone sensor and transmit them to the signal filter;

[0032] Filter out the noise of the sleep physiological signals through the signal filter and transmit them to the signal amplifier;

[0033] Amplify the sleep physiological signals through the signal amplifier and transmit them to the analog-to-digital converter;

[0034] The sleep physiological signals are analog signals. Convert the analog signals into digital signals through the analog-to-digital converter and save them to the memory;

[0035] At regular intervals, read the data from the memory through the control unit and process it.

[0036] Furthermore, the driving unit adopts a hidden-wheel drive. When the elderly chair needs to be moved, the control unit issues a control instruction to the driving unit through the human-machine interaction unit or gravity sensing, so as to achieve movement;

[0037] Human-machine interaction is that the human-machine interaction unit issues a control signal to the control unit through voice, gesture or button. The control unit issues a control instruction to the driving unit, so as to achieve movement;

[0038] Gravity sensing realizes the weight detection and judgment of the elderly chair through a weighing sensor. When the weight is higher than the set threshold, a control signal is sent to the control unit. The control unit issues a control instruction to the driving unit, and the hidden wheels extend and wait for the user's next instruction.

[0039] The beneficial effects of the present invention are:

[0040] (1) The leisure elderly chair. A leisure chair with a massage function can relieve lumbar pressure. When the elderly sit on it, they can not only relax physically and mentally, but also enjoy the cervical care function. The intelligent leisure chair adjusts the state of the seat according to the correct sitting posture set in the user's computer system program and the actual body shape of the user, and adjusts the state of the seat in combination with the actual body type of the user to enable people to maintain the most comfortable sitting posture that is least likely to cause physical damage. The intelligent leisure elderly chair not only improves the elderly's life, but also brings a new intelligent experience to the design of elderly products.

[0041] (2) Adopt a hidden wheel structure. When the elderly chair needs to be moved, through the human-computer interaction unit or gravity sensing, the control unit issues a control instruction to the drive unit to achieve movement, which can greatly ensure the stability and safety of the elderly chair, and at the same time is more intelligent and beautiful. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a schematic diagram of the leisure elderly chair system of the present invention;

[0043] Figure 2 is a schematic diagram of the detection unit of the present invention;

[0044] Figure 3 is a schematic diagram of the pose detection module of the present invention;

[0045] Figure 4 is a schematic diagram of the sitting posture detection of the present invention;

[0046] Figure 5 is a schematic diagram of the behavior pose detection of the present invention;

[0047] Figure 6 is a schematic diagram of the sleep detection module of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0048] In order to deepen the understanding of the present invention, the following will further describe the present invention in conjunction with embodiments. These embodiments are only used to explain the present invention and do not constitute a limitation on the protection scope of the present invention.

[0049] Embodiment 1

[0050] According to Figures 1-6 shown, this embodiment provides a hidden roller type leisure elderly chair based on an intelligent control algorithm, including:

[0051] A control unit, which is composed of an embedded microprocessor board and undertakes the real-time calculation work of the control system operation;

[0052] A detection unit, which is used to detect various physical indicators, body poses, and states of the user, and feedback the detection data to the control unit;

[0053] An execution unit that adjusts the device state and gives reminders according to the control unit instructions based on the detection results of the detection unit;

[0054] A human-computer interaction unit that collects signals through touch perception, voice, and gestures and sends them to the control unit. The control unit performs calculations and issues control instructions to the execution unit;

[0055] A driving unit that, when the elderly chair needs to be moved, drives the motor to automatically lift the wheels for movement. When the elderly chair is placed at a certain position, the wheels are automatically hidden through the control unit.

[0056] More specifically, the detection unit includes:

[0057] A pose detection module that automatically detects the height and sitting postures of different elderly people and sends them to the control unit, and automatically adjusts the tilt angle of the backrest according to the reliance and supine conditions of the elderly people in the sitting posture;

[0058] A sleep detection module that collects sleep physiological signals through a microphone sensor and sends them to the control unit for judgment, and adjusts the sitting posture of the seat according to the results;

[0059] A health detection module that, through multiple sensors, automatically detects data on the sitting posture, heart rate, and respiratory rate of the elderly people when they sit down, and analyzes the physical conditions of the elderly people according to the collected data and sends them to the control unit;

[0060] A temperature detection module that detects the seat temperature and ambient temperature through a thermostat and sends the signal data to the control unit, and controls the seat temperature in combination with the weather conditions.

[0061] More specifically, the pose detection module includes: behavior pose detection and sitting posture detection;

[0062] The sitting posture detection is for the adjustment of the sitting height and the tilt angle of the backrest. It automatically adjusts the height according to the height and sitting postures of different elderly people, and automatically adjusts the tilt angle of the backrest according to the reliance and supine conditions of the elderly people in the sitting posture to ensure the comfort of the elderly people's sitting postures;

[0063] The behavior pose detection collects the actions of the elderly people on the chair and performs behavior recognition and motion analysis. First, it locates the head and shoulders of the human body; then it calculates the head pose using the planar imaging characteristics of the human head, and at the same time calculates the torso pose using the contour change characteristics of the human shoulders; finally, it estimates the pose of the human body in motion by combining the head and torso poses.

[0064] More specifically, the specific steps of the sitting posture detection are as follows:

[0065] Collect images of the user on the elderly chair through a camera, and output the coordinate information of the joint points of the person in the image to the control unit;

[0066] Use a convolutional neural network to predict the confidence map corresponding to the body parts, and use a geometric transformation kernel to achieve the transmission of information between adjacent joint points;

[0067] Again, use the convolutional neural network to integrate the bidirectional graph structure information model to achieve further information transmission;

[0068] Finally, superimpose the joint information obtained through the geometric transformation kernel and the joint point information obtained from the bidirectional graph structure information model to obtain the specific positions of all the final joint points, completing the sitting posture detection.

[0069] More specifically, the behavior pose detection first performs human contour feature modeling. Model the human contour features through 3 fitting functions. When modeling, first obtain the smallest vertical rectangle of the human body area and extract the human contour within the rectangular area. Define the rectangle height as h; then start from the highest point of the human body and use h horizontal lines to intercept the human contour. Each horizontal line can intercept two cut points, one on the left and one on the right, and obtain the following two feature vectors:

[0070] R = (r 1 , r 2 , … r n )

[0071] L = (l 1 , l 2 , … l n )

[0072] Among them, r n represents the abscissa of the right intersection point of the nth horizontal line and the human contour; l n represents the abscissa of the left intersection point of the nth horizontal line and the human contour; and define the vector as:

[0073] W = (w 1 , w 2 , … w n )

[0074] w n = r n - l n

[0075] Among them, w n represents the human body width at the nth horizontal line;

[0076] The feature vectors R, L, and W can be respectively fitted into functions f R (x), f L (x) and f W (x).

[0077] More specifically, when the human body contour feature modeling is completed, the image is successively subjected to the positioning of the human head and shoulders, the calculation of the human head pose, and the calculation of the human torso pose;

[0078] Since the positioning of the human head and shoulders is based on the elliptical characteristic presented by the human head contour, the function f W (x) representing the human body width will have a local minimum at the head boundary line. Let X min be the set of all x values at which the function f W (x) obtains local minimum points. Then the position of the human head boundary line is defined as:

[0079] X HL = min(x n ), x n ∈ X

[0080] That is, the minimum value in the set X min ; the shoulder position is defined as:

[0081] X SL = X HL + K * X HL

[0082] K is a proportionality coefficient, thus completing the segmentation and positioning of the human head and shoulders;

[0083] The calculation of the human head pose is to let S LF , S RF , S H represent the area of the skin color region of the left half of the human head, the area of the skin color region of the right half of the human head, and the area of the entire head region respectively. Then the human head pose variable P H is defined as:

[0084]

[0085] where T is a pre-set threshold. When P H = 1, it means the face is facing away from the camera. When P H = 0, it means the face is facing the camera. When P H > 0, it means the face is facing the left side of the camera. When P H < 0, it means the face is facing the right side of the camera; First, calculate S LF , S RF . When the two are relatively close, it is determined that the face is facing the camera or away from the camera. On this basis, calculate When then the proportion of the skin color region in the head region is relatively large, that is, it is determined that the face is facing the camera, otherwise it is determined that the face is facing away from the camera; If S LF , SRF If there is a large difference, it can be determined that the face is facing the left or right side of the camera; by calculating S LF -S RF and giving the final judgment based on the value. If the value is greater than 0, it means the face is facing the left side of the camera; if the value is less than 0, it means the face is facing the right side of the camera;

[0086] The calculation of the human torso pose first defines the human torso pose variable P B The calculation method is as follows:

[0087]

[0088] where T is a preset threshold, X HL represents the position of the human head boundary line, X SL represents the position of the shoulder boundary line, and X SL >X HL , in the formula represents the gradient of the left contour line of the human body in the shoulder area, represents the gradient of the right contour line in the shoulder area, P B <0 indicates that the gradient of the left contour line in the shoulder area is larger, and the left human contour is closer to the camera; P B >0 indicates that the right human contour is closer to the camera; P B =0 indicates that the left and right contour lines are basically symmetric in the shoulder area, and at this time the human body is facing or facing away from the camera.

[0089] More specifically, the sleep detection module includes a microphone sensor and a sleep physiological signal acquisition module;

[0090] The output end of the microphone sensor is connected to the input end of the sleep physiological signal acquisition module, and the output end of the sleep physiological signal acquisition module is wirelessly connected to the signal processing terminal; the sleep physiological signal acquisition module includes a signal filter, a signal amplifier, a processor, a memory and a wireless transmission module. The output end of the microphone sensor is connected to the input end of the signal filter, the output end of the signal filter is connected to the input end of the signal amplifier, the output end of the signal filter is connected to the control unit, and the signal filter includes a pulse signal filter, a respiratory signal filter and a displacement signal filter.

[0091] More specifically, the working method of the sleep detection module is as follows:

[0092] Collect sleep physiological signals through the microphone sensor and transmit them to the signal filter;

[0093] Filter out the noise of the sleep physiological signals through the signal filter and transmit them to the signal amplifier;

[0094] The sleep physiological signals are amplified by a signal amplifier and transmitted to an analog-to-digital converter;

[0095] The sleep physiological signals are analog signals, which are converted into digital signals by an analog-to-digital converter and saved to a memory;

[0096] At regular intervals, the control unit reads out the data from the memory and processes it.

[0097] More specifically, the driving unit adopts a hidden-wheel drive. When the elderly chair needs to be moved, the control unit issues a control instruction to the driving unit through the human-computer interaction unit or gravity sensing, so as to achieve movement;

[0098] Human-computer interaction means that the human-computer interaction unit issues a control signal to the control unit through voice, gesture or button. The control unit issues a control instruction to the driving unit, so as to achieve movement;

[0099] Gravity sensing realizes the weight detection and judgment of the elderly chair through a weighing sensor. When the weight is higher than the set threshold, a control signal is sent to the control unit. The control unit issues a control instruction to the driving unit, and the hidden wheels extend and wait for the user's next instruction.

[0100] More specifically, the gravity sensing includes: an equal-strength cantilever beam, a weighing sensor composed of strain gauges and a conversion module; when there is a person on the elderly chair, the cantilever beam generates a small deformation due to the force, so that the weighing sensor outputs a pair of differential-mode voltage signals. The conversion module first amplifies the signal and then converts it into a digital quantity for the control unit to use;

[0101] High-precision resistance strain gauges are pasted at the equal-strength area of the cantilever beam, with two distributed at equal distances on each of the upper and lower sides. Then the resistance strain gauges are connected in the form of a bridge. When the cantilever beam is subjected to a downward pressure, the upper strain gauge will be subjected to tensile stress and the resistance will increase; the lower strain gauge will be subjected to compressive stress and the resistance will decrease;

[0102] The detection signal of the weighing sensor is sent to the conversion module. The conversion module first performs data conversion. Let S be the sensitivity of the weighing sensor, F max be its maximum range, u s be the power supply voltage of the weighing sensor, M min be the minimum load that the weighing sensor can distinguish. Therefore, the output u min of the weighing sensor under the minimum load is:

[0103]

[0104] A 24-bit A / D converter is adopted, and the minimum input voltage uin min resolved is:

[0105]

[0106] When the load cell bears the full load, the output voltage u of the sensor max is:

[0107]

[0108] The maximum input voltage uin of the A / D converter max = Vref, so the maximum amplification factor A that the signal can achieve is max is:

[0109]

[0110] Filter the converted signal, and use the previous estimated value and the current actual measured value to estimate the current value of the signal, realizing continuous prediction and correction. Let the linear discrete system equation be:

[0111] x(n) = Ax(n - 1) + Bu(n) + w(n)

[0112] z(n) = H(n)x(n) + v(n)

[0113] where x(n) is the state vector of the system, u(n) is the input vector, w(n) is the estimation noise vector, A and B are the system matrices, z(n) is the measurement vector, H(n) is the measurement transfer matrix, and v(n) is the observation noise vector; among them, both the estimation noise and the observation noise are white noises that conform to the Gaussian distribution. Extract the white noise and perform the state prediction result at time (n) of the signal. Combining the measured value and the predicted value, the optimal estimated value of the system at time n can be calculated, and the steps are repeated to complete the filtering;

[0114] The filtered data is sent to the control unit through the communication module; the control unit controls the lifting of the drive wheel according to the signal.

[0115] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A hidden roller leisure chair for the elderly based on intelligent control algorithm, characterized in that: include: The control unit, which consists of an embedded microprocessor board, is responsible for the real-time calculation of the control system operation; The detection unit is used to detect various body indicators, body posture, and state of the user, and feed the detection data back to the control unit; The detection unit includes a posture detection module; The posture detection module includes: behavioral posture detection and sitting posture detection; Behavior posture detection uses three fitting functions to model the human body contour features. First, the minimum vertical rectangle of the human body area is obtained and the human body contour in the rectangular area is extracted to define the rectangular height as ; Then start from the highest point of the human body with If we use horizontal lines to intercept the human body contour, each horizontal line can intercept two points, one on the left and one on the right, and obtain the following two eigenvectors: ； ； in, Indicates The horizontal coordinate of the right intersection point of the horizontal line and the human body contour; Indicates The horizontal coordinate of the left intersection point of the horizontal line and the human body contour; and the vector is defined as: ； ； in Indicates The width of the human body at the horizontal line; Eigenvector Can be fitted into functions , and ; make For function All local minima The value set, then the position of the human head boundary is defined as: ； in, Indicates the position of the boundary line of the human head, that is, the set The minimum value in ; the shoulder position is defined as: ； in, Indicates the position of the shoulder boundary line, is the proportional coefficient, thus completing the segmentation and positioning of the human head and shoulders; The calculation of the human head posture is , , represents the area of ​​the skin color area on the left half of the human head, the area of ​​the skin color area on the right half of the human head, and the area of ​​the entire head area, respectively. The human head position variable Defined as: ； in, is a pre-set threshold. Indicates that the person is facing away from the camera. Indicates that the face is facing the camera. It means the face is facing the left side of the camera. It means the face is facing the right side of the camera; Calculation of human torso posture: First define the human torso posture variables The calculation method is as follows: ； in, It means that the human silhouette on the left is closer to the camera; It means that the human body silhouette on the right is closer to the camera; Indicates that the left and right contours are basically symmetrical in the shoulder area, and the human body is facing or facing away from the camera; The execution unit adjusts the equipment status and issues reminders according to the control unit instructions and the detection results of the detection unit; The human-computer interaction unit collects signals through touch perception, voice, and gestures and sends them to the control unit, which performs calculations and issues control instructions to the execution unit; The driving unit, when the elderly chair needs to be moved, the motor drives the wheels to automatically rise and fall to achieve movement. When the elderly chair is placed in a certain position, the wheels are automatically hidden through the control unit.

2. According to claim 1, a hidden roller type leisure chair for the elderly based on intelligent control algorithm is characterized in that: The posture detection module is used to automatically detect the height and sitting posture of different elderly people and send them to the control unit, and automatically adjust the inclination angle of the backrest according to the leaning and supine conditions of the elderly in the sitting state; The detection unit also includes: The sleep detection module collects sleep physiological signals through a microphone sensor and sends them to the control unit for judgment, and adjusts the seat posture according to the results; The health detection module uses multiple sensors to automatically detect the elderly's sitting posture, heart rate, and breathing rate when they sit down, and sends the collected data to the control unit to analyze the elderly's physical condition; The temperature detection module detects the seat temperature and ambient temperature through the thermostat and sends the signal data to the control unit to control the seat temperature according to the weather conditions.

3. The hidden roller type leisure chair for the elderly based on intelligent control algorithm according to claim 2 is characterized in that: The sitting posture detection is aimed at adjusting the sitting height and the tilt angle of the backrest. The height is automatically adjusted according to the height and sitting posture of different elderly people. The tilt angle of the backrest is automatically adjusted according to the leaning and supine conditions of the elderly when sitting, ensuring the comfort of the elderly sitting posture; The behavior posture detection collects the movements of the elderly on the chair and performs behavior recognition and motion analysis. First, the head and shoulders of the human body are located; then the head posture is calculated using the planar imaging characteristics of the human head, and the trunk posture is calculated using the contour change characteristics of the human shoulder; finally, the posture of the human body in motion is estimated by combining the postures of the head and trunk.

4. According to the hidden roller type leisure chair for the elderly based on intelligent control algorithm as claimed in claim 1, it is characterized in that: The specific steps of the sitting posture detection are as follows: The camera captures the image of the user on the elderly chair and outputs the coordinate information of the joint points of the person in the image to the control unit; Use convolutional neural networks to predict the confidence maps corresponding to body parts, and use geometric transformation kernels to realize the transfer of information between adjacent joints; The bidirectional graph structure information model is integrated into the convolutional neural network to achieve further information transmission; Finally, the joint information obtained by the geometric transformation kernel is superimposed with the joint point information obtained by the bidirectional graph structure information model to obtain the final specific positions of all joint points and complete the sitting posture detection.

5. The hidden roller type leisure chair for the elderly based on intelligent control algorithm according to claim 2 is characterized in that: The sleep detection module includes a microphone sensor and a sleep physiological signal acquisition module; The output end of the microphone sensor is connected to the input end of the sleep physiological signal acquisition module, and the output end of the sleep physiological signal acquisition module is wirelessly connected to the signal processing terminal; The sleep physiological signal acquisition module includes a signal filter, a signal amplifier, a processor, a memory and a wireless transmission module. The output end of the microphone sensor is connected to the input end of the signal filter, the output end of the signal filter is connected to the input end of the signal amplifier, and the output end of the signal filter is connected to the control unit. The signal filter includes a pulse signal filter, a breathing signal filter and a displacement signal filter.

6. The hidden roller type leisure chair for the elderly based on intelligent control algorithm according to claim 5 is characterized in that: The working method of the sleep detection module is as follows: The sleeping physiological signal is collected through the microphone sensor and transmitted to the signal filter; The noise of sleep physiological signals is filtered out through the signal filter and transmitted to the signal amplifier; The sleep physiological signal is amplified by a signal amplifier and transmitted to an analog-to-digital converter; The sleep physiological signal is an analog signal, which is converted into a digital signal through an analog-to-digital converter and saved in a memory; At regular intervals, the control unit reads data from the memory and processes it.

7. The hidden roller type leisure chair for the elderly based on intelligent control algorithm according to claim 1 is characterized in that: The driving unit is driven by hidden wheels. When the elderly chair needs to be moved, the control unit sends a control command to the driving unit through the human-computer interaction unit or gravity sensing, thereby achieving movement. Human-machine interaction is that the human-machine interaction unit sends control signals to the control unit through voice, gestures or buttons, and the control unit sends control instructions to the drive unit to achieve movement; Gravity sensing detects and determines the weight of the elderly chair through a weighing sensor. When the weight is higher than the set threshold, a control signal is sent to the control unit. The control unit sends a control command to the drive unit, the hidden wheels extend, and wait for the next command from the user.

Images