CN108836012A - A kind of soft or hard adjustable intelligent mattress and its control method with human muscle's state-detection - Google Patents
A kind of soft or hard adjustable intelligent mattress and its control method with human muscle's state-detection Download PDFInfo
- Publication number
- CN108836012A CN108836012A CN201810990599.6A CN201810990599A CN108836012A CN 108836012 A CN108836012 A CN 108836012A CN 201810990599 A CN201810990599 A CN 201810990599A CN 108836012 A CN108836012 A CN 108836012A
- Authority
- CN
- China
- Prior art keywords
- muscle
- human
- soft
- signal
- action
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47C—CHAIRS; SOFAS; BEDS
- A47C27/00—Spring, stuffed or fluid mattresses or cushions specially adapted for chairs, beds or sofas
- A47C27/08—Fluid mattresses or cushions
- A47C27/10—Fluid mattresses or cushions with two or more independently-fillable chambers
-
- A—HUMAN NECESSITIES
- A47—FURNITURE; DOMESTIC ARTICLES OR APPLIANCES; COFFEE MILLS; SPICE MILLS; SUCTION CLEANERS IN GENERAL
- A47C—CHAIRS; SOFAS; BEDS
- A47C31/00—Details or accessories for chairs, beds, or the like, not provided for in other groups of this subclass, e.g. upholstery fasteners, mattress protectors, stretching devices for mattress nets
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention discloses the soft or hard adjustable intelligent mattresses and its control method of a kind of human muscle's state-detection, pass through the particular functional layer of soft or hard adjustable intelligent mattress, human body surface muscle electric signal can be acquired in a manner of non-contact coupling, and muscle electric signal that is faint to this and mixing implements effective interference shielding, disappear mixed filtering, amplification and digital sample；The present invention extracts frequency domain character amount using the muscle electric signal, and then integrate the signal characteristic group of whole-body muscle movement, the determination of action mode type is realized using BP neural network, and then analyze sleep quality of human body, the human body deliberate action that accurate measurements can be difficult to for the prior art realizes that high reliability tests and analyzes, Whole Body type of action is judged more accurately, and then the adjusting of soft or hard adjustable intelligent mattress is made to match with sleep quality state, it plays significantly superior sleep quality and promotes effect.
Description
Technical field
The present invention relates to Smart Home technical field, more particularly to it is a kind of can with the soft or hard of human muscle's state-detection
Adjust intelligent mattress and its control method.
Background technique
Sleep is that everyone requires the important physiological activity carried out, and sleep can help human body to recover from fatigue, alleviate feelings
Thread, sufficient sleep are very necessary to the normal life of people.In modern society, the people with symptoms such as insomnia, somnolences does not exist
Minority usually causes very big influence to the life on daytime of people, leads to other mind & body problems.Meanwhile the health of human body
Situation will have a direct impact on and determine again many health problems such as sleep quality, such as myalgia, frequent micturition can all cause insomnia or
Person's sleep quality is low.
Mattress is the important tool of sleep, closely bound up with sleep quality.The mattress type that modern people use gradually becomes
To diversification, mainly have：Sping mattress, palm mattress, latex bed mattress, water bed mattress, magnetic mattress etc..In recent years, soft or hard adjustable intelligence
Energy mattress is once release, just by consumers.As its name suggests, the hardness energy of soft or hard adjustable intelligent mattress and human contact
Enough to adjust, therefore, it can adapt to everyone different physiological curves, sleeping position and habit, provide personalization for everyone
Comfortable sleeping environment.The structure of soft or hard adjustable intelligent mattress generally comprise soft or hard regulating course, comfortable contact layer and one or
Multiple functional layers；Wherein, soft or hard regulating course realizes the adjusting to the soft or hard degree of mattress using the principle for changing aeration quantity；Comfortably connect
Contact layer is in contact with human body, provides more comfortable contact sense of touch；And pressure state detection, temperature may be implemented in each functional layer
A variety of functions relevant to intelligence such as adjusting.
Human body body in sleep procedure is also constantly making various movements, also can be under unconscious after especially falling asleep
Generate a series of actions.Wherein, the frequency and amplitude acted under sleeping state is all bigger, and under deep sleep
The occurrence frequency and amplitude of movement can be relatively small.Meanwhile it is excessively high in temperature or too low cause that body-sensing is uncomfortable, suppresses urine, body
When posture is not comfortable enough, human body also can stress and act.Duration prison is carried out to human action under sleep state
It surveys, it will be appreciated that the Depth of sleep of user rule the reason of analyzing influence user's sleep quality, and gives personalized solution party
Case, including to the soft or hard regulating course of intelligent mattress and self adaptive control of functional layer etc..
Currently, pressure can be arranged in the functional layer of soft or hard adjustable intelligent mattress to the monitoring acted under sleep quality state
Sensor measures pressure change caused by human action, the width of the amplitude and frequency representative human action that are changed by pressure value
Degree and frequency.But the human action monitoring accuracy based on pressure measurement is lower, to the monitoring precision of human body deliberate action and
Real-time is all inadequate, is also difficult to effectively be judged to human action position and manner of execution.
The various movements of human body are all to be drawn to realize by muscle, and muscle movement can generate subtle telecommunications in human body surface
Number, and as different, movement amplitudes and frequency are different with mode at the position of human action, which also can be therewith
Generate otherness.Also, intelligent mattress is close with human contact in sleep procedure and contact area is big, for mentioning for muscle electric signal
It takes and provides convenience, if it is possible to be monitored using the electric signal to human muscle's state in sleep procedure, Jin Erfen
Sleep quality of human body is analysed, significantly superior technical effect can be played compared with prior art.However, muscle electric signal is more faint
And it is mingled with much noise, the difficulty that extraction, the identification of the electric signal are all solved in the presence of needs.
Summary of the invention
In view of this, the present invention provides the soft or hard adjustable intelligent mattress and its controlling party of a kind of human muscle's state-detection
Method.
The present invention provides a kind of soft or hard adjustable intelligent mattress of human muscle's state-detection, which is characterized in that including：It is soft or hard
Regulating course, comfortable contact layer and at least one functional layer；The soft or hard regulating course includes that several can be independent inflatable and to put
The inflating cells of gas, by the inflated for deflated change chamber interior atmospheric pressure value of each inflating cells, to adjust the chamber pair
Answer the hardness in mattress region；Comfortable contact layer is arranged on soft or hard regulating course, is directly in contact with human body；The functional layer
It is embedded within the comfortable contact layer or lays the surface layer with comfortable contact layer, each functional layer includes several functions
The flexible wire of unit and each functional unit of connection；Wherein, at least one described functional layer includes muscular states detection function
Ergosphere, the functional unit of the muscular states detection function layer include coupled capacitor patch electrode, interference shielding shell, signal amplification
Module, disappear mixed filter module, analog-to-digital conversion module；The coupled capacitor patch electrode to be formed by coupling with human muscle surface
Capacitor, the potential difference signal that induction muscle movement generates, and the potential difference signal is transmitted to the signal amplification module；
The potential difference signal that the signal amplification module is used to generate muscle movement amplifies, and amplified potential difference is believed
Number it is transmitted to the mixed filter module that disappears；The mixed filter module that disappears is filtered amplified potential difference signal, and elimination mixes letter
Number, it is then sent to the analog-to-digital conversion module；The analog-to-digital conversion module carries out analog-to-digital conversion, and generating indicates muscle movement
Digital signal, and the output of the flexible wire by connecting each functional unit；The interference shielding shell is used for dry to the external world
Signal is disturbed to be shielded；
The soft or hard adjustable intelligent mattress further includes：Human action recognition unit, the human action recognition unit are used for
The digital signal of the expression muscle movement of each functional unit output of muscular states detection function layer is received, and according to whole
The digital signal identifies human action mode；
Sleep state analytical unit, for analyzing sleep quality state according to human action mode；
Mattress adjusts unit, the soft or hard adjusting for being adapted to the sleep quality state, to soft or hard adjustable intelligent mattress
Layer and other functional layers are adjusted.
Preferably, the human action recognition unit specifically includes：Muscle movement signal characteristic abstraction module, for from
Characteristic quantity is extracted in the digital signal of the expression muscle movement of each functional unit output, whole characteristic quantities is integrated and is formed
Whole-body muscle action signal feature group；Action mode Classification Neural, using record whole-body muscle action signal feature group with
It is special according to the whole-body muscle action signal inputted in real time after the training sample of human action mode type corresponding relationship is trained
Sign group judges current human action mode type.
It may further be preferable that the muscle movement signal characteristic abstraction module is used for the institute exported to each functional unit
Stating indicates that the digital signal of muscle movement calculates characteristic quantity as follows：
Wherein XPiIndicate the characteristic quantity of the digital signal of the expression muscle movement of i-th of functional unit output, Pi(f) it indicates
The power spectrum of the digital signal of the expression muscle movement of i-th of functional unit output；Wherein the value range of i is 1-n；Also,
The characteristic quantity of the muscle movement signal characteristic abstraction block combiner all functional units generates whole-body muscle action signal
Feature group XP={ XP1, XP2…XPi…XPn}。
It may further be preferable that the action mode Classification Neural includes input layer, hidden layer and output layer；Input
Layer corresponds to the whole-body muscle action signal feature group of input；The hidden layer is used for by being instructed with the training sample
Practice, obtains the corresponding relationship between whole-body muscle action signal feature group and human action mode type, and then according to defeated in real time
The whole-body muscle action signal feature group entered determines human action mode type；The output layer is used for human action mode class
Type is exported.
It may further be preferable that the action mode Classification Neural utilizes training sample, whole-body muscle is acted and is believed
The sample and corresponding human action mode type of number feature group are input to BP neural network as learning sample；Execute forward direction
Conduction calculates：The learning sample is updated to BP neural network, successively calculates the numerical value of hidden layer and output layer；Judge epicycle
Whether deviation, which is less than or equal to scheduled tolerance ε, if the determination result is YES then stops iteration, if judging result be it is no, after
It is continuous to execute retrospectively calculate, weight is changed according to learning rate；By repetition learning, between the neuron for constantly adjusting BP neural network
Weighted value, until deviation is less than equal to scheduled tolerance, then BP neural network training is completed；For current real-time
The whole-body muscle action signal feature group of acquisition, is input to the trained BP neural network as an input vector,
The BP neural network is set to export current human action mode type.
Preferably, the sleep state analytical unit is according to the human action mode type in scheduled statistics duration
Distribution situation calculates its degree of deviation being distributed with human body action mode type under ideal sleep state.
It may further be preferable that the sleep state analytical unit is according to every kind of human action mode type in the statistics
The difference of frequency under frequency and ideal sleep state in duration, in conjunction with the normalization power of every kind of human action mode
Weight parameter, calculates the degree of deviation.
It may further be preferable that the mattress adjusts unit according to the degree of deviation, determine to soft or hard regulating course and its
The adjustment parameter that its functional layer is adjusted.
Preferably, the interference shielding shell successively includes insulating layer, metal screen layer and silica gel outer cover from the inside to the outside
Layer.
Preferably, other functional layers include temperature regulation layer, and the mattress adjust unit be adapted to it is described
The temperature of the temperature regulation layer is adjusted in sleep quality state.
In turn, the soft or hard adjustable intelligent mattress control method based on human muscle's state-detection that the present invention provides a kind of,
It is characterized by comprising the following steps：
The coupled capacitor patch electrode that functional unit using soft or hard adjustable intelligent mattress functional layer includes, by with human body
Muscle surface couples to form capacitor, the potential difference signal that induction muscle movement generates；
The potential difference signal generated to muscle movement amplifies, and amplified potential difference signal is filtered,
Clutter is eliminated, analog-to-digital conversion is then carried out, generates the digital signal for indicating muscle movement；
The digital signal of the expression muscle movement of each functional unit output is received, and according to all digital signals
Human action mode is identified；
For analyzing sleep quality state according to human action mode；
Be adapted to the sleep quality state, the soft or hard regulating course and other functional layers to soft or hard adjustable intelligent mattress into
Row is adjusted.
Preferably, identification is carried out to human action mode to specifically include：The expression exported from each functional unit
Characteristic quantity is extracted in the digital signal of muscle movement, is integrated whole characteristic quantities and is formed whole-body muscle action signal feature group；Benefit
With the training sample of record whole-body muscle action signal feature group and human action mode type corresponding relationship to BP neural network
After being trained, according to the whole-body muscle action signal feature group inputted in real time, current human action mode type is judged.
It may further be preferable that being calculated as follows the digital signal of the expression muscle movement of each functional unit output
Characteristic quantity：
Wherein XPiIndicate the characteristic quantity of the digital signal of the expression muscle movement of i-th of functional unit output, Pi(f) it indicates
The power spectrum of the digital signal of the expression muscle movement of i-th of functional unit output；Wherein the value range of i is 1-n；Also,
The characteristic quantity for combining all functional units generates whole-body muscle action signal feature group XP={ XP1, XP2…XPi…XPn}。
It may further be preferable that by the sample of whole-body muscle action signal feature group and corresponding human action mode class
Type is input to BP neural network as learning sample；Forward conduction is executed to calculate：The learning sample is updated to BP neural network,
Successively calculate the numerical value of hidden layer and output layer；Judge whether the deviation of epicycle is less than or equal to scheduled tolerance ε, if judgement
As a result be it is yes, then stop iteration, if judging result be it is no, continue to execute retrospectively calculate, according to learning rate change weight；By
Repetition learning constantly adjusts the weighted value between the neuron of BP neural network, permits until deviation is less than equal to scheduled
Perhaps deviation, then BP neural network training are completed；For the whole-body muscle action signal feature group currently acquired in real time, as
One input vector is input to the trained BP neural network, and the BP neural network is made to export current human action mode class
Type.
Preferably, for analyzing sleep quality state and specifically including according to human action mode：According in scheduled system
The distribution situation of human action mode type, calculates it and is distributed with human body action mode type under ideal sleep state in timing is long
The degree of deviation.
It may further be preferable that frequency and reason according to every kind of human action mode type in the statistics duration
The difference for thinking frequency under sleep state calculates the deviation in conjunction with the normalized weight parameter of every kind of human action mode
Degree.
It may further be preferable that tool is adjusted in the soft or hard regulating course and other functional layers to soft or hard adjustable intelligent mattress
Body includes：According to the degree of deviation, the adjustment parameter that soft or hard regulating course and other functional layers are adjusted is determined.
As it can be seen that the application in sleep procedure by the particular functional layer of soft or hard adjustable intelligent mattress, can be with non-contact
The mode of coupling acquires human body surface muscle electric signal, and muscle electric signal that is faint to this and mixing implements effective interference
It shields, disappear mixed filtering, amplification and digital sample；The present invention extracts frequency domain character amount using the muscle electric signal, and then integrates
The signal characteristic group of whole-body muscle movement is realized the determination of action mode type using BP neural network, and then analyzes human body sleeping
Dormancy quality carries out the side of human action detection and sleep state estimating using pressure change or air pressure sensing compared with prior art
Formula, the human body deliberate action that accurate measurements can be difficult to for the prior art realize that high reliability tests and analyzes, more accurately
Judge Whole Body type of action, and then the adjusting of soft or hard adjustable intelligent mattress is made to match with sleep quality state, plays bright
Aobvious more preferably sleep quality promotes effect.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention,
And it can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can
It is clearer and more comprehensible, the followings are specific embodiments of the present invention.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field
Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention
Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings：
Fig. 1 shows the overall frame of the soft or hard adjustable bed mattess of human muscle's status detection function provided in an embodiment of the present invention
Structure schematic diagram；
Fig. 2 shows soft or hard adjustable bed mattess body construction schematic diagrames provided in an embodiment of the present invention；
Fig. 3 shows soft or hard adjustable bed mattess functional layer detection network schematic diagram provided in an embodiment of the present invention；
Fig. 4 shows the muscular states detection function layer function unit knot of soft or hard adjustable bed mattess provided in an embodiment of the present invention
Structure schematic diagram；
Fig. 5 shows the circuit modular structure of the functional unit of muscular states detection function layer provided in an embodiment of the present invention
Figure；
Fig. 6 shows human action recognition unit function structure chart provided in an embodiment of the present invention.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing
Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here
It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure
It is fully disclosed to those skilled in the art.
The embodiment of the invention provides a kind of soft or hard adjustable intelligent mattress with human muscle's status detection function.Referring to
Fig. 1 shows the general frame schematic diagram of the system, and the system comprises soft or hard adjustable bed mattess ontology 1, human action identification are single
Member 2, sleep state analytical unit 3 and mattress adjust unit 4.
The soft or hard adjustable bed mattess ontology 1 is from its structure, as shown in Fig. 2, including soft or hard regulating course 101, comfortably connecing
Contact layer 102 and one or more functional layer 103.Wherein, soft or hard regulating course 101 is using the principle realization pair for changing aeration quantity
The adjusting of the soft or hard degree of mattress.Specifically, soft or hard regulating course be divided into several can be with the inflation of independent inflatable and deflation
Chamber 101A, each inflating cells 101A is installed into outlet valve, and the valve is connected to by gas-guide tube with inflator pump, is passed through
Its air pressure inside can be increased to inflating cells 101A inflation, so that the increase of mattress hardness is shown as, and to inflating cells 101A
Deflation can reduce its air pressure inside, to show as the reduction of mattress hardness；Since each chamber 101A can be carried out independently firmly
Degree is adjusted, so that the different location of entire mattress be made to can have different hardness, therefore can adapt to the physiological curve of human body
With feeling preference.Comfortable contact layer 102 is arranged on soft or hard regulating course 101, including latex, filling cotton, surface pro-skin fabric
Equal multilayered structures, the layer are in contact with human body, provide more comfortable contact sense of touch.One or more functional layer 103 can be with
It is embedded within the comfortable contact layer, each functional layer 103 includes functional unit 103A and each functional unit of connection
The flexible wire 103B of 103A.As shown in figure 3, the functional layer 103 of soft or hard adjustable bed mattess ontology 1 is divided into several matrixes point
The detection grid 103C of cloth, and a functional unit 103A is set in each detection grid 103C.The flexible wire 103B
For linkage function unit 103A, the input and output of functional unit 103A signal are realized.For example, the functional layer 103 can be with
It is temp regulating function layer, the functional unit 103A of the functional layer is electric heater, the temperature transmitted according to flexible wire 103B
Adjustment signal, functional unit 103A adjust the heating temperature of its own, to build and the suitable sleep temperature of user's body-sensing
Environment.
Also, the functional layer 103 of the soft or hard adjustable bed mattess ontology includes at least a muscular states detection function layer, institute
Surface layer with comfortable contact layer can be laid by stating functional layer.The structure of the functional unit 103A of the muscular states detection function layer
As shown in figure 4, including coupled capacitor patch electrode 103A1, interference shielding shell 103A2 and circuit board 103A3.Wherein, such as Fig. 5
Shown, setting signal amplification module 103A4 on circuit board 103A3, disappear mixed filter module 103A5, analog-to-digital conversion module 103A6.
The coupled capacitor patch electrode 103A1 with human muscle surface by coupling to form capacitor, the electricity of induction muscle movement generation
Potentiometer signal, and the potential difference signal is transmitted to the signal amplification module 103A4；Coupled capacitor patch electrode
103A1 does not need to directly fit with human epidermal skin, can be coupled across clothing and human body surface and form capacitance body, from
And incude human body surface electric signal.It is ultra-weak electronic signal that human muscle, which moves the electric signal generated, and mixes various noise telecommunications
Number, vulnerable to interference, signal-to-noise ratio is relatively low, and frequency range is generally between 50Hz-2KHz, signal amplitude 0-2mV.The letter
The potential difference signal that number amplification module 103A4 is used to generate muscle movement amplifies, and amplified potential difference is believed
Number it is transmitted to the mixed filter module 103A5 that disappears.The mixed filter module 103A5 that disappears is using low-pass filter, and cutoff frequency is in 2KHz
Left and right, is filtered amplified potential difference signal, eliminates clutter, is then sent to the analog-to-digital conversion module
103A5.The analog-to-digital conversion module 103A5 uses high-precision AD converter, carries out modulus to filtered potential difference signal and turns
Change, generate the digital signal for indicating muscle movement, and by the flexible wire 103B of each functional unit of connection export to
The human action recognition unit 2.The interference shielding shell 103A2 guarantees muscle for shielding to extraneous interference signal
Movement generate electric signal accuracy, the interference shielding shell 103A2 from the inside to the outside successively include insulating layer, metal screen layer with
And silica gel enclosing cover layer.
Human action recognition unit 2 is used to receive the table of each functional unit 103A output of muscular states detection function layer
Show the digital signal of muscle movement, and human action mode is identified according to all digital signals.Such as Fig. 6 institute
Show, human body action recognition unit 2 specifically includes：Muscle movement signal characteristic abstraction module 201 and action mode classification mind
Through network 202.
The expression muscle that muscle movement signal characteristic abstraction module 201 is used to export from each functional unit 103A is dynamic
Characteristic quantity is extracted in the digital signal of work；Equipped with n functional unit 103A, the muscle movement signal characteristic abstraction module
The digital signal of the expression muscle movement of 201 pairs of each functional unit output calculates characteristic quantity as follows：
Wherein XPiIndicate the characteristic quantity of the digital signal of the expression muscle movement of i-th of functional unit output, Pi(f) it indicates
The power spectrum of the digital signal of the expression muscle movement of i-th of functional unit output；Wherein the value range of i is 1-n.In turn,
Muscle movement signal characteristic abstraction module 201 integrates the characteristic quantity of whole n functional units, and it is special to form whole-body muscle action signal
Sign group XP={ XP1, XP2…XPi…XPn}。
Action mode Classification Neural 202 is a BP neural network, which includes input layer, hidden layer
And output layer；Input layer corresponds to the whole-body muscle action signal feature group X of inputP；The hidden layer is used for by with described
Training sample is trained, and obtains the corresponding relationship between whole-body muscle action signal feature group and human action mode type,
And then according to the whole-body muscle action signal feature group X inputted in real timePDetermine human action mode type；The output layer is used for
Human action mode type is exported.The training sample has recorded whole-body muscle action signal feature group and human action
Mode type corresponding relationship, for example, in the mattress design stage various human actions can be executed by tester by being surveyed
Mode type, such as by prostrate become to lie on one's side, become that right side is sleeping, the systemic human action such as becomes to lie down by lying on one's side by lying on the left side
The fine motions people such as is slightly trembleed in type, the upper limbs human action type such as lower limb human action type, swing arm such as kicking, leg
Body type of action etc., and whole-body muscle action signal feature group caused by every kind of human action mode type is recorded, from
And the corresponding relationship item of a human body action mode type and whole-body muscle action signal feature group is generated, pass through accumulative number enough
The corresponding relationship item of amount, such as 10000, form the training sample.In turn, by the whole body in the training sample of the accumulation
Muscle movement signal characteristic group inputs BP neural network, to the weight and deviation progress between each neuron of the BP neural network
Adjusting training repeatedly, the human action mode of the human action mode type for exporting the BP neural network and training sample record
Type is as close possible to when degree of closeness reaches expected, then it is assumed that algorithm training is completed, and corresponding weight and deviation are saved.
The BP neural network after training can be according to the whole-body muscle action signal feature group X acquired in real timeP={ XP1, XP2…
XPi…XPn, export corresponding human action mode type.More specifically, training action pattern classification neural network 202 is simultaneously
The process for carrying out the identification of human action mode type includes the following steps：
(1) training sample data are utilized, the sample of whole-body muscle action signal feature group and corresponding human body are moved
Operation mode type is input to the BP neural network as learning sample；Wherein, the whole-body muscle action signal feature group X of inputp
={ xp1...xpN, wherein xp1,xp2,……xpNAs a dimension of input vector, i-th of functional unit is corresponded respectively to
The characteristic quantity of the digital signal of the expression muscle movement of output；Desired output corresponding with the feature group of the input is tpm, should
Desired output is human action mode type corresponding with this feature group, t in samplepmDifferent values represent human action
Mode type.
(2) forward conduction is executed to calculate：BP neural network input layer has N number of input neuron, and hidden layer has K a hidden
Layer neuron, output layer have M output neuron, then successively calculate hidden layer and the numerical value of output layer is as follows：
Wherein w1nkIt is the weight between n-th of neuron of input layer and k-th of neuron of hidden layer, O1pkIt is hidden
Hide the output of k-th of neuron of layer；w2kmIt is the weight between m-th of neuron of k-th of neuron of hidden layer and output layer,
O2pmIt is the output of m-th of output layer neuron, activation primitiveI indicates the i-th wheel training；
(3) implementation deviation calculates：Judge whether the deviation of epicycle (the i-th wheel) is less than or equal to
Scheduled tolerance ε if the determination result is YES then stops iteration, if judging result be it is no, continue following process；
(4) retrospectively calculate is executed：
Wherein learning rate is μ,
-δpm(i)=(tpm-O2pm(i))O2pm(i)(1-O2pm(i)),
It is as follows to change weight：
w1nk(i+1)=w1nk(i)+Δw1nk(i+1)
w2km(i+1)=w2km(i)+Δw2km(i+1)
(5) (2) step is returned, the study of i+1 wheel is re-started.
By repetition learning, the weighted value between neuron is constantly adjusted, is permitted until deviation is less than equal to scheduled
Perhaps deviation ε, then BP neural network training are completed.To for the whole-body muscle action signal feature group X currently acquired in real timeP
={ XP1, XP2...XPi...XPn, it can be input to trained BP neural network as an input vector, by the mind
Output through the network human action mode type current as user.
Therefore, it is possible to which a detection time point is arranged in the unit time (such as 5 seconds) every agreement, examined by muscular states
Each functional unit 103A of measurement of power ergosphere acquires and exports the primary digital signal for indicating muscle movement in the detection time point,
And then the human action mode type of the detection time point is identified by human action recognition unit 2.
The reception of sleep state analytical unit 3 is moved in each detection time point by the human body that human action recognition unit 2 exports
Operation mode type；In turn, the statistics of sleep state analytical unit 3 is in scheduled statistics duration (such as apart from current time 5 minutes
It is interior) in every kind of human action mode type frequency count value, so that it is determined that the distribution feelings of human action mode type
Condition.For example, systemic human action Class1, lower limb human action Class1 5 times, upper limb occurs in statistics duration
Human action type 2 times, fine motion human action type 38 times.Sleep state analytical unit 3 calculates the distribution situation and ideal
The degree of deviation that human body action mode type is distributed under sleep state；It is, the sleep state analytical unit 3 is according to every kind of people
The difference of body action mode type frequency under the frequency and ideal sleep state in the statistics duration, in conjunction with every
The normalized weight parameter of kind human action mode, calculates the degree of deviation.For example, systemic people under default ideal sleep state
Body type of action 0 time, lower limb human action type 5 times, upper limb human action type 5 times, fine motion human action type
10 times, then various human action mode types are in the frequency in the statistics duration and frequency under ideal sleep state
Difference be respectively systemic human action type+1, lower limb human action type+10, upper limb human action type -3,
Fine motion human action type+28, by the difference further combined with the normalized weight parameter of every kind of human action mode type,
Such as the weight parameter of systemic human action type is 10, lower limb human action type and upper limb human action type
Weight parameter is 5, and the weight parameter of fine motion human action type is 1, then the degree of deviation calculated be (+1*10)+(+10*5)+
(- 3*5)+(28*1)=73.
The mattress adjusts unit 4 according to the degree of deviation, and soft or hard regulating course and other functional layers are adjusted in determination
The adjustment parameter of section, and then the adjusting to soft or hard regulating course and other functional layers is executed according to adjustment parameter.For example, above-mentioned deviation
Angle value is higher, then illustrates that the action frequency in user's sleep procedure is more, amplitude is bigger, show user's Depth of sleep state not
Good, then the aeration quantity that mattress adjusting unit 4 can reduce soft or hard regulating course is conducive to user so that mattress is more soft
Into deep sleep, and the reduction amplitude of aeration quantity then direct ratio degree of deviation, that is, the degree of deviation are higher, then aeration quantity
Reduction amplitude is bigger, and mattress also just becomes more soft；, whereas if the degree of deviation is lower, then the reduction amplitude of aeration quantity is smaller；
If the degree of deviation is 0 or negative value, adjusting can not be executed to the aeration quantity of soft or hard regulating course.Other functional layer packets
Temperature regulation layer is included, and mattress adjusting unit can be according to the degree of deviation, to the temperature value of the temperature regulation layer
It is adjusted, equally, deviation angle value is higher, then illustrates that user's Depth of sleep is not in good state, then bigger to the knots modification of temperature；Instead
It, the degree of deviation is lower, then the knots modification of temperature is smaller；If the degree of deviation is 0 or negative value, tune can not be executed to temperature
Section.
As it can be seen that the application in sleep procedure by the particular functional layer of soft or hard adjustable intelligent mattress, can be with non-contact
The mode of coupling acquires human body surface muscle electric signal, and muscle electric signal that is faint to this and mixing implements effective interference
It shields, disappear mixed filtering, amplification and digital sample；The present invention extracts frequency domain character amount using the muscle electric signal, and then integrates
The signal characteristic group of whole-body muscle movement is realized the determination of action mode type using BP neural network, and then analyzes human body sleeping
Dormancy quality carries out the side of human action detection and sleep state estimating using pressure change or air pressure sensing compared with prior art
Formula, the human body deliberate action that accurate measurements can be difficult to for the prior art realize that high reliability tests and analyzes, more accurately
Judge Whole Body type of action, and then the adjusting of soft or hard adjustable intelligent mattress is made to match with sleep quality state, plays bright
Aobvious more preferably sleep quality promotes effect.
Various component embodiments of the invention can be implemented in hardware, or to run on one or more processors
Software module realize, or be implemented in a combination thereof.It will be understood by those of skill in the art that can be used in practice
Microprocessor or digital signal processor (DSP) are realized.The present invention is also implemented as described herein for executing
Some or all device or device programs (for example, computer program and computer program product) of method.In this way
Realization program of the invention can store on a computer-readable medium, or can have the shape of one or more signal
Formula.Such signal can be downloaded from an internet website to obtain, and perhaps be provided on the carrier signal or with any other shape
Formula provides.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and ability
Field technique personnel can be designed alternative embodiment without departing from the scope of the appended claims.In the claims,
Any reference symbol between parentheses should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not
Element or step listed in the claims.Word "a" or "an" located in front of the element does not exclude the presence of multiple such
Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real
It is existing.In the unit claims listing several devices, several in these devices can be through the same hardware branch
To embody.The use of word first, second, and third does not indicate any sequence.These words can be explained and be run after fame
Claim.
Claims (10)
1. a kind of soft or hard adjustable intelligent mattress of human muscle's state-detection, which is characterized in that including：It is soft or hard regulating course, comfortable
Contact layer and at least one functional layer；The soft or hard regulating course includes that several can be with the inflatable chamber of independent inflatable and deflation
Room, by the inflated for deflated change chamber interior atmospheric pressure value of each inflating cells, so that adjusting the chamber corresponds to mattress region
Hardness；Comfortable contact layer is arranged on soft or hard regulating course, is directly in contact with human body；The functional layer is embedded into this and relaxes
Within the suitable contact layer or surface layer of laying and comfortable contact layer, each functional layer includes several functional units and company
Connect the flexible wire of each functional unit；Wherein, at least one described functional layer includes muscular states detection function layer, the flesh
The functional unit of meat status detection function layer includes coupled capacitor patch electrode, interference shielding shell, signal amplification module, disappear mixed filter
Wave module, analog-to-digital conversion module；The coupled capacitor patch electrode incudes flesh by coupling to form capacitor with human muscle surface
The potential difference signal that meat movement generates, and the potential difference signal is transmitted to the signal amplification module；The signal is put
The potential difference signal that big module is used to generate muscle movement amplifies, and amplified potential difference signal is transmitted to and is disappeared
Mixed filter module；The mixed filter module that disappears is filtered amplified potential difference signal, eliminates clutter, then conveys
To the analog-to-digital conversion module；The analog-to-digital conversion module carries out analog-to-digital conversion, generates the digital signal for indicating muscle movement, and
By the flexible wire output for connecting each functional unit；The interference shielding shell is for shielding extraneous interference signal
It covers；
The soft or hard adjustable intelligent mattress further includes：Human action recognition unit, the human action recognition unit is for receiving
The digital signal of the expression muscle movement of each functional unit output of muscular states detection function layer, and according to whole
Digital signal identifies human action mode；
Sleep state analytical unit, for analyzing sleep quality state according to human action mode；
Mattress adjusts unit, for being adapted to the sleep quality state, to the soft or hard regulating course of soft or hard adjustable intelligent mattress with
And other functional layers are adjusted.
2. the soft or hard adjustable intelligent mattress of human muscle's state-detection according to claim 1, which is characterized in that the people
Body action recognition unit specifically includes：Muscle movement signal characteristic abstraction module, for from described in the output of each functional unit
It indicates to extract characteristic quantity in the digital signal of muscle movement, integrates whole characteristic quantities and form whole-body muscle action signal feature
Group；Action mode Classification Neural, it is corresponding with human action mode type using record whole-body muscle action signal feature group
After the training sample of relationship is trained, according to the whole-body muscle action signal feature group inputted in real time, current human body is judged
Action mode type.
3. the soft or hard adjustable intelligent mattress of human muscle's state-detection according to claim 2, which is characterized in that the flesh
The digital signal for the expression muscle movement that meat action signal characteristic extracting module is used to export each functional unit is as follows
Calculate characteristic quantity：
Wherein XPiIndicate the characteristic quantity of the digital signal of the expression muscle movement of i-th of functional unit output, Pi(f) i-th is indicated
The power spectrum of the digital signal of the expression muscle movement of a functional unit output；Wherein the value range of i is 1-n；Also, it is described
The characteristic quantity of muscle movement signal characteristic abstraction block combiner all functional units generates whole-body muscle action signal feature
Group XP={ XP1, XP2…XPi…XPn}。
4. the soft or hard adjustable intelligent mattress of human muscle's state-detection according to claim 3, which is characterized in that described dynamic
Operation mode Classification Neural includes input layer, hidden layer and output layer；The whole-body muscle that input layer corresponds to input acts letter
Number feature group；The hidden layer is used to obtain whole-body muscle action signal feature group by being trained with the training sample
With the corresponding relationship between human action mode type, and then determined according to the whole-body muscle action signal feature group that inputs in real time
Human action mode type；The output layer is for exporting human action mode type.
5. the soft or hard adjustable intelligent mattress of human muscle's state-detection according to claim 4, which is characterized in that described to sleep
Dormancy state analysis unit calculates it and sleeps with ideal according to the distribution situation of the human action mode type in scheduled statistics duration
The degree of deviation that human body action mode type is distributed under dormancy state.
6. the soft or hard adjustable intelligent mattress of human muscle's state-detection according to claim 5, which is characterized in that the bed
Pad adjusts unit according to the degree of deviation, determines the adjustment parameter that soft or hard regulating course and other functional layers are adjusted.
7. a kind of soft or hard adjustable intelligent mattress control method based on human muscle's state-detection, which is characterized in that including following
Step：
The coupled capacitor patch electrode that functional unit using soft or hard adjustable intelligent mattress functional layer includes, by with human muscle
Surface couples to form capacitor, the potential difference signal that induction muscle movement generates；
The potential difference signal generated to muscle movement amplifies, and amplified potential difference signal is filtered, and eliminates
Then clutter carries out analog-to-digital conversion, generate the digital signal for indicating muscle movement；
The digital signal of the expression muscle movement of each functional unit output is received, and according to all digital signals to people
Body action mode is identified；
For analyzing sleep quality state according to human action mode；
It is adapted to the sleep quality state, the soft or hard regulating course and other functional layers of soft or hard adjustable intelligent mattress are adjusted
Section.
8. the soft or hard adjustable intelligent mattress control method according to claim 7 based on human muscle's state-detection, special
Sign is, carries out identification to human action mode and specifically includes：The expression muscle movement exported from each functional unit
Characteristic quantity is extracted in digital signal, is integrated whole characteristic quantities and is formed whole-body muscle action signal feature group；Utilize record whole body
After the training sample of muscle movement signal characteristic group and human action mode type corresponding relationship is trained BP neural network,
According to the whole-body muscle action signal feature group inputted in real time, current human action mode type is judged.
9. the soft or hard adjustable intelligent mattress control method according to claim 8 based on human muscle's state-detection, special
Sign is, calculates characteristic quantity as follows to the digital signal of the expression muscle movement of each functional unit output：
Wherein XPiIndicate the characteristic quantity of the digital signal of the expression muscle movement of i-th of functional unit output, Pi(f) i-th is indicated
The power spectrum of the digital signal of the expression muscle movement of a functional unit output；Wherein the value range of i is 1-n；Also, it combines
The characteristic quantity of all functional units generates whole-body muscle action signal feature group XP={ XP1, XP2…XPi…XPn}。
10. the soft or hard adjustable intelligent mattress control method according to claim 9 based on human muscle's state-detection, special
Sign is, for analyzing sleep quality state and specifically including according to human action mode：According to the people in scheduled statistics duration
The distribution situation of body action mode type calculates its degree of deviation being distributed with human body action mode type under ideal sleep state.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810990599.6A CN108836012B (en) | 2018-08-28 | 2018-08-28 | Hardness-adjustable intelligent mattress with human muscle state detection function and control method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810990599.6A CN108836012B (en) | 2018-08-28 | 2018-08-28 | Hardness-adjustable intelligent mattress with human muscle state detection function and control method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108836012A true CN108836012A (en) | 2018-11-20 |
| CN108836012B CN108836012B (en) | 2021-09-07 |
Family
ID=64189738
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810990599.6A Active CN108836012B (en) | 2018-08-28 | 2018-08-28 | Hardness-adjustable intelligent mattress with human muscle state detection function and control method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108836012B (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102247063A (en) * | 2010-05-21 | 2011-11-23 | 许汉忠 | Active supporting device of healthy bed and method thereof |
| US20120096648A1 (en) * | 2010-10-25 | 2012-04-26 | Cheng-Yang Chan | Air Tool With Modularized Air Pad |
| CN102458239A (en) * | 2009-06-05 | 2012-05-16 | 皇家飞利浦电子股份有限公司 | Capacitive sensing system |
| WO2014057029A1 (en) * | 2012-10-10 | 2014-04-17 | Kuster Raffael | Movement cushion |
| CN104257159A (en) * | 2014-09-05 | 2015-01-07 | 杨松 | Method for acquiring human-body physiological signals, spring mattress and system |
| CN204218446U (en) * | 2014-09-05 | 2015-03-25 | 杨松 | Gather mat and the system of physiology signal |
| CN104732250A (en) * | 2015-03-25 | 2015-06-24 | 安徽农业大学 | Non-sleep-interference sleeping posture and sleeping behavior testing identification method |
| CN105310681A (en) * | 2014-06-23 | 2016-02-10 | 三星电子株式会社 | Bioelectrode, and method and apparatus for processing biosignal using the same |
| CN106068097A (en) * | 2014-02-20 | 2016-11-02 | 佛吉亚汽车座椅有限责任公司 | Integrate the seat having sensor |
| CN106388819A (en) * | 2016-10-28 | 2017-02-15 | 许昌学院 | Human upper limb muscle state monitoring system based on surface electromyogram signals and judging method thereof |
| CN106667435A (en) * | 2016-12-17 | 2017-05-17 | 复旦大学 | Intelligent sensing mattress for monitoring sleep |
| CN107126005A (en) * | 2017-07-03 | 2017-09-05 | 广东葆康信息科技有限公司 | Biomethanics mattress and system |
| CN108420228A (en) * | 2018-04-13 | 2018-08-21 | 浙江想能云软件股份有限公司 | A kind of soft or hard adjustable bed mattess of intelligence and its monitoring method of sleep state monitoring |
-
2018
- 2018-08-28 CN CN201810990599.6A patent/CN108836012B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102458239A (en) * | 2009-06-05 | 2012-05-16 | 皇家飞利浦电子股份有限公司 | Capacitive sensing system |
| CN102247063A (en) * | 2010-05-21 | 2011-11-23 | 许汉忠 | Active supporting device of healthy bed and method thereof |
| US20120096648A1 (en) * | 2010-10-25 | 2012-04-26 | Cheng-Yang Chan | Air Tool With Modularized Air Pad |
| WO2014057029A1 (en) * | 2012-10-10 | 2014-04-17 | Kuster Raffael | Movement cushion |
| CN106068097A (en) * | 2014-02-20 | 2016-11-02 | 佛吉亚汽车座椅有限责任公司 | Integrate the seat having sensor |
| CN105310681A (en) * | 2014-06-23 | 2016-02-10 | 三星电子株式会社 | Bioelectrode, and method and apparatus for processing biosignal using the same |
| CN104257159A (en) * | 2014-09-05 | 2015-01-07 | 杨松 | Method for acquiring human-body physiological signals, spring mattress and system |
| CN204218446U (en) * | 2014-09-05 | 2015-03-25 | 杨松 | Gather mat and the system of physiology signal |
| CN104732250A (en) * | 2015-03-25 | 2015-06-24 | 安徽农业大学 | Non-sleep-interference sleeping posture and sleeping behavior testing identification method |
| CN106388819A (en) * | 2016-10-28 | 2017-02-15 | 许昌学院 | Human upper limb muscle state monitoring system based on surface electromyogram signals and judging method thereof |
| CN106667435A (en) * | 2016-12-17 | 2017-05-17 | 复旦大学 | Intelligent sensing mattress for monitoring sleep |
| CN107126005A (en) * | 2017-07-03 | 2017-09-05 | 广东葆康信息科技有限公司 | Biomethanics mattress and system |
| CN108420228A (en) * | 2018-04-13 | 2018-08-21 | 浙江想能云软件股份有限公司 | A kind of soft or hard adjustable bed mattess of intelligence and its monitoring method of sleep state monitoring |
Non-Patent Citations (5)
| Title |
|---|
| YIKUN GU ETAL: "Robust EMG pattern recognition in the presence of confounding", 《EXPERT SYSTEMS WITH APPLICATIONS》 * |
| 任丹: "基于表面肌电信号特征的动作识别", 《辽东学院学报》 * |
| 王海宁: "《基于多通道生理信号的情绪识别技术研究》", 31 August 2016 * |
| 电线的绝缘层和屏蔽层: "《航天器电源系统》", 31 December 2013 * |
| 韩晓新, 邢绍邦等: "基于 EMG 频域特征的假肢机电信号识别研究", 《测试技术学报》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108836012B (en) | 2021-09-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109157194B (en) | System and method for acquiring and analyzing human health data of mattress with adjustable hardness | |
| EP3735173B1 (en) | Multi-functional tubular worn garment | |
| CN108420228A (en) | A kind of soft or hard adjustable bed mattess of intelligence and its monitoring method of sleep state monitoring | |
| CN109744785B (en) | Intelligent self-adaptive mattress based on flexible pressure sensor array | |
| US10973344B2 (en) | Bedding system with a CNN based machine vision process | |
| CN101115438B (en) | Device for determining a stress level of a person and providing feedback on the basis of the stress level as determined | |
| Kiaghadi et al. | Phyjama: Physiological sensing via fiber-enhanced pyjamas | |
| CN110072432A (en) | Improve the adaptivity sleep system of personal sleep condition using data analysis and learning art | |
| CN108308942A (en) | The soft or hard adjustable bed mattess of intelligence and its control method of a kind of fusion family Internet of Things | |
| US20190328277A1 (en) | Device for Measuring Sleep Apnea and Method Therefor | |
| CN105877713A (en) | Method for automatically adjusting sleeping postures by fusing multivariate information | |
| CN105595672A (en) | Intelligent mattress system and method for accurately acquiring human body vital sign data on mattress | |
| CN205728862U (en) | Intelligent bed mattress system | |
| CN109091126A (en) | A kind of human body physiological state differentiation method for visualizing and equipment based on multisensor | |
| CN112869710A (en) | Bed with sleeping and physical therapy functions | |
| CN106725466A (en) | A kind of bow-backed alarm set and based reminding method based on surface myoelectric technology | |
| CN108836012A (en) | A kind of soft or hard adjustable intelligent mattress and its control method with human muscle's state-detection | |
| CN105854147A (en) | Intelligent home environment regulation system and method | |
| JPH08322804A (en) | Biological signal detector | |
| CN117017295A (en) | Pressure analysis relieving method and device based on sleep monitoring and computer equipment | |
| CN210354022U (en) | Intelligent self-adaptive mattress based on flexible pressure sensor array | |
| CN207821962U (en) | A kind of O-shaped leg apparatus for correcting of inflatable | |
| JP4172193B2 (en) | Massage machine | |
| CN207721937U (en) | A kind of O-shaped leg apparatus for correcting detected using muscle signals | |
| CN106923949B (en) | A kind of pigeon breast apparatus for correcting based on muscle signals detection |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |