CN105832348A - Intelligent sensing blanket - Google Patents

Abstract

The invention discloses an intelligent sensing blanket, which comprises a central processing unit as well as an emotion sensing module, a brain wave sensing module, a body temperature sensing module and a sitting posture sensing module which are in wireless connection to the central processing unit, wherein the emotion sensing module is used for determining the variety of sensed emotion in accordance with analyzed target emotion physiological response characteristics and combinations thereof; the brain wave sensing module comprises a plurality of brain wave sensors, a brain wave processing unit, a power management module and a brain wave sleep promoting unit; a non-contact infrared temperature sensor is arranged in the body temperature sensing module; and an acceleration meter and a tension sensor are arranged in the sitting posture sensing module. The intelligent sensing blanket disclosed by the invention, by virtue of the emotion sensing module, the brain wave sensing module and the body temperature sensing module, can sense emotion, brain wave signal and body temperature of passengers so as to regulate environment in a cabin; and by virtue of the sitting posture sensing module, the sitting posture of the passengers can be sensed and the posture of seats can be automatically regulated, so as to offer a more comfortable flying experience to the passengers in a long airline journey.

Description

A kind of Intellisense woollen blanket

Technical field

The invention belongs to smart machine field, particularly relate to a kind of Intellisense woollen blanket.

Background technology

The time difference is the most irritating, and it has upset original biological clock, allows people be much troubled.Environment on aircraft is the most uncomfortable, empty Between also ratio narrow, airline be difficult to carrys out the environment in adjustment machine according to the health of passenger, it is provided that comfortable experience.

Summary of the invention

It is an object of the invention to provide a kind of Intellisense woollen blanket, it is intended to solve airline and be difficult to the body according to passenger The problem that body situation carrys out the environment in adjustment machine.

The present invention is achieved in that a kind of Intellisense woollen blanket, and described Intellisense woollen blanket includes central processing unit And with the mood sensing module of described central processing unit wireless connections, brain wave sensing module, body temperature sensing module, sitting posture sense Know module；

Described mood sensing module is for perception and the emotional information of collection measurand, and emotion described in Treatment Analysis Information is with tension value of producing a feeling, and then controls letter according to described nervous value to produce the unlatching controlling recording module recording Number；

Described mood sensing module also includes communication module and signal processing module, this communication module and described emotion sense Know that module is connected, for sending described unlatching control signal to described recording module；And described recording module, logical with described Letter module is connected, for carrying out recording according to described unlatching control signal；

Described signal processing module is for carrying out emotion physiology by the physiological responses signal that the target emotion of acquisition reacts Response characteristic is analyzed, and analyzes physiological signal key character and the group thereof of reaction target emotion change from physiological responses signal Close, according to analyzing target emotion physiological responses feature of obtaining and combinations thereof and determine the kind of perception emotion, including defeating, emerging Put forth energy, be sick of, nervous；

Tension value T=k1 of described intense strain × E1 (HRV)+k × E (P)+k × E (R), wherein,

k₁+k₂+k₃=1；

E₂(P)=(P (t)-P (t-1))/P₀, 0 < E2 (P) < 1；

E₃(R)=(A R (t))/A, 0 < E₃(R)<1；

HRV, P and R represent heart rate change value, pressure value and epidermis conduction resistance, k respectively₁, k₂, k₃For weight coefficient, divide The existing changes in heart rate of complicated variant, blood pressure and the contribution to nervous degree metric of the epidermis electric conductivity, E₁(HRV) it is to become according to heart rate Change the nervous degree calculated, E₂(P) for the nervous degree calculated according to blood pressure,_E3(R) it is according to skin The nervous degree that electric conductivity change calculations goes out, t is current time, and t-1 is the previous moment of current time, and t-2 is current Front two moment in moment,For the heart rate change value sum of t-2 moment, t-1 moment and current time, HRV (t-2) is The heart rate change value in t-2 moment, HRV (t-1) is the heart rate change value in t-1 moment, and HRV (t) is the changes in heart rate of current time Value, H0 is the heart rate value under the normal emotional state of measurand, and P (t) is the pressure value of current time, and P (t-1) is the t-1 moment Pressure value, P0 is measurand pressure value under normal emotional state, A be measurand measure in advance skin resistance ginseng Examining value, R (t) is current time skin resistance；

Described brain wave sensing module includes:

Eeg signal classification identification module, for classifying to EEG signals and identifying；

Multiple brain wave sensors, configuration detects the brain electricity of the user wearing described brain wave detection device in use Ripple, and produce eeg signal；

Brain wave processing unit, for being amplified obtaining with Filtering Processing by the simulation eeg signal collected The simulation eeg signal including α ripple, β ripple, θ ripple and δ ripple in the range of 0.5Hz-100Hz；Simulation eeg signal is carried out Analog digital conversion carries out Fourier transformation and respectively obtains α ripple, β ripple, θ ripple and the Fourier spectrum of δ ripple after becoming digital brain electrical ripple signal, will Signal from space field transformation to frequency domain；The digital brain electrical ripple signal including α ripple, β ripple, θ ripple and δ ripple is carried out at the window of triumphant pool Reason, obtains the indices parameter of eeg signal through amplitude analysis, time domain analysis and frequency domain analysis；

Power management module, configuration is connected with the plurality of brain wave sensor, and detects from the plurality of brain The eeg signal of radio wave sensor；

Brain wave sleeping unit, including the electric field generation coil in medicated pillow, medicated pillow, supply unit, medicated pillow is built-in with vibration Sensor, pulse amplifying change-over circuit, clock generating circuit, frequency dividing circuit, frequency control circuit, sine wave filter, brain electricity Wave filter, selection switch, drive amplification circuit；Insomniac is tossed about what motion was converted into by described vibrating sensor Electronic impulse is input to the pulse amplifying change-over circuit one end being attached thereto, and inputs from the other end after it is converted into control signal To the clock generating circuit being attached thereto；Clock generating circuit on the one hand be input to standard second pulse signal to be attached thereto point Frequency circuit one end, shallow sleep and deep sleep brain wave frequency of oscillation that frequency dividing is gone out by the divided circuit other end are 0.4 9Hz's The frequency control circuit interconnected therewith it is input to Deng wide pulse signal；

Described body temperature sensing module is built-in with non-contact infrared temperature sensor, this non-contact infrared temperature sensing Device is connected with temperature-difference thermoelectric heap amplifying circuit and temperature-compensating and amplifying circuit respectively, temperature-difference thermoelectric heap amplifying circuit and temperature Degree compensates and amplifying circuit is connected is connected respectively to A/D convertor circuit, and described A/D convertor circuit is a multi-channel A/D conversion electricity Road, A/D convertor circuit is connected with governor circuit, and governor circuit is connected with display circuit and warning circuit；Described is contactless Infrared temperature sensor uses thermopile IR temperature sensor to realize temperature signals and ambient temperature signal i.e. temperature-difference thermoelectric Voltage signal that heap is faint and the non-contact detecting of the thermistor signal of electro-hot regulator；

Described sitting posture sensing module includes sitting position corrector body, is provided with acceleration in sitting position corrector body Meter, tension pick-up；

The frequency control circuit that output timing signal also is controlled to be attached thereto by described brain wave sleeping unit is by setting Fixed time sequencing timing working, make frequency control circuit by input from frequency dividing circuit whole wait wide pulse signals be converted to The most some minutes is interval, from the beginning of shallow sleeping brain wave frequency of oscillation point 9Hz, gradually reduces frequency to 4Hz, and automatic Transition Starting to deep sleep brain wave frequency of oscillation 4Hz, gradually reduce the frequency wide output of pulse signal to 0.4Hz, clock occurs The triggering of the control signal that circuit is also sent by pulse amplifying change-over circuit controls, and often triggers and once just extends the time set The timing point of order extends some minutes, so that its frequency control circuit controlled is at this shallow sleep or deep sleep signal frequency point The working time set extends some minutes；

Described body temperature sensing module is built-in with body temperature calibration module, and this body temperature calibration module includes that infrared body temperature monitoring sets Standby, calibration data processing center, temperature measuring equipment, environmental parameter monitoring apparatus, environment parameter monitoring system, infrared body temperature monitoring is System；Wherein

(1) temperature measuring equipment, is used for gathering auditory meatus or shell temperature data T0, and transmits to calibration data processing center；

(2) environmental parameter monitoring apparatus: be used for gathering context dependant information, and environmental information changes into the signal of telecommunication, passes Transport to environment parameter monitoring system；

(3) environment parameter monitoring system, after receiving the signal of telecommunication collected, is processed by data, and the cyclization that converts Border supplemental characteristic, and transmit to calibration data processing center；

(4) infrared body surface temperature monitoring equipment, the body surface infrared information of same individual in acquisition step (1), and will Infrared information changes into the signal of telecommunication, transmits to infrared body temperature monitoring system；

(5) infrared body temperature monitoring system, for receiving the signal of telecommunication that infrared body surface temperature monitoring equipment transmits, will The signal of telecommunication carries out data process, is converted into shell temperature Ty incoming calibration data processing center；

(6) calibration data processing center, for by the auditory canal temperature collected or shell temperature, ambient parameter, infrared body Table temperature carries out data analysis and process, sets up the calibration parameter curve of ambient parameter-shell temperature-body temperature, is sent out by this curve Give central processing unit.

Further, the eeg signal classification recognition methods of described eeg signal classification identification module concretely comprises the following steps:

Step one, chooses main examination and time examination, to the EEG signal number consecutively that 5 experimenters are corresponding be EEG_data_al, EEG_data_aa, EEG_data_av, EEG_data_ay, EEG_data_aw, selected al experimenter is the most tested, and other are four years old Position experimenter is secondary tested；

Step 2, frequency domain filtering, use one 8～the band filter of 30Hz, the eeg data gathered is filtered Pretreatment, and this frequency band has obvious ERD/ERS physiological phenomenon；

Step 3, chooses training sample, after signal filtering, from the A class and B class EEG signal of major experimental person al Choose 11 EEG signal respectively as training sample, then want the A class of experimenter and B class EEG signal point from other 4 precedences Do not choose 10 EEG signal and be 40 as training sample, A class and the B class training sample sum of the most all examination persons；

Step 4, obtains the A class of experimenter and covariance matrix sum RA of B class training sample and RB respectively, all times Examination person's A class and the covariance matrix sum of B class training sampleWith

$R_A=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Ai}{X_{Ai}}^T}{tr\left(X_{Ai}{X_{Ai}}^T\right)}$

$R_B=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Bi}{X_{Bi}}^T}{tr\left(X_{Bi}{X_{Bi}}^T\right)}$

Wherein, X_Ai(i=1,2...10) represents that experimenter's i & lt imagines left chirokinesthetic EEG signal, X_Bi(i=1, 2...10) represent that experimenter's i & lt imagines right chirokinesthetic EEG signal, X_(i,A) ^TRepresent the transposition of X (i, A), tr (X_(i,A) X_(i,A) ^T) representing matrix X_(i,A)X_(i,A) ^TMark,

${\hat{R}}_A=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T}{tr\left({\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T\right)}$

${\hat{R}}_B=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T}{tr\left({\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T\right)}$

Wherein,Represent that time examination person's i & lt imagines left chirokinesthetic EEG signal,Represent that time examination person's i & lt imagines right chirokinesthetic EEG signal；

Step 5, seeks regularized covariance matrix

Introducing regularization parameter and β, span is α ∈ [0,1] and β ∈ [0,1], and α takes 0,0.001,0.01 respectively, 0.1,0.2；β takes 0,0.01,0.1,0.2,0.4,0.6 respectively, under the effect of regularization parameter, by the covariance square of experimenter Battle array sum combines with the covariance matrix sum of time examination person, constructs two classes average regularized covariance matrix, the following institute of formula Show:

$Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

$Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

Wherein,RepresentMark, I is the unit of N × N Matrix, N is channel acquisition number；

Step 6, by the two classes average regularized covariance Matrix Calculating in step 5 and and carry out Eigenvalues Decomposition, solve Canonical whitening matrix, as follows:

$Z(\mathrm{\&alpha;},\mathrm{\&beta;})=Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})+Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=\hat{U}\&CenterDot;\widehat{\mathrm{\&Lambda;}}\&CenterDot;{\hat{U}}^T$

Wherein,It is characterized value diagonal matrix,For characteristic of correspondence vector matrix, then canonical whitening matrix is:

$P={\widehat{\mathrm{\&Lambda;}}}^{(-\frac{1}{2})}\&CenterDot;{\hat{U}}^T$

Step 7, to the Z of gained in step 6_A(α, β) and Z_B(α, β) changes as follows:

${\stackrel{\&OverBar;}{Z}}_A(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_A\&CenterDot;{\mathrm{\&Lambda;}}_A\&CenterDot;{U_A}^T$

${\stackrel{\&OverBar;}{Z}}_B(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_B\&CenterDot;{\mathrm{\&Lambda;}}_B\&CenterDot;{U_B}^T$

Wherein, Λ_AAnd Λ_BBeing characterized value diagonal matrix, UA and UB is characteristic of correspondence vector matrix, chooses diagonal matrix Λ_A、Λ_BMiddle eigenvalue of maximum characteristic of correspondence vector, structure spatial filter is as follows:

W_A=U_A ^T·P

W_B=U_B ^T·P

Step 8, by two class EEG signal X of training sample_AAnd X_BThrough corresponding wave filter W_A、W_BHave:

F_A=W_A ^T·X_A

F_B=W_B ^T·X_B

Step 9, in step 8 through regularization common space mode filtering EEG signals calculate power spectral density, Ask for the power spectral density value that frequency is 8～15Hz, utilize structural learning dictionary B=[F_A F_B]；

Step 10, chooses one group of data in the training sample of experimenter successively as test sample y, enters by above step Row filtering, projection, the test sample data after reservation process；

Step 11, the rarefaction representation solving test sample as the following formula is vectorial:

$\begin{array}{ccc}{\hat{x}}_1=min\left|\right|x|{|}_1& subjectto& \left|\right|\hat{y}-\hat{B}x|{|}_2\&le;\mathrm{\&epsiv;}\end{array}$

Wherein x is the rarefaction representation vector of test Mental imagery sample to be solved, test Mental imagery sample to be solved for y Notebook data, is error threshold for ε；

Step 12, for Mental imagery i each time, according to the rarefaction representation vector of test sampleCalculate residual error

$r_i\left(\hat{y}\right)=\left|\right|\hat{y}-{\mathrm{BT}}_i\left({\hat{x}}_i\right)\left|\right|$

WhereinIt is by rarefaction representation vectorThe new vector obtained, in this vector, the i-th type games imagination is corresponding Element entry identical with corresponding element entry in rarefaction representation vector, other element entries are zero；

Step 13, by the minimum classification of residual error as the recognition result of final Mental imagery classification:

It is test sample data.

Further, the communication means of described wireless communication module includes launching n road source signal, receiving terminal receives aliasing letter Number, piece-rate system separation multichannel aliasing signal；

After described transmitting n road source signal refers to the mixing of n road source signal channel, hybrid system is referred to as A, transmitting terminal by N root antenna is in spatial emission；

Described receiving terminal receives aliasing signal and refers to that receiving terminal utilizes m (m >=n ＞ 1) root sky bundle of lines aliasing signal to receive Getting off, receive signal and be referred to as observation signal, receiving terminal first carries out the pretreatment of observation signal, and pretreatment comprises two parts, i.e. Centralization processes and spherization process；

Described piece-rate system separation multichannel aliasing signal refers to that piece-rate system W can be according to each road source signal information entropy Different this multichannel aliasing signals that separate in entropy territory, wherein the criterion of information entropy uses negentropy；

Wherein, the expression formula of negentropy approximate calculation is as follows:

Wherein, k_jFor some normal numbers, M is the gaussian variable with zero-mean, unit variance, function G_jFor non-secondary letter Number；

As all of G_jDuring=G, approximate expression becomes:

J_G(x)≈C[E{G(x)}-E{G(M)}]²(formula 2)

Wherein, G is any non-quadratic function, and C is a constant；

Then using above-mentioned formula to carry out the calculating of negentropy, the difference according to each road signal negentropy value can be each road signal Extract, it is achieved the multiplexing of channel.

Further, described piece-rate system separation multichannel aliasing signal connects with multi-path antenna owing to using multi-path antenna to launch The MIMO technique received, MIMO-EDM wireless communication system is along with the increase of dual-mode antenna number, limit channels capacity Also can increase the most linearly, the calculating process of MIMO-EDM wireless communication system power-carrying is as follows:

Laguerre polynomial computation is utilized to obtain:

Wherein, m=min (N_t,N_r)

N=max (N_t,N_r)

For the Laguerre multinomial that number of times is k

If making λ=n/m, the channel capacity expression after following normalization can be derived；

Wherein,

$v_2={(\sqrt{\mathrm{\&tau;}}+1)}^2$

In the case of fast Rayleigh declines, make m=n=N_t=N_r, then v₁=0, v₂=4；

Progressive channel capacity is:

Utilize inequality:

log₂(1+x)≥log₂(x) (formula 6)

Formula (5) is reduced to:

The present invention passes through mood sensing module, brain wave sensing module, body temperature sensing module sensing passengers emotion, brain wave Signal, body temperature, so adjust in cabin environment, it is provided that comfortable experience, by sitting posture sensing module sensing passengers sitting posture, And power seat posture so that the passenger of long haul aircraft travelling has and more comfortable seizes the opportunity experience.The present invention is by using R- Tradition common space pattern (CSP) algorithm carries out feature extraction to Mental imagery EEG signals, it is achieved that dimensionality reduction, effectively reduces Calculating when using the Method of EEG signals classification SRC of rarefaction representation based on signal to Mental imagery EEG's Recognition Complexity, R-tradition common space pattern (CSP) algorithm is to utilize the theory of matrix simultaneous diagonalization on algebraically, finds one group of sky Between wave filter so that this group wave filter effect under, the variance of a class signal reaches very big, and another kind of signal is that variance reaches Minimum, thus reach the purpose of classification, R-tradition common space pattern (CSP) is by introducing regularization parameter and β, by multidigit The training data weighting summation of experimenter combines, and effectively avoids the drawback that little training sample feature extraction is unstable, reduces The individual difference of data, takes full advantage of the experimental data of other subjectss.The inventive method is simple, easy to operate preferably Solve existing EEG feature extraction exist construction feature vector eigenvalue stability low, discrimination compares Difference, identifies that the characteristic vector obtained that classification exists is difficult to have linear separability, classification causes the biggest puzzlement, causes identification The problem that rate reduces.The multichannel that present invention achieves channel is multiple, has bigger channel capacity and higher spectrum utilization efficiency, Such that it is able to realize two-forty, high-quality broadband wireless communications.

Accompanying drawing explanation

Fig. 1 is the built-in system composition diagram of the Intellisense woollen blanket that the embodiment of the present invention provides；

In figure: 1, central processing unit；2, mood sensing module；3, brain wave sensing module 3-1, brain wave sensor；3- 2, brain wave processing unit；3-3, power management module；3-4, brain wave sleeping unit；4, body temperature sensing module；5, sitting posture sense Know module.

Detailed description of the invention

For the summary of the invention of the present invention, feature and effect can be further appreciated that, hereby enumerate following example, and coordinate accompanying drawing Describe in detail as follows.

As it is shown in figure 1, the present invention is achieved in that a kind of Intellisense woollen blanket includes central processing unit 1 and with described The mood sensing module 2 of central processing unit 1 wireless connections, brain wave sensing module 3, body temperature sensing module 4, sitting posture perception mould Block 5；

Described mood sensing module 2 is for perception and the emotional information of collection measurand, and feelings described in Treatment Analysis Thread information is with tension value of producing a feeling, and then controls to produce the unlatching controlling recording module recording according to described nervous value Signal；

Described mood sensing module 2 also includes communication module and signal processing module, this communication module and described emotion Sensing module is connected, for sending described unlatching control signal to described recording module；And described recording module, with described Communication module is connected, for carrying out recording according to described unlatching control signal；

Described signal processing module is for carrying out emotion physiology by the physiological responses signal that the target emotion of acquisition reacts Response characteristic is analyzed, and analyzes physiological signal key character and the group thereof of reaction target emotion change from physiological responses signal Close, according to analyzing target emotion physiological responses feature of obtaining and combinations thereof and determine the kind of perception emotion, including defeating, emerging Put forth energy, be sick of, nervous；

Tension value T=k1 of described intense strain × E1 (HRV)+k × E (P)+k × E (R), wherein,

k₁+k₂+k₃=1；

E₂(P)=(P (t)-P (t-1))/P₀, 0 < E2 (P) < 1；

E₃(R)=(A R (t))/A, 0 < E₃(R)<1；

HRV, P and R represent heart rate change value, pressure value and epidermis conduction resistance, k respectively₁, k₂, k₃For weight coefficient, divide The existing changes in heart rate of complicated variant, blood pressure and the contribution to nervous degree metric of the epidermis electric conductivity, E₁(HRV) it is to become according to heart rate Change the nervous degree calculated, E₂(P) for the nervous degree calculated according to blood pressure,_E3(R) it is according to skin The nervous degree that electric conductivity change calculations goes out, t is current time, and t-1 is the previous moment of current time, and t-2 is current Front two moment in moment,For the heart rate change value sum of t-2 moment, t-1 moment and current time, HRV (t-2) is The heart rate change value in t-2 moment, HRV (t-1) is the heart rate change value in t-1 moment, and HRV (t) is the changes in heart rate of current time Value, H0 is the heart rate value under the normal emotional state of measurand, and P (t) is the pressure value of current time, and P (t-1) is the t-1 moment Pressure value, P0 is measurand pressure value under normal emotional state, A be measurand measure in advance skin resistance ginseng Examining value, R (t) is current time skin resistance；

Described brain wave sensing module 3 includes:

Eeg signal classification identification module, for classifying to EEG signals and identifying；

Multiple brain wave sensor 3-1, configuration detects the user wearing described brain wave detection device in use Brain wave, and produce eeg signal；

Brain wave processing unit 3-2, is used for: be amplified obtaining with Filtering Processing by the simulation eeg signal collected The simulation eeg signal including α ripple, β ripple, θ ripple and δ ripple in the range of 0.5Hz-100Hz；Simulation eeg signal is carried out Analog digital conversion carries out Fourier transformation and respectively obtains α ripple, β ripple, θ ripple and the Fourier spectrum of δ ripple after becoming digital brain electrical ripple signal, will Signal from space field transformation to frequency domain；The digital brain electrical ripple signal including α ripple, β ripple, θ ripple and δ ripple is carried out at the window of triumphant pool Reason, obtains the indices parameter of eeg signal through amplitude analysis, time domain analysis and frequency domain analysis；

Power management module 3-3, configuration is connected with the plurality of brain wave sensor, and detects from the plurality of The eeg signal of brain wave sensor；

Brain wave sleeping unit 3-4, including the electric field generation coil in medicated pillow, medicated pillow, supply unit, medicated pillow is built-in with and shakes Dynamic sensor, pulse amplifying change-over circuit, clock generating circuit, frequency dividing circuit, frequency control circuit, sine wave filter, brain Electric wave filter, selection switch, drive amplification circuit；The motion of insomniac being tossed about of described vibrating sensor is converted into Electronic impulse be input to pulse amplifying change-over circuit one end of being attached thereto, defeated from the other end after it is converted into control signal Enter to the clock generating circuit being attached thereto；On the one hand standard second pulse signal is input to be attached thereto by clock generating circuit Frequency dividing circuit one end, shallow sleep and deep sleep brain wave frequency of oscillation that frequency dividing is gone out by the divided circuit other end are 0.4 9Hz The wide pulse signal that waits be input to the frequency control circuit that interconnects therewith；

Described body temperature sensing module 4 is built-in with non-contact infrared temperature sensor, and this non-contact infrared temperature passes Sensor is connected with temperature-difference thermoelectric heap amplifying circuit and temperature-compensating and amplifying circuit respectively, temperature-difference thermoelectric heap amplifying circuit and Temperature-compensating and amplifying circuit are connected and are connected respectively to A/D convertor circuit, and described A/D convertor circuit is a multi-channel A/D conversion electricity Road, A/D convertor circuit is connected with governor circuit, and governor circuit is connected with display circuit and warning circuit；Described is contactless Infrared temperature sensor uses thermopile IR temperature sensor to realize temperature signals and ambient temperature signal i.e. temperature-difference thermoelectric Voltage signal that heap is faint and the non-contact detecting of the thermistor signal of electro-hot regulator；

Described sitting posture sensing module 5 includes sitting position corrector body, is provided with acceleration in sitting position corrector body Meter, tension pick-up；

Described tension pick-up includes dynamometer link, strip potsherd and pottery fixing seat, tension signal processing module； The bonding dynamometer link in described strip potsherd front, seat fixed by the back side bonding elastomer pottery；Described strip potsherd is as from pottery Porcelain fixes seat to the moment between dynamometer link；

Described tension signal processing means is connected with described side force bar and described strip potsherd, this tension signal Processing means include two groups of tensile strain sheets and compression strain sheet composition tension force inductive component, built-in signal processor, two Surveying a roller, three jockey pulleys and sensor installation seat, described sensor installation seat is arranged side by side three jockey pulleys, two-by-two Survey roller, three jockey pulley place straight lines and two survey roller place straight line parallels, described survey it is provided with between adjacent jockey pulley Open installation tension inductive component on roller；

Described strip potsherd is a quick elastomer of power, and this power quick elastomer positive and negative respectively arranges two strain resistors, with Time by the through hole arranged on the quick elastomer of power, these four strain resistors interconnections are formed resistance bridges；And pass through laser Resistance bridge is adjusted to zero-bit by the system that trims；After described dynamometer link perception tension signal, micro-by making the quick elastomer of power produce Amount deformation, so that whole tension pick-up is the most electric by the millivolt level that electric bridge output is relevant to tension signal precision linear Pressure.

Further, output timing signal is also controlled the frequency control being attached thereto by described brain wave sleeping unit 3-4 Circuit processed sequence timing according to set time works, and makes frequency control circuit the broad pulses such as whole input from frequency dividing circuit be believed Number be converted to the most some minutes for interval, from the beginning of shallow sleeping brain wave frequency of oscillation point 9Hz, gradually reduce frequency to 4Hz, And automatic Transition starts to deep sleep brain wave frequency of oscillation 4Hz, gradually reduce frequency defeated to the wide pulse signal such as grade of 0.4Hz Going out, the triggering of the control signal that clock generating circuit is also sent by pulse amplifying change-over circuit controls, and often triggers and once just extends The timing point of the time sequencing set extends some minutes, so that its frequency control circuit controlled is in this shallow sleep or deep The working time that sleep signal frequency sets extends some minutes so that its frequency control circuit controlled in this shallow sleep or The working time that deep sleep signal frequency point sets extends some minutes；Frequency control circuit while by timing, gradually change and shake The selection that the wide pulse signal such as shallow, deep sleep brain wave swinging frequency is input to be attached thereto switchs one end, by selecting switch certainly Surely connect sine wave filter or brain wave filter, after guiding insomniac to go to sleep peacefully, send off signal Control end to the supply unit being attached thereto and close power supply shutdown；Another of described sine wave filter or brain wave filter Hold the sine wave that will filter out or simulating human brain wave shape signal to be input to the drive amplification circuit on one side being attached thereto, drive Be input to the electric field generation coil being attached thereto after amplifying circuit amplification from the other end, driving electric field generation coil forms simulation people Class to the bio-electric field of deep sleep brain wave, thus reaches without Drug therapy from shallow sleeping brain wave, just can make insomniac Break away from insomnia and enter the purpose of deep sleep.

Further, described body temperature sensing module 4 is built-in with body temperature calibration module, and this body temperature calibration module includes infrared body Temperature monitoring device, calibration data processing center, temperature measuring equipment, environmental parameter monitoring apparatus, environment parameter monitoring system, infrared body Temperature monitoring system；Wherein

(1) temperature measuring equipment, is used for gathering auditory meatus or shell temperature data T0, and transmits to calibration data processing center；

(2) environmental parameter monitoring apparatus: be used for gathering context dependant information, and environmental information changes into the signal of telecommunication, passes Transport to environment parameter monitoring system；

(3) environment parameter monitoring system, after receiving the signal of telecommunication collected, is processed by data, and the cyclization that converts Border supplemental characteristic, and transmit to calibration data processing center；

(4) infrared body surface temperature monitoring equipment, the body surface infrared information of same individual in acquisition step (1), and will Infrared information changes into the signal of telecommunication, transmits to infrared body temperature monitoring system；

(5) infrared body temperature monitoring system, for receiving the signal of telecommunication that infrared body surface temperature monitoring equipment transmits, will The signal of telecommunication carries out data process, is converted into shell temperature Ty incoming calibration data processing center；

(6) calibration data processing center, for by the auditory canal temperature collected or shell temperature, ambient parameter, infrared body Table temperature carries out data analysis and process, sets up the calibration parameter curve of ambient parameter-shell temperature-body temperature, is sent out by this curve Give central processing unit.

Further, the eeg signal classification recognition methods of described eeg signal classification identification module concretely comprises the following steps:

Step one, chooses main examination and time examination, to the EEG signal number consecutively that 5 experimenters are corresponding be EEG_data_al, EEG_data_aa, EEG_data_av, EEG_data_ay, EEG_data_aw, selected al experimenter is the most tested, and other are four years old Position experimenter is secondary tested；

Step 2, frequency domain filtering, use one 8～the band filter of 30Hz, the eeg data gathered is filtered Pretreatment, and this frequency band has obvious ERD/ERS physiological phenomenon；

Step 3, chooses training sample, after signal filtering, from the A class and B class EEG signal of major experimental person al Choose 11 EEG signal respectively as training sample, then want the A class of experimenter and B class EEG signal point from other 4 precedences Do not choose 10 EEG signal and be 40 as training sample, A class and the B class training sample sum of the most all examination persons；

Step 4, obtains the A class of experimenter and covariance matrix sum RA of B class training sample and RB respectively, all times Examination person's A class and the covariance matrix sum of B class training sampleWith

$R_A=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Ai}{X_{Ai}}^T}{tr\left(X_{Ai}{X_{Ai}}^T\right)}$

$R_B=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Bi}{X_{Bi}}^T}{tr\left(X_{Bi}{X_{Bi}}^T\right)}$

Wherein, X_Ai(i=1,2...10) represents that experimenter's i & lt imagines left chirokinesthetic EEG signal, X_Bi(i=1, 2...10) represent that experimenter's i & lt imagines right chirokinesthetic EEG signal, X_(i,A) ^TRepresent the transposition of X (i, A), tr (X_(i,A) X_(i,A) ^T) representing matrix X_(i,A)X_(i,A) ^TMark,

${\hat{R}}_A=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T}{tr\left({\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T\right)}$

${\hat{R}}_B=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T}{tr\left({\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T\right)}$

Wherein,Represent that time examination person's i & lt imagines left chirokinesthetic EEG signal,Represent that time examination person's i & lt imagines right chirokinesthetic EEG signal；

Step 5, seeks regularized covariance matrix

Introducing regularization parameter and β, span is α ∈ [0,1] and β ∈ [0,1], and α takes 0,0.001,0.01 respectively, 0.1,0.2；β takes 0,0.01,0.1,0.2,0.4,0.6 respectively, under the effect of regularization parameter, by the covariance square of experimenter Battle array sum combines with the covariance matrix sum of time examination person, constructs two classes average regularized covariance matrix, the following institute of formula Show:

$Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

$Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

Wherein,RepresentMark, I is the unit of N × N Matrix, N is channel acquisition number；

Step 6, by the two classes average regularized covariance Matrix Calculating in step 5 and and carry out Eigenvalues Decomposition, solve Canonical whitening matrix, as follows:

$Z(\mathrm{\&alpha;},\mathrm{\&beta;})=Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})+Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=\hat{U}\&CenterDot;\widehat{\mathrm{\&Lambda;}}\&CenterDot;{\hat{U}}^T$

Wherein,It is characterized value diagonal matrix,For characteristic of correspondence vector matrix, then canonical whitening matrix is:

$P={\widehat{\mathrm{\&Lambda;}}}^{(-\frac{1}{2})}\&CenterDot;{\hat{U}}^T$

Step 7, to the Z of gained in step 6_A(α, β) and Z_B(α, β) changes as follows:

${\stackrel{\&OverBar;}{Z}}_A(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_A\&CenterDot;{\mathrm{\&Lambda;}}_A\&CenterDot;{U_A}^T$

${\stackrel{\&OverBar;}{Z}}_B(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_B\&CenterDot;{\mathrm{\&Lambda;}}_B\&CenterDot;{U_B}^T$

Wherein, Λ_AAnd Λ_BBeing characterized value diagonal matrix, UA and UB is characteristic of correspondence vector matrix, chooses diagonal matrix Λ_A、Λ_BMiddle eigenvalue of maximum characteristic of correspondence vector, structure spatial filter is as follows:

W_A=U_A ^T·P

W_B=U_B ^T·P

Step 8, by two class EEG signal X of training sample_AAnd X_BThrough corresponding wave filter W_A、W_BHave:

F_A=W_A ^T·X_A

F_B=W_B ^T·X_B

Step 9, in step 8 through regularization common space mode filtering EEG signals calculate power spectral density, Ask for the power spectral density value that frequency is 8～15Hz, utilize structural learning dictionary B=[F_A F_B]；

Step 10, chooses one group of data in the training sample of experimenter successively as test sample y, enters by above step Row filtering, projection, the test sample data after reservation process；

Step 11, the rarefaction representation solving test sample as the following formula is vectorial:

$\begin{array}{ccc}{\hat{x}}_1=min\left|\right|x|{|}_1& subjectto& \left|\right|\hat{y}-\hat{B}x|{|}_2\&le;\mathrm{\&epsiv;}\end{array}$

Wherein x is the rarefaction representation vector of test Mental imagery sample to be solved, test Mental imagery sample to be solved for y Notebook data, is error threshold for ε；

Step 12, for Mental imagery i each time, according to the rarefaction representation vector of test sampleCalculate residual error

$r_i\left(\hat{y}\right)=\left|\right|\hat{y}-{\mathrm{BT}}_i\left({\hat{x}}_i\right)\left|\right|$

WhereinIt is by rarefaction representation vectorThe new vector obtained, in this vector, the i-th type games imagination is corresponding Element entry identical with corresponding element entry in rarefaction representation vector, other element entries are zero；

Step 13, by the minimum classification of residual error as the recognition result of final Mental imagery classification:

It is test sample data.

Further, the communication means of described wireless communication module includes launching n road source signal, receiving terminal receives aliasing letter Number, piece-rate system separation multichannel aliasing signal；

After described transmitting n road source signal refers to the mixing of n road source signal channel, hybrid system is referred to as A, transmitting terminal by N root antenna is in spatial emission；

Described receiving terminal receives aliasing signal and refers to that receiving terminal utilizes m (m >=n ＞ 1) root sky bundle of lines aliasing signal to receive Getting off, receive signal and be referred to as observation signal, receiving terminal first carries out the pretreatment of observation signal, and pretreatment comprises two parts, i.e. Centralization processes and spherization process；

Described piece-rate system separation multichannel aliasing signal refers to that piece-rate system W can be according to each road source signal information entropy Different this multichannel aliasing signals that separate in entropy territory, wherein the criterion of information entropy uses negentropy；

Wherein, the expression formula of negentropy approximate calculation is as follows:

Wherein, k_jFor some normal numbers, M is the gaussian variable with zero-mean, unit variance, function G_jFor non-secondary letter Number；

As all of G_jDuring=G, approximate expression becomes:

J_G(x)≈C[E{G(x)}-E{G(M)}]²(formula 2)

Wherein, G is any non-quadratic function, and C is a constant；

Then using above-mentioned formula to carry out the calculating of negentropy, the difference according to each road signal negentropy value can be each road signal Extract, it is achieved the multiplexing of channel.

Further, described piece-rate system separation multichannel aliasing signal connects with multi-path antenna owing to using multi-path antenna to launch The MIMO technique received, MIMO-EDM wireless communication system is along with the increase of dual-mode antenna number, limit channels capacity Also can increase the most linearly, the calculating process of MIMO-EDM wireless communication system power-carrying is as follows:

Laguerre polynomial computation is utilized to obtain:

Wherein, m=min (N_t,N_r)

N=max (N_t,N_r)

For the Laguerre multinomial that number of times is k

If making λ=n/m, the channel capacity expression after following normalization can be derived；

Wherein,

$v_2={(\sqrt{\mathrm{\&tau;}}+1)}^2$

In the case of fast Rayleigh declines, make m=n=N_t=N_r, then v₁=0, v₂=4；

Progressive channel capacity is:

Utilize inequality:

log₂(1+x)≥log₂(x) (formula 6)

Formula (5) is reduced to:

The present invention passes through mood sensing module, brain wave sensing module, body temperature sensing module sensing passengers emotion, brain wave Signal, body temperature, so adjust in cabin environment, it is provided that comfortable experience, by sitting posture sensing module sensing passengers sitting posture, And power seat posture so that the passenger of long haul aircraft travelling has and more comfortable seizes the opportunity experience.

The above is only to presently preferred embodiments of the present invention, and the present invention not makees any pro forma restriction, Every technical spirit according to the present invention, to any simple modification made for any of the above embodiments, equivalent variations and modification, belongs to In the range of technical solution of the present invention.

Claims (4)

1. an Intellisense woollen blanket, it is characterised in that described Intellisense woollen blanket includes central processing unit and with described The mood sensing module of central processing unit wireless connections, brain wave sensing module, body temperature sensing module, sitting posture sensing module；

Described mood sensing module is for perception and the emotional information of collection measurand, and emotional information described in Treatment Analysis With tension value of producing a feeling, and then according to described nervous value to produce the unlatching control signal controlling recording module recording；

Described mood sensing module also includes communication module and signal processing module, this communication module and described mood sensing mould Block is connected, for sending described unlatching control signal to described recording module；And described recording module, with described communication mould Block is connected, for carrying out recording according to described unlatching control signal；

Described signal processing module is for carrying out emotion physiological responses by the physiological responses signal that the target emotion of acquisition reacts Feature analysis, analyzes physiological signal key character of reaction target emotion change and combinations thereof, root from physiological responses signal According to analyzing target emotion physiological responses feature of obtaining and combinations thereof and determine the kind of perception emotion, including defeating, excited, detest Tired, nervous；

Tension value T=k1 of described intense strain × E1 (HRV)+k × E (P)+k × E (R), wherein,

k₁+k₂+k₃=1；

E₂(P)=(P (t)-P (t-1))/P₀, 0 < E2 (P) < 1；

E₃(R)=(A R (t))/A, 0 < E₃(R)<1；

HRV, P and R represent heart rate change value, pressure value and epidermis conduction resistance, k respectively₁, k₂, k₃For weight coefficient, respectively body Existing changes in heart rate, blood pressure and the contribution to nervous degree metric of the epidermis electric conductivity, E₁(HRV) it is according to changes in heart rate meter The nervous degree calculated, E₂(P) for the nervous degree calculated according to blood pressure,_E3(R) it is according to skin conductivity The nervous degree that property change calculations goes out, t is current time, and t-1 is the previous moment of current time, and t-2 is current time Front two moment,For the heart rate change value sum of t-2 moment, t-1 moment and current time, HRV (t-2) is t-2 The heart rate change value in moment, HRV (t-1) is the heart rate change value in t-1 moment, and HRV (t) is the heart rate change value of current time, H0 is the heart rate value under the normal emotional state of measurand, and P (t) is the pressure value of current time, and P (t-1) is the blood in t-1 moment Pressure value, P0 is measurand pressure value under normal emotional state, and A is the skin resistance reference that measurand is measured in advance Value, R (t) is current time skin resistance；

Described brain wave sensing module includes:

Eeg signal classification identification module, for classifying to EEG signals and identifying；

Multiple brain wave sensors, configuration detects the brain wave of the user wearing described brain wave detection device in use, And produce eeg signal；

Brain wave processing unit, for being amplified obtaining 0.5Hz-with Filtering Processing by the simulation eeg signal collected The simulation eeg signal including α ripple, β ripple, θ ripple and δ ripple in the range of 100Hz；Simulation eeg signal is carried out modulus turn Carry out Fourier transformation after changing digital brain electrical ripple signal into and respectively obtain α ripple, β ripple, θ ripple and the Fourier spectrum of δ ripple, by signal from Space field transformation is to frequency domain；The digital brain electrical ripple signal including α ripple, β ripple, θ ripple and δ ripple carries out triumphant pool window process, through width Value analysis, time domain analysis and frequency domain analysis obtain the indices parameter of eeg signal；

Power management module, configuration is connected with the plurality of brain wave sensor, and detects from the plurality of brain wave The eeg signal of sensor；

Brain wave sleeping unit, including the electric field generation coil in medicated pillow, medicated pillow, supply unit, medicated pillow is built-in with vibrating sensing Device, pulse amplifying change-over circuit, clock generating circuit, frequency dividing circuit, frequency control circuit, sine wave filter, brain wave are filtered Ripple device, selection switch, drive amplification circuit；Insomniac is tossed about the electronics that is converted into of motion by described vibrating sensor Pulse is input to the pulse amplifying change-over circuit one end being attached thereto, after it is converted into control signal from the other end be input to The clock generating circuit of connection；On the one hand standard second pulse signal is input to the frequency dividing electricity being attached thereto by clock generating circuit One end, road, shallow sleep and deep sleep brain wave frequency of oscillation that frequency dividing is gone out by the divided circuit other end are the wide of 0.4 9Hz Pulse signal is input to the frequency control circuit interconnected therewith；

Described body temperature sensing module is built-in with non-contact infrared temperature sensor, and this non-contact infrared temperature sensor divides Not be not connected with temperature-difference thermoelectric heap amplifying circuit and temperature-compensating and amplifying circuit, temperature-difference thermoelectric heap amplifying circuit and temperature benefit Repay and amplifying circuit be connected be connected respectively to A/D convertor circuit, described A/D convertor circuit is a multi-channel A/D change-over circuit, AD Change-over circuit is connected with governor circuit, and governor circuit is connected with display circuit and warning circuit；Described non-contact infrared It is micro-to temperature signals and the i.e. temperature difference heat pile of ambient temperature signal that temperature sensor uses thermopile IR temperature sensor to realize Weak voltage signal and the non-contact detecting of the thermistor signal of electro-hot regulator；

Described sitting posture sensing module includes sitting position corrector body, is provided with accelerometer, opens in sitting position corrector body Force transducer；

The frequency control circuit that output timing signal also is controlled to be attached thereto by described brain wave sleeping unit is by setting Time sequencing timing working, if making frequency control circuit be converted to the whole wide pulse signals such as grade inputted from frequency dividing circuit with every Dry minute is interval, from the beginning of shallow sleeping brain wave frequency of oscillation point 9Hz, gradually reduces frequency to 4Hz, and automatic Transition is to deeply Sleeping brain wave frequency of oscillation 4Hz starts, and gradually reduces the frequency wide output of pulse signal to 0.4Hz, clock generating circuit The triggering of the control signal also sent by pulse amplifying change-over circuit controls, and often triggers and once just extends the time sequencing set Timing point extend some minutes, so that its frequency control circuit controlled sets at this shallow sleep or deep sleep signal frequency point Working time extend some minutes；

Described body temperature sensing module is built-in with body temperature calibration module, and this body temperature calibration module includes infrared body temperature monitoring device, school Quasi-data processing centre, temperature measuring equipment, environmental parameter monitoring apparatus, environment parameter monitoring system, infrared body temperature monitoring system；Its In

(1) temperature measuring equipment, is used for gathering auditory meatus or shell temperature data T0, and transmits to calibration data processing center；

(2) environmental parameter monitoring apparatus: be used for gathering context dependant information, and environmental information changes into the signal of telecommunication, transmission is extremely Environment parameter monitoring system；

(3) environment parameter monitoring system, after receiving the signal of telecommunication collected, is processed by data, and is converted into environment ginseng Number data, and transmit to calibration data processing center；

(4) infrared body surface temperature monitoring equipment, the body surface infrared information of same individual in acquisition step (1), and by infrared Information changes into the signal of telecommunication, transmits to infrared body temperature monitoring system；

(5) infrared body temperature monitoring system, for receiving the signal of telecommunication that infrared body surface temperature monitoring equipment transmits, by telecommunications Number carry out data process, be converted into shell temperature Ty incoming calibration data processing center；

(6) calibration data processing center, for by the auditory canal temperature collected or shell temperature, ambient parameter, infrared body surface temperature Degree carries out data analysis and process, sets up the calibration parameter curve of ambient parameter-shell temperature-body temperature, is sent to by this curve Central processing unit.

2. Intellisense woollen blanket as claimed in claim 1, it is characterised in that the brain electricity of described eeg signal classification identification module Modulation recognition recognition methods concretely comprises the following steps:

Step one, chooses main examination and time examination, is EEG_data_al, EEG_ to the EEG signal number consecutively that 5 experimenters are corresponding Data_aa, EEG_data_av, EEG_data_ay, EEG_data_aw, selected al experimenter is the most tested, other four realities The person of testing is secondary tested；

Step 2, frequency domain filtering, use one 8～the band filter of 30Hz, the eeg data gathered is filtered pre-place Manage, and this frequency band has obvious ERD/ERS physiological phenomenon；

Step 3, chooses training sample, after signal filtering, from the A class and B class EEG signal of major experimental person al respectively Choose 11 EEG signal as training sample, then want the A class of experimenter and B class EEG signal are selected respectively from other 4 precedences Take 10 EEG signal and be 40 as training sample, A class and the B class training sample sum of the most all examination persons；

Step 4, obtains the A class of experimenter and covariance matrix sum RA of B class training sample and RB, all examination person A respectively Class and the covariance matrix sum of B class training sampleWith

$R_A=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Ai}{X_{Ai}}^T}{tr\left(X_{Ai}{X_{Ai}}^T\right)}$

$R_B=\underset{i=1}{\overset{10}{\&Sigma;}}\frac{X_{Bi}{X_{Bi}}^T}{tr\left(X_{Bi}{X_{Bi}}^T\right)}$

Wherein, X_Ai(i=1,2...10) represents that experimenter's i & lt imagines left chirokinesthetic EEG signal, X_Bi(i=1,2...10) Represent that experimenter's i & lt imagines right chirokinesthetic EEG signal, X_(i,A) ^TRepresent the transposition of X (i, A), tr (X_(i,A)X_(i,A) ^T) represent Matrix X_(i,A)X_(i,A) ^TMark,

${\hat{R}}_A=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T}{tr\left({\hat{X}}_{Ai}{{\hat{X}}_{Ai}}^T\right)}$

${\hat{R}}_B=\underset{i=1}{\overset{40}{\&Sigma;}}\frac{{\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T}{tr\left({\hat{X}}_{Bi}{{\hat{X}}_{Bi}}^T\right)}$

Wherein,Represent that time examination person's i & lt imagines left chirokinesthetic EEG signal,Represent that time examination person's i & lt imagines right chirokinesthetic EEG signal；

Step 5, seeks regularized covariance matrix

Introducing regularization parameter and β, span is α ∈ [0,1] and β ∈ [0,1], and α takes 0,0.001,0.01,0.1 respectively, 0.2；β takes 0,0.01,0.1,0.2,0.4,0.6 respectively, under the effect of regularization parameter, by the covariance matrix of experimenter it And combine with the covariance matrix sum of secondary examination person, constructing two classes average regularized covariance matrix, formula is as follows:

$Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_A+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_A}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

$Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=(1-\mathrm{\&alpha;})\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}+\frac{\mathrm{\&alpha;}}{N}tr\&lsqb;\frac{(1-\mathrm{\&beta;})\&CenterDot;R_B+\mathrm{\&beta;}\&CenterDot;{\hat{R}}_B}{(1-\mathrm{\&beta;})\&CenterDot;m+\mathrm{\&beta;}\&CenterDot;(n-1)\&CenterDot;m}\&rsqb;\&CenterDot;I$

Wherein,RepresentMark, I is the unit matrix of N × N, N is channel acquisition number；

Step 6, by the two classes average regularized covariance Matrix Calculating in step 5 and and carry out Eigenvalues Decomposition, solve canonical Whitening matrix, as follows:

$Z(\mathrm{\&alpha;},\mathrm{\&beta;})=Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})+Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})=\hat{U}\&CenterDot;\widehat{\mathrm{\&Lambda;}}\&CenterDot;{\hat{U}}^T$

Wherein,It is characterized value diagonal matrix,For characteristic of correspondence vector matrix, then canonical whitening matrix is:

$P={\widehat{\mathrm{\&Lambda;}}}^{(-\frac{1}{2})}\&CenterDot;{\hat{U}}^T$

Step 7, to the Z of gained in step 6_A(α, β) and Z_B(α, β) changes as follows:

${\stackrel{\&OverBar;}{Z}}_A(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_A(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_A\&CenterDot;{\mathrm{\&Lambda;}}_A\&CenterDot;{U_A}^T$

${\stackrel{\&OverBar;}{Z}}_B(\mathrm{\&alpha;},\mathrm{\&beta;})=P\&CenterDot;Z_B(\mathrm{\&alpha;},\mathrm{\&beta;})\&CenterDot;P^T=U_B\&CenterDot;{\mathrm{\&Lambda;}}_B\&CenterDot;{U_B}^T$

Wherein, Λ_AAnd Λ_BBeing characterized value diagonal matrix, UA and UB is characteristic of correspondence vector matrix, chooses diagonal matrix Λ_A、Λ_B Middle eigenvalue of maximum characteristic of correspondence vector, structure spatial filter is as follows:

W_A=U_A ^T·P

W_B=U_B ^T·P

Step 8, by two class EEG signal X of training sample_AAnd X_BThrough corresponding wave filter W_A、W_BHave:

F_A=W_A ^T·X_A

F_B=W_B ^T·X_B

Step 9, in step 8 through regularization common space mode filtering EEG signals calculate power spectral density, ask for Frequency, in the power spectral density value of 8～15Hz, utilizes structural learning dictionary B=[F_A F_B]；

Step 10, chooses one group of data in the training sample of experimenter successively as test sample y, filters by above step Ripple, projection, the test sample data after reservation process；

Step 11, the rarefaction representation solving test sample as the following formula is vectorial:

$\begin{array}{ccc}{\hat{x}}_1=min\left|\right|x|{|}_1& subjectto& \left|\right|\hat{y}-\hat{B}x|{|}_2\&le;\mathrm{\&epsiv;}\end{array}$

Wherein x is the rarefaction representation vector of test Mental imagery sample to be solved, test Mental imagery sample number to be solved for y According to, it is error threshold for ε；

Step 12, for Mental imagery i each time, according to the rarefaction representation vector of test sampleCalculate residual error

$r_i\left(\hat{y}\right)=\left|\right|\hat{y}-{\mathrm{BT}}_i\left({\hat{x}}_i\right)\left|\right|$

WhereinIt is by rarefaction representation vectorThe new vector obtained, the unit in this vector, corresponding to the i-th type games imagination Prime implicant is identical with corresponding element entry in rarefaction representation vector, and other element entries are zero；

Step 13, by the minimum classification of residual error as the recognition result of final Mental imagery classification:

It is test sample data.

3. Intellisense woollen blanket as claimed in claim 1, it is characterised in that the communication means of described wireless communication module includes Launch n road source signal, receiving terminal receives aliasing signal, piece-rate system separation multichannel aliasing signal；

After described transmitting n road source signal refers to the source signal channel mixing of n road, hybrid system is referred to as A, at transmitting terminal by n root Antenna is in spatial emission；

Described receiving terminal receives aliasing signal and refers to that receiving terminal utilizes m (m >=n ＞ 1) root sky bundle of lines aliasing signal to receive, Receiving signal and be referred to as observation signal, receiving terminal first carries out the pretreatment of observation signal, and pretreatment comprises two parts, i.e. centralization Process and spherization process；

Described piece-rate system separation multichannel aliasing signal refers to that piece-rate system W can be according to the difference of each road source signal information entropy Separate this multichannel aliasing signal in entropy territory, wherein the criterion of information entropy uses negentropy；

Wherein, the expression formula of negentropy approximate calculation is as follows:

Wherein, k_jFor some normal numbers, M is the gaussian variable with zero-mean, unit variance, function G_jFor non-quadratic function；

As all of G_jDuring=G, approximate expression becomes:

J_G(x)≈C[E{G(x)}-E{G(M)}]²(formula 2)

Wherein, G is any non-quadratic function, and C is a constant；

Then using above-mentioned formula to carry out the calculating of negentropy, the difference according to each road signal negentropy value can be the signal extraction of each road Out, it is achieved the multiplexing of channel.

4. Intellisense woollen blanket as claimed in claim 3, it is characterised in that described piece-rate system separation multichannel aliasing signal Due to use multi-path antenna launch and multi-path antenna receive MIMO technique, MIMO-EDM wireless communication system along with The increase of dual-mode antenna number, limit channels capacity also can increase the most linearly, and the MIMO-EDM wireless communication system limit is held The calculating process of amount is as follows:

Laguerre polynomial computation is utilized to obtain:

Wherein, m=min (N_t,N_r)

N=max (N_t,N_r)

For the Laguerre multinomial that number of times is k

If making λ=n/m, derive the channel capacity expression after following normalization；

Wherein,

$v_2={(\sqrt{\mathrm{\&tau;}}+1)}^2$

In the case of fast Rayleigh declines, make m=n=N_t=N_r, then v₁=0, v₂=4；

Progressive channel capacity is:

Utilize inequality:

log₂(1+x)≥log₂(x) (formula 6)

Formula (5) is reduced to: