CN107184205B - Automatic knowledge memory traction method based on memory scale and induction capture of brain - Google Patents

Automatic knowledge memory traction method based on memory scale and induction capture of brain Download PDF

Info

Publication number
CN107184205B
CN107184205B CN201710290775.0A CN201710290775A CN107184205B CN 107184205 B CN107184205 B CN 107184205B CN 201710290775 A CN201710290775 A CN 201710290775A CN 107184205 B CN107184205 B CN 107184205B
Authority
CN
China
Prior art keywords
memory
brain
data
knowledge
scale
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.)
Active
Application number
CN201710290775.0A
Other languages
Chinese (zh)
Other versions
CN107184205A (en
Inventor
吴一兵
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Easymonitor Technology Development Co ltd
Original Assignee
Beijing Easymonitor Technology Development Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Easymonitor Technology Development Co ltd filed Critical Beijing Easymonitor Technology Development Co ltd
Priority to CN201710290775.0A priority Critical patent/CN107184205B/en
Publication of CN107184205A publication Critical patent/CN107184205A/en
Application granted granted Critical
Publication of CN107184205B publication Critical patent/CN107184205B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A61B5/369—Electroencephalography [EEG]
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
    • A61B5/316—Modalities, i.e. specific diagnostic methods
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B5/00—Measuring for diagnostic purposes; Identification of persons
    • A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
    • A61B5/7235—Details of waveform analysis

Abstract

The invention discloses an automatic knowledge memory traction method based on memory scales and induction capture of a brain. The method comprises a wearable bioelectricity signal acquisition and transmission terminal and a computer or a server. The signal acquisition and transmission terminal acquires multi-lead brain wave signals in real time, transmits the signals to a computer, calculates the brain wave signals through a calculation algorithm program in the computer, extracts characteristic components related to brain memory processing in the brain waves in real time, takes dimensionless data expressed as 0-100 as quantitative scales of the memory state of the brain, converts the scales into stimulation input signals of natural senses of a human, performs closed-loop feedback induction and traction, captures the segment moment when the brain achieves the memory scale maximization in real time, and synchronously triggers the playing of knowledge through a multimedia unit of the computer, thereby realizing the bidirectional resonance of the memory state and knowledge display of the brain, and achieving the purpose of an automatic process of deep-etching, long-term and high-efficiency memory of the knowledge.

Description

Automatic knowledge memory traction method based on memory scale and induction capture of brain
Technical Field
The invention belongs to a working method of an electronic robot, and relates to a method and intelligent equipment for automatically acquiring knowledge, which are applied to the field of education and learning.
Background
Education and learning are the eternal topics of mankind and the impetus for human progress. In the modern times, brain science is the leading scientific field of important attention of people, and learning knowledge is one of the important functions of the brain. Since the birth of mankind, with the propagation of culture and the progress of science, the development of society creates the prosperity of modern industry and modern science, and a scientific, quantitative, normative and efficient technology and method for using the brain, especially the memory in learning, do not appear for thousands of years. The root of this reason is the laggard of brain science, people's understanding of the brain is very early, especially the working principle of the brain is always the blank of the scientific community.
To establish a scientific, quantitative and normative memory method, brain measurement technology must be developed to measure the memory state of the brain, and the current measurement aiming at the brain mainly focuses on image examination and blood flow measurement in the medical field, so that the problems of tumor and angiosclerosis are solved. The functional imaging can also be applied to basic research of memory function, but aiming at memory in learning, people can only measure the memory by subjective scale and complete psychological tasks, and the subjective scale and the psychological tasks are realized by subjective self-judgment and the completed task quality of people. The method is full of uncertainty and irreproducibility, has no real-time property, and cannot be applied to the memory process of daily learning. In recent decades, people have preliminary results on the measurement of brain functional states through electrophysiological techniques, and functional state indexes including sedation, pain, anxiety and the like can be extracted from the brain, and meanwhile, basic research of brain science also proves that effective components such as pain, sedation, emotion and the like exist in brain waves. Under the background, a quantitative method for scientifically and normatively memorizing knowledge can be developed.
Disclosure of Invention
The invention aims to provide an automatic knowledge memory traction method based on brain memory scale and induction capture, which aims to solve the technical problems of extracting characteristic components reflecting the pre-memory preparation state of the brain and quantitatively expressing (memory scale) by acquiring brain wave signals and using the memory scale as a feedback control signal on the basis of the method.
In order to achieve the purpose of the invention, the technical scheme adopted by the invention is as follows:
a brain-based memory scale and induction capture automatic knowledge memory traction method comprises the steps of acquiring brain waves under various environmental states and sleep and disease treatment states, extracting brain concentration memory quantitative scales based on the acquisition of the brain waves, and traction inducing and capturing the optimal learning state of the brain to realize the automatic knowledge memory method and equipment, discretizing and acquiring the brain waves into a computer by using a processing unit of the brain waves as original signals, applying various mathematical calculation algorithms such as wavelet, pattern recognition, power spectrum and the like, decomposing the brain waves into components covering low frequency, medium frequency, high frequency and ultrahigh frequency units under different scales, extracting characteristic components related to concentration memory in the components, expressing dimensionless data of 0-100 through normalization processing, reflecting the brain concentration memory quantitative scales in real time, and controlling the computer to output audio frequency by using the size of the memory scales, The change of the video signal and the computer extension output are used as stimulation signals, the stimulation signals are fed back to the sensory organs of the brain in real time to induce and draw the brain to quickly enter a learning state, when a special concentration memory state is captured, the computer presents knowledge content by sound and video, and the two states output by the computer are alternated to form a closed loop mode, so that the traction and induction processes of relaxation, entertainment and game are realized when the learning state is not in the learning state, and the learning and memory of the knowledge are realized when the memory state is in the concentration, so that the automatic memory of the knowledge is realized; the brain electricity sensing electrode is worn on the forehead and ear parts of the head, the potential signals of a plurality of parts of the brain are collected in a non-invasive way, the signals enter a computing unit through a preamplifier, a multistage integrated circuit and an analog-to-digital converter, the signal bandwidth comprises ultrahigh frequency components (brain wave components more than 30 hz) in brain waves, the brain electricity collecting part is set to be one to a plurality of lead parts, the brain wave signals of each lead part are amplified in a preposed way, converted in an analog-to-digital way, and directly transmitted to a computer system in a wired way or packaged and transmitted to an internet data server through a TCP/IP protocol of a wireless communication control unit, a special computing control software system in the computer or the server processes, displays, stores and forwards the received brain wave signals in real time, wherein the computing part mainly processes and computes by wavelet analysis, combining a mode recognition algorithm and a power spectrum algorithm, decomposing effective components and pseudo-differential components related to waveform dimension, waveform special points, waveform variation, frequency spectrum and the like in brain waves, extracting characteristic indexes of a preparation state before memory of the brain in real time, obtaining dimensionless data from 0 to 100 through normalization processing, using the dimensionless data as scales for objectively and quantitatively reflecting the memory capacity of the brain in real time, expressing and playing the scales in the form of external signals such as music, videos, toys and the like by a software system, feeding back the scales to a sensory channel of an auditory, visual or body position of the brain in real time, and controlling the expressions of the music, the videos and the toys by the scales and the frequency of the brain waves; setting a threshold value of a memory scale, immediately starting to play the knowledge content required to be memorized through listening and watching once a software system catches a pre-memory preparation state exceeding the threshold value, continuously playing contents such as music video games and the like once the software system is lower than the threshold value level, automatically circulating and reciprocating the process, editing the played knowledge content in a standard format, and automatically replacing the content according to the setting of the repetition times; in the learning and memorizing process, no operation action of a user exists, and all processes are automatically completed by the equipment.
The method for extracting real-time continuous quantitative scales reflecting the brain focusing on the memory state by the acquisition and the calculation processing of brain wave signals can be used as real-time feedback control signals to drive the processes of induction, traction and knowledge playing, and is characterized in that: the method is characterized in that a digital sequence of a section of brain wave signal in a fixed-length or variable-time window acquired in real time is utilized, a plurality of mathematical calculation algorithms are applied to the digital sequence, wavelet analysis calculation is taken as a main part, a pattern recognition algorithm and a power spectrum algorithm are combined, effective components and pseudo-differential components related to waveform dimension, waveform special points, waveform variation degree, frequency spectrum and the like in brain waves are decomposed, the current change characteristics of a real-time continuous brain memory state are extracted and used as scale expression of brain memory, the change precision and speed of scale values meet the requirements of real-time continuous control signals, and the algorithm covers nonlinear and linear numerical signal processing algorithms such as wavelet analysis, power spectrum analysis, pattern recognition and the like.
The method can be used as a scale for objectively and quantitatively reflecting the concentration and memory state of the brain in real time, and can control the change of external expression forms of music, videos, toys and the like which can be accepted by natural sense organs of human bodies, the change is used as an external feedback stimulation signal of the sense organs to induce and draw the change of the concentration and memory state of the brain, and the method is characterized in that: aiming at the feedback of the auditory system, the robot automatically reads the song names of the music files stored in the system in advance, the format is not limited, the number is determined by the computer capacity configured in the robot, according to the size of the memory scale, the robot automatically adjusts the played songs and the volume of the songs, the change of the songs is determined by a certain offline threshold value of the memory scale, the sequence of the played songs is determined by the state of the brain state under the songs, other quantitative data including the memory scale and the relaxation and concentration of the brain state, the change of the music and the volume are fed back to the auditory nervous system of the human, the reverse reflection of the system is caused, the change of the memory state of the brain area is triggered, namely the memory scale is changed, the volume is changed, the circulation is repeated, and the traction and the induction of the memory state are generated; aiming at the feedback of a visual system, the change of the size of the memory scale synchronously controls the process change of video playing, including the change of video display with knowledge content and the process change of an electronic game; the change of video knowledge content display is mainly presented by the brightness change of a screen, the frequency of the brightness change is synchronous with the size of a memory scale, and the brightness change and the memory scale form a synchronous action; the change of the electronic game process adopts a professional customized game, the real-time exchange of data is formed between the operation of the game and the robot software, and the result of the game is changed according to the change of the memory scale, so that the traction and the induction of the memory state are realized and strengthened in a game entertainment mode.
When the robot software system catches the preparation state before memory which exceeds the set threshold, the robot software system triggers the playing of preset knowledge content, and continues to play entertainment content such as music games and the like when the preset knowledge content is lower than the threshold, and the robot software system is characterized in that: the memory of knowledge is generated in a two-way resonance mechanism of knowledge playing and brain focusing on memory states, a computing software system of the robot computes memory scales of the brain in real time, compares preset memory state thresholds, switches and plays prestored knowledge contents when the preset memory state thresholds are exceeded, switches and plays music game contents when the preset memory state thresholds are lower than the preset threshold, and the memory state and the knowledge playing occur only when the preset memory state thresholds are exceeded, and circulates to form resonance; the set memory scale threshold value can be constant or can be changed in a self-adaptive manner, the memory state of the brain at present can be automatically adapted, the adaptive change is determined by the overlong time lower than the threshold value condition and the degree of deviation of the scale, the knowledge is presented according to a certain rule, and the knowledge content is edited and generated through a database or a character system of a computer or a server; the presentation of knowledge comprises auditory and visual expressions, the switching of auditory and visual expressions or the simultaneous expression is determined by opening eyes and closing eyes, the knowledge is played according to items, the playing of each item sets the repetition times, the repetition times can be constant or can be related to the self-adaptive change of a scale threshold and the occurrence of a sleep state; the functions of opening and closing eyes and recognizing the sleep state by the robot can be used as options, with or without, which is represented by that the repeated times of playing the knowledge content are fixed or variable; the playing of visual knowledge is matched with the flickering of a screen, so that the differentiated acceptance of a visual nervous system is triggered, and the visual fatigue is relieved.
In wavelet processing calculations, two bioelectric signal sequences can be decomposed for use as signature features of signal strength and interference signals.
Adopting a wavelet formula:
Figure GDA0002665955190000041
a, tau, which is the scaling factor and the translation parameter of the wavelet transform;
ω, the angular rating of the wavelet transform;
ψ(aω) Is the mother function of the wavelet transform;
x (w) is a wavelet transform result sequence of the original brain wave discrete sequence;
WTx is the time domain signal sequence after inverse transformation, i.e. the final result sequence;
for electroencephalogram vector group
bi(t)＝[x1 x2 x3 … xm-2 xm-1 xm]
i: number of brain wave leads, m: number of vector elements, x: electroencephalogram data, t: point in time
Real-time computing, decomposing wavelet basis function under each scale window by multi-scale filter bank algorithm
(Wb(2^j,wj(x)))j∈z
(Wb(2^0,w0(x)))，(Wb(2^1,w1(x)))…(Wb(2^N,wN(x)))
wj(x) Representing a set of resulting data sequences after wavelet transform, x representing a sequence number in the data sequence, z: a time domain space;
j: wavelet basis number (dimension number)
A group of time domain reconstruction functions are obtained by inverse transformation of wavelet basis functions and data of all scales as follows:
fi(t)＝∑Wb(2^i,wi)*Ψ2^i(x)
i: dimension
Ψ 2^0(x), Ψ 2^1 (x.. Ψ 2^ N (x): mother function scale wavelet data points
N: the order;
each reconstruction function represents the performance of brain waves, eye movement waves and muscle waves under different scales; the scale also corresponds to the frequency component of the signal, distributed in the conventional rhythm and high-frequency rhythm of the electroencephalogram; for each sequence data of the decomposed reconstruction function, extracting the characteristic points of the data by adopting a waveform recognition algorithm in a pattern recognition algorithm:
Tj(x)∈z；
t: eigenvalue vector, j: dimension, x: discrete data representing waveform characteristics; z: a time domain space;
the feature data in the vector tj (x) includes the results of the computation of the characteristic points, amplitudes, variations, slopes, areas, auto-correlation, and cross-correlation, and the computation is from the basic algorithm:
y(t，i)＝(fi(j)-fi(j-1))/Δt
j: discrete data subscript, i: dimension;
obtaining the maximum value in the sequence y (t, i) to obtain one of the characteristic indexes, wherein the positive and negative inversion points are special points, and the number of the special points is represented by the value of t;
to fi(x) The sequence data applies an iterative differential algorithm:
d(j，k)＝∑(fi(j+k)-fi(j+k-1)/(Δt+k))
k: delta of Δ t, from 1.. N, j: numerical serial numbers; i: dimension
For each vector in the matrix d (j, k), sorting and adding data points in the vector, and selecting the maximum sum in each vector as a slope and an amplitude;
for each scale vector f in the reconstruction functioni(x)＝[x1 x2 x3 … xm-2 xm-1 xm]Generating mode types omega 1, omega 2, … and omega c from the processed result vectors y (t, i) and d (j, k), and then calculating the distance of each reconstruction function by using the distance function between the modes; obtaining the variation, autocorrelation and cross-correlation values of the reconstruction function;
calculating an integral of a reconstruction function
Si(x)＝∫fi(x)*Δx
Obtaining the area value of each function;
Tj(x) The vector expresses the waveform characteristics and rules of the reconstruction function; for the acquired eye movement electric signals and frontal muscle electric signals in the brain waves, the eye movement electric signals and the frontal muscle electric signals are distributed in reconstruction functions with specific scales, and for the reconstruction functions, first, a first derivative of the reconstruction functions is obtained:
Di(x)＝(fi(x)-fi(x+m))/Δx
x: abscissa, Δ x: an abscissa increment;
to Di(x) Sorting to obtain maximum and minimum values, setting threshold Q (setting constant), and obtaining
Di(x) Obtaining a group of extreme point vectors by the positive and negative polarity change points:
Mi(j) (ii) a A high point;
mi(j) (ii) a A low point;
j: the number of extreme points;
and (3) adopting an integral algorithm for the correlation reconstruction function:
E＝∫fi(t)^2*Δt
obtaining the power values of myoelectricity and eye movement; aiming at the result data, through a normalization combination algorithm:
Sq＝exp(a*abs(Mi-mi)+b*E)；
a, b: weighting coefficients determined by the signal expression ranges;
quantitative expression Sq of signal intensity indicating interference can be obtained as a real-time identification condition of open eyes and closed eyes;
and (3) extracting the complexity characteristic value of the waveform power of each scale aiming at the reconstruction functions under different scales:
Figure GDA0002665955190000061
i, j: dimension
Sum of complexity:
Ai＝{Pi(j)}；
j: serial number of characteristic value
Eliminating the brain wave wavelet reconstruction component of eye movement and myoelectricity in the reconstruction function, adopting a power spectrum algorithm:
Figure GDA0002665955190000071
fi(x) The method comprises the following steps Brain wave wavelet reconstruction function, x (w): the magnitude of the spectral power;
various components of the power spectrum in brain waves, including the values of α β δ θ band, sef, mef edge frequency, dominant frequency value:
Fi＝{α，β，δ，θ…sef,mef}；
combining the characteristic vectors obtained by waveform identification to form a group of data vectors covering time domain, frequency domain and nonlinear complexity results:
Gi(x)＝{Tij，Fij}；
i: dimension; j: the serial number of the characteristic value.
Data vector Gi(x) A set, which is a primary processing result of the brain wave wavelet reconstruction function, named as a metadata set of brain wave primary processing, may be used as basic data of secondary calculation;
for a data sequence, the following calculation formula is obtained through data weighting:
E＝{c，d}*{Ai，Gi}；
c, d: a weighting coefficient;
applying a normalization calculation to the E data:
memory scale (exp (e)) x 100
Obtaining real-time objective quantitative characteristic indexes of the preparation state before memory in the brain waves.
The invention has the advantages that:
the method of the invention can be used for realizing the automatic control technology of the robot. The robot consists of a wearable bioelectrical signal acquisition and transmission terminal (a bioelectrical signal acquisition sensor, a signal amplifier, a single chip computing unit and a communication control unit), a computer or a server and a group of mathematical computation programs. The signal acquisition and transmission terminal acquires multi-lead brain wave signals in real time, transmits the signals to a computer through signal amplification, shaping, filtering, analog-to-digital conversion and communication, calculating brain wave signals through a calculation algorithm program in a computer, extracting characteristic components related to brain memory processing (preparation state before memory) in the brain waves in real time, taking dimensionless data expressed as 0-100 as quantitative scales of the memory state of the brain, converting the scales into stimulation input signals of natural senses of human, performing closed-loop feedback induction and traction, capturing the segment time when the brain reaches the maximization of the memory scales in real time, and through the multimedia unit of the computer, the playing of the knowledge is synchronously triggered, the memory state of the brain and the bidirectional resonance of knowledge display are realized, and the aim of the automatic process of deep, long-term and high-efficiency memory of the knowledge is fulfilled. The invention realizes the ability of automatic learning and memory in any state, in particular realizes the functions of inducing and drawing memory state, capturing and utilizing the segment of the memory processing state and automatically filling knowledge content in the entertainment state. The invention has important practical value for improving the learning and memory ability of knowledge, delaying the brain aging and the like.
Drawings
Description of the drawings:
FIG. 1 is a functional diagram of the automatic knowledge memory distraction method of the present invention.
FIG. 2 is a block diagram of the components of an embodiment of the automated knowledge memory distraction method of the present invention.
Detailed Description
The working principle of the invention is shown in figures 1 and 2.
A brain-based memory scale and induction capture automatic knowledge memory traction method comprises the steps of acquiring brain waves under various environmental states and sleep and disease treatment states, extracting brain concentration memory quantitative scales based on the acquisition of the brain waves, and traction inducing and capturing the optimal learning state of the brain to realize the automatic knowledge memory method and equipment, discretizing and acquiring the brain waves into a computer by using a processing unit of the brain waves as original signals, applying various mathematical calculation algorithms such as wavelet, pattern recognition, power spectrum and the like, decomposing the brain waves into components covering low frequency, medium frequency, high frequency and ultrahigh frequency units under different scales, extracting characteristic components related to concentration memory in the components, expressing dimensionless data of 0-100 through normalization processing, reflecting the brain concentration memory quantitative scales in real time, and controlling the computer to output audio frequency by using the size of the memory scales, The change of the video signal and the computer extension output are used as stimulation signals, the stimulation signals are fed back to the sensory organs of the brain in real time to induce and draw the brain to quickly enter a learning state, when a special concentration memory state is captured, the computer presents knowledge content by sound and video, and the two states output by the computer are alternated to form a closed loop mode, so that the traction and induction processes of relaxation, entertainment and game are realized when the learning state is not in the learning state, and the learning and memory of the knowledge are realized when the memory state is in the concentration, so that the automatic memory of the knowledge is realized; the brain electricity sensing electrode is worn on the forehead and ear parts of the head, the potential signals of a plurality of parts of the brain are collected in a non-invasive way, the signals enter a computing unit through a preamplifier, a multistage integrated circuit and an analog-to-digital converter, the signal bandwidth comprises ultrahigh frequency components (brain wave components more than 30 hz) in brain waves, the brain electricity collecting part is set to be one to a plurality of lead parts, the brain wave signals of each lead part are amplified in a preposed way, converted in an analog-to-digital way, and directly transmitted to a computer system in a wired way or packaged and transmitted to an internet data server through a TCP/IP protocol of a wireless communication control unit, a special computing control software system in the computer or the server processes, displays, stores and forwards the received brain wave signals in real time, wherein the computing part mainly processes and computes by wavelet analysis, combining a mode recognition algorithm and a power spectrum algorithm, decomposing effective components and pseudo-differential components related to waveform dimension, waveform special points, waveform variation, frequency spectrum and the like in brain waves, extracting characteristic indexes of a preparation state before memory of the brain in real time, obtaining dimensionless data from 0 to 100 through normalization processing, using the dimensionless data as scales for objectively and quantitatively reflecting the memory capacity of the brain in real time, expressing and playing the scales in the form of external signals such as music, videos, toys and the like by a software system, feeding back the scales to a sensory channel of an auditory, visual or body position of the brain in real time, and controlling the expressions of the music, the videos and the toys by the scales and the frequency of the brain waves; setting a threshold value of a memory scale, immediately starting to play the knowledge content required to be memorized through listening and watching once a software system catches a pre-memory preparation state exceeding the threshold value, continuously playing contents such as music video games and the like once the software system is lower than the threshold value level, automatically circulating and reciprocating the process, editing the played knowledge content in a standard format, and automatically replacing the content according to the setting of the repetition times; in the learning and memorizing process, no operation action of a user exists, and all processes are automatically completed by the equipment.
The method for extracting real-time continuous quantitative scales reflecting the brain focusing on the memory state by the acquisition and the calculation processing of brain wave signals can be used as real-time feedback control signals to drive the processes of induction, traction and knowledge playing, and is characterized in that: the method is characterized in that a digital sequence of a section of brain wave signal in a fixed-length or variable-time window acquired in real time is utilized, a plurality of mathematical calculation algorithms are applied to the digital sequence, wavelet analysis calculation is taken as a main part, a pattern recognition algorithm and a power spectrum algorithm are combined, effective components and pseudo-differential components related to waveform dimension, waveform special points, waveform variation degree, frequency spectrum and the like in brain waves are decomposed, the current change characteristics of a real-time continuous brain memory state are extracted and used as scale expression of brain memory, the change precision and speed of scale values meet the requirements of real-time continuous control signals, and the algorithm covers nonlinear and linear numerical signal processing algorithms such as wavelet analysis, power spectrum analysis, pattern recognition and the like.
The method can be used as a scale for objectively and quantitatively reflecting the concentration and memory state of the brain in real time, and can control the change of external expression forms of music, videos, toys and the like which can be accepted by natural sense organs of human bodies, the change is used as an external feedback stimulation signal of the sense organs to induce and draw the change of the concentration and memory state of the brain, and the method is characterized in that: aiming at the feedback of the auditory system, the robot automatically reads the song names of the music files stored in the system in advance, the format is not limited, the number is determined by the computer capacity configured in the robot, according to the size of the memory scale, the robot automatically adjusts the played songs and the volume of the songs, the change of the songs is determined by a certain offline threshold value of the memory scale, the sequence of the played songs is determined by the state of the brain state under the songs, other quantitative data including the memory scale and the relaxation and concentration of the brain state, the change of the music and the volume are fed back to the auditory nervous system of the human, the reverse reflection of the system is caused, the change of the memory state of the brain area is triggered, namely the memory scale is changed, the volume is changed, the circulation is repeated, and the traction and the induction of the memory state are generated; aiming at the feedback of a visual system, the change of the size of the memory scale synchronously controls the process change of video playing, including the change of video display with knowledge content and the process change of an electronic game; the change of video knowledge content display is mainly presented by the brightness change of a screen, the frequency of the brightness change is synchronous with the size of a memory scale, and the brightness change and the memory scale form a synchronous action; the change of the electronic game process adopts a professional customized game, the real-time exchange of data is formed between the operation of the game and the robot software, and the result of the game is changed according to the change of the memory scale, so that the traction and the induction of the memory state are realized and strengthened in a game entertainment mode.
When the robot software system catches the preparation state before memory which exceeds the set threshold, the robot software system triggers the playing of preset knowledge content, and continues to play entertainment content such as music games and the like when the preset knowledge content is lower than the threshold, and the robot software system is characterized in that: the memory of knowledge is generated in a two-way resonance mechanism of knowledge playing and brain focusing on memory states, a computing software system of the robot computes memory scales of the brain in real time, compares preset memory state thresholds, switches and plays prestored knowledge contents when the preset memory state thresholds are exceeded, switches and plays music game contents when the preset memory state thresholds are lower than the preset threshold, and the memory state and the knowledge playing occur only when the preset memory state thresholds are exceeded, and circulates to form resonance; the set memory scale threshold value can be constant or can be changed in a self-adaptive manner, the memory state of the brain at present can be automatically adapted, the adaptive change is determined by the overlong time lower than the threshold value condition and the degree of deviation of the scale, the knowledge is presented according to a certain rule, and the knowledge content is edited and generated through a database or a character system of a computer or a server; the presentation of knowledge comprises auditory and visual expressions, the switching of auditory and visual expressions or the simultaneous expression is determined by opening eyes and closing eyes, the knowledge is played according to items, the playing of each item sets the repetition times, the repetition times can be constant or can be related to the self-adaptive change of a scale threshold and the occurrence of a sleep state; the functions of opening and closing eyes and recognizing the sleep state by the robot can be used as options, with or without, which is represented by that the repeated times of playing the knowledge content are fixed or variable; the playing of visual knowledge is matched with the flickering of a screen, so that the differentiated acceptance of a visual nervous system is triggered, and the visual fatigue is relieved.
The invention relates to a wearable wireless brain wave signal acquisition sensor and a computer, or a mobile communication terminal and a cloud computing server which are equivalent to the computer, and related signal processing, computing and control management software. The brain wave characteristic component extraction method comprises a multi-lead brain wave acquisition and processing part, a noninvasive electrode, a communication control module, a brain wave characteristic component calculation and extraction method and a normalization expression mode of a preparation state before memory, a standardized structure of knowledge needing to be memorized, multimedia output and other software and hardware parts.
The high-level nervous system of brain is composed of neuron cells, the discharge of neuron is the expression of nerve conduction, and the potential signal of nerve conduction reaches the surface of head skin, namely the brain wave signal. The brain wave signals contain various components of the brain state, and are also a physical source of cognitive memory. The collected brain wave signals are utilized, a plurality of mathematical calculation algorithms are applied to the brain wave signals through a group of algorithm programs configured in advance on a computer, characteristic indexes of memory processing components in the brain wave signals are found and extracted, and the characteristic indexes can be used as quantitative scales of a preparation state before memory after normalization processing, so that the automatic knowledge memory traction robot is formed.
The wearable wireless brain wave signal acquisition sensor comprises a preamplifier circuit, a singlechip control circuit, a combined wireless communication access control circuit, a power supply circuit and the like. The connection relationship among the parts is as follows:
the multi-lead brain wave signals are collected and transmitted to the filtering, noise control and amplification input part of the preamplifier circuit through brain wave receptor electrodes (bioelectricity signal sensors) of a user, and are respectively transmitted to the analog-digital conversion circuit of the singlechip computer control part of the sensor through corresponding channels. The single chip computer control part obtains digital brain wave (including high-frequency brain wave over 30 hz) data, and the data is encrypted and compressed to obtain a processed data stream which is sent to a storage queue. Under the control of a reading instruction of the single chip control circuit, data in the queue is output to the Bluetooth module through the data port and sent, and after receiving Bluetooth communication data, a computer or a mobile communication terminal (mobile phone) synchronously pushes electroencephalogram data to the mobile internet through the wifi, 3G or 4G port. If a computer (or a mobile panel) is adopted, an algorithm program and a management control program configured in the computer process and calculate the received data of the electroencephalogram sensor in real time, capture and induce the memory state, and complete all functions of the robot. If a mobile communication terminal (mobile phone) is adopted, all tasks of calculation and management control of brain wave data are completed through a cloud computing server arranged on an internet platform, and the mobile communication terminal completes the presentation of results including music and videos. The data receiving and control instructions of the cloud computing server are all interactive on a mobile internet platform, and each electroencephalogram sensor, including a configured computer or a mobile communication terminal, has a unique device number, and corresponds to an address and a port of the server.
The unique device number consists of the machine number, the server number plus the fixed IP address of the server:
the EEG sensor equipment number (including configured computer or mobile phone) is the unique address number within the range of server number + fixed IP address of networking server + server port number
65535 corresponding to each cloud computing server
The wearable wireless brain wave sensor performs encryption processing and compression on data converted into digital signals. The data window is 7-15 beta
The L stream window is m + addr + asyn + data
m is an encryption mark transmitted by the module, addr is a machine address, asyn is synchronization, and data is encryption data comprising numerical data and waveform data.
When a computer at a short distance is used as a processing platform of the electroencephalogram sensor, the sensor and the computer form the main part of the robot. The computer needs to download the special robot system software, music tracks, game videos and knowledge contents, and during downloading, the identity information of the user needs to be established, and the identity card number is used as the unique user identification code. The sensor data directly receives data through a wifi or Bluetooth communication port configured on the computer, and an operating system and robot application software of the computer complete all work of data calculation, memory induction capture and knowledge learning. Meanwhile, the computer synchronously sends the original brain wave data to the Internet platform, the set cloud computing server receives the data to form machine learning big data, all functions of the robot are synchronously completed, and the result data are sent to the Internet platform in the same encryption format for relevant personnel to share and monitor the results on different terminals at different geographic positions.
Aiming at mobile application, each mobile communication terminal (mobile phone) receives real-time data of an electroencephalogram sensor, the data are synchronously transmitted to a cloud computing server in a mobile internet area through wifi, 3G and 4G ports by APP software configured on the mobile phone, and the server runs a special robot function software system such as real-time data computing processing software, induced traction capture control software, user identity synchronous management software, knowledge content editing software and the like. The server completes the functions of data calculation, identity matching, original and result data storage, forwarding, command generation, big data formation and the like, and the result data and the control command are returned to the mobile communication terminal (mobile phone) through the mobile internet in the same encryption format. And when the APP software in the mobile phone receives data and commands, the functions of presenting sound and video are completed. Meanwhile, the holders of other terminals distributed on the Internet share and monitor the automatic learning and memorizing result under the condition of identity authorization.
One of the mobile communication terminals is a mobile phone. The mobile phone receives the encrypted data stream sent by the electroencephalogram sensor through the Bluetooth function configured on the mobile phone and under the support of the APP software of the robot. And the data is synchronously transmitted to the Internet platform in real time through the processes of displaying, calling and forwarding. The method comprises the steps of setting Bluetooth communication between a mobile phone and a sensor, matching, adopting a Socket thread in a mobile phone APP, initializing a Bluetooth channel, entering a receiving flow, decoding received data signals, and selectively displaying the data signals on a mobile phone screen in real time. Meanwhile, for the decoded data, the cloud computing server address and the port address are extracted, and the received original data is sent to the corresponding cloud computing server without any processing through a WiFi (wireless fidelity) interface, a 3G interface or a 4G interface of the mobile phone. The APP receives command data and result data sent by the corresponding server by using another Socket thread, and the decryption calculation format of the data is the same as the data calculation format sent by the sensor. And for the result data, the APP displays the data at the fixed screen position of the mobile phone in real time, so that a user can observe the quantitative scale of the memory state in real time. For command data, the APP controls the playing of music, video and knowledge needing to be memorized according to different commands. The content of music tracks, game videos and knowledge, and APP software, all need to be downloaded from the server. The identification number is used as the unique user identification code, the personal file of the user is established while downloading, and once the user information is generated, the mobile phone APP software can become a part of the robot.
The brain wave signal is calculated by utilizing a calculation algorithm program, characteristic components related to brain memory processing (preparation state before memory) in the brain waves are extracted in real time, dimensionless data expressed as 0-100 are used as quantitative scales of the memory state of the brain, the scales are converted into stimulation input signals of natural senses of people, closed-loop feedback induction and traction are performed, the segment moment when the brain reaches the memory scale maximization is captured in real time, the playing of knowledge is synchronously triggered through a multimedia unit of a computer, the two-way resonance of the memory state and knowledge display of the brain is realized, and the automatic process of deep-etching, long-term and efficient memory of the knowledge is achieved. The memory scale is expressed as a stimulation input signal of a natural sense organ of a human, the game entertainment instinct of the human is utilized, the scale is displayed in the forms of listening, seeing and feeling of music, toys, electronic competitions and the like, and the brain is naturally stimulated to generate a preparation state before memory under the background of pleasure, joy and concentration. The scale size controls the change of music, video pictures and electronic games, the change forms the stimulation input of sense organs, the preparation state before the memory of the brain is adversely affected, the scale change causes the stimulation change, and the process of closed-loop feedback induction and traction is repeated continuously when the scale change is caused. The expression of memory scale, also the capture of the memory processing (preparation state before memory) state of brain, as long as the machine captures the scale of memory processing and reaches certain condition (threshold), suspend the production of the stimulus signal immediately, broadcast the knowledge content that needs to be memorized repeatedly in order, lower than the condition, continue to start the emergence of the stimulus signal, the human natural sense organ accepts the information such as music, drawing, game that can make people happy again, two kinds of states switch over at random, the determinant factor is the change of the memory scale, once the scale controls the broadcast of knowledge to begin, the brain prepares the resonance that state and knowledge display before memory, knowledge in brain is in the best, most efficient state. The whole process is realized completely, namely a factory for automatically producing knowledge memory is established in the brain, external knowledge is continuously converted into deep and long-term memory in the brain, meanwhile, the input of stimulation signals of music, games and the like acts on the brain, the optimal process of dormancy and fatigue recovery of a brain nervous system is realized, automatic memory and fatigue recovery are alternated, so that the robot and the human brain form an organic whole, and the automatic knowledge memory traction robot and the human brain form the automatic knowledge memory traction robot together.
After receiving real-time brain wave data, a computer or a cloud computing server applies various mathematical algorithms and control processes, wherein a computing part mainly analyzes and calculates wavelets aiming at a digital sequence of a section of brain wave signals in a fixed-length or variable-time window acquired in real time, combines a mode recognition algorithm and a power spectrum algorithm, decomposes effective components and pseudo-differential components related to waveform dimensionality, waveform special points, waveform variation, frequency spectrum and the like in brain waves, decomposes the brain waves into components covering low-frequency, medium-frequency, high-frequency and ultrahigh-frequency units in different scales, extracts characteristic indexes of a brain preparation state before memory in real time, obtains dimensionless data from 0 to 100 through normalization processing, and uses the dimensionless data as scales for objectively and quantitatively reflecting the memory capacity of the brain in real time. The software system then expresses and plays the scales in the form of external signals such as music, video, toys and the like, the content of the music and the video changes along with the size of the scales to form a stimulation signal, and the stimulation signal is fed back to a sensory channel of the hearing, the vision or the body position of the brain in real time to induce and draw the brain to quickly enter a state of concentration and memory. The expression of music, video and toy is controlled by the scale and the frequency of brain wave, and the presentation form is different along with the difference of the numerical value, thus deepening the response of brain. The software system calculates the size of the memory scale in real time and compares the size with a preset scale threshold value, and once the software system captures the scale value of the preparation state before memory which exceeds the threshold value, the software system immediately starts to play the knowledge content needing to be memorized through audios and videos. Once the memory scale value is lower than the threshold level, the contents such as music video games and the like are continuously played, and the memory traction inducing process is started. The alternation of two states output by the computer forms a closed-loop mode, when the computer is not in a learning state, the computer relaxes, entertains, and draws and induces games, when the computer is in a memory state, the learning and the memory of knowledge are realized, the automatic memory process of the knowledge is automatically circulated and reciprocated, the played knowledge content is edited in a standard format, and is automatically changed according to the setting of the repeated times. In the learning and memorizing process, no operation action of a user exists, and all processes are automatically completed by the equipment.
The scale which can objectively and quantitatively reflect the concentration and memory state of the brain in real time can be used as a feedback control signal for controlling the audio-visual signal change of external expression forms which can be accepted by the natural sense organs of the human body, such as music, video, toys, and the like. This change acts as an external feedback stimulus to the sensory organs, inducing a change in the state of concentration and memory of the drawn brain. The software system in the robot can automatically read the song names of the music files stored in the system in advance, the format is not limited, and the number is determined by the computer capacity configured in the robot. According to the size of the memory scale, the robot automatically adjusts the size of the played music and the volume of the music, changes the music, is determined by a certain offline threshold value of the memory scale, plays the sequence of the music, is determined by the state of the brain state under the music, and contains other quantitative data of the memory scale and the brain state, such as relaxation, concentration and the like. The change of music song and volume is fed back into auditory nerve system of human body to make system retroreflex, and can trigger the change of memory state of brain region, i.e. change of memory scale and control volume change, and can produce traction and induction of memory state. The system comprises video game files, and synchronously controls the process change of video playing according to the feedback of a visual system and the size change of memory scales, wherein the process change comprises the change of video display with knowledge content and the process change of the electronic game. The change of video knowledge content display is mainly presented by the brightness change of a screen, the frequency of the brightness change is synchronous with the size of the memory scale, and the two display functions synchronously. The change of the electronic game process adopts a professional customized game, the real-time exchange of data is formed between the operation of the game and the robot software, and the result of the game is changed according to the change of the memory scale, so that the traction and the induction of the memory state are realized and strengthened in a game entertainment mode.
When the scale change of the preparation state before the memory which exceeds the set threshold value is captured by the robot software system, the preset knowledge content is triggered to be played, and the entertainment content such as the music game and the like is continuously played when the scale change is lower than the threshold value. The memory of knowledge is generated by a mechanism of bidirectional resonance of knowledge playing and brain focusing on a memory state, when the brain is in a high memory state, the playing of the knowledge reaches the maximum volume, and the maximization of the two phenomena is the resonance of two signals, so that the brain is in a state of the optimal learning effect. The robot's calculation software system calculates the memory scale of brain in real time, compares the threshold value of the memory state that is set up in advance, exceeds the threshold value, switches to broadcast the knowledge content that is stored in advance, is less than the threshold value, switches to broadcast the content of music game, the memory state appears, the knowledge broadcast appears, the circulation is reciprocal, form the learning mode of automatic memory. The set memory scale threshold value can be constant or can be adaptively changed, and the memory scale threshold value is automatically adapted to the current brain memory state. The adaptation is determined by the length of time below the threshold condition and the degree of scale drift. The knowledge content is edited and generated through a database or a word system of a computer or a server according to certain rules. The presentation of the knowledge comprises auditory and visual expressions, switching between auditory and visual expressions or simultaneous expressions is determined by eyes opening and eyes closing, the knowledge is played according to items, the playing of each item sets the repetition times, the repetition times can be constant or can be related to the adaptive change of the scale threshold and the occurrence of the sleep state. The functions of opening and closing eyes and recognizing the sleep state by the robot may be used as options, with or without, in that the number of repetitions of the playing of the knowledge content is fixed or variable. In the adaptive state, the lower the scale threshold, the more times the repetition is required. The playing of visual knowledge is matched with the flickering of a screen, so that the differentiated acceptance of a visual nervous system is triggered, and the visual fatigue is relieved.
For brain wave processing, calculating different component sequences of the brain and time domain characteristics and multi-scale complexity characteristics of each sequence by adopting waveform identification and wavelet analysis algorithms through discretization processing with the sampling frequency of 1400/s, the sampling time window of 1.25s and the sampling precision of 10-bit. And extracting rhythm components in the brain waves and the power of each rhythm wave band by adopting a power spectrum algorithm to obtain a calculation result sequence of primary processing calculation, wherein the sequence corresponds to hundreds of independent or weakly related regular characteristics in the brain waves. The brain waves collected from both sides of the forehead include an eye movement electric component and a frontal muscle electric component. In wavelet processing calculations, two bioelectric signal sequences can be decomposed for use as signature features of signal strength and interference signals.
Adopting a wavelet formula:
Figure GDA0002665955190000151
a, tau, which is the scaling factor and the translation parameter of the wavelet transform;
ω, the angular rating of the wavelet transform;
ψ(aω) Is the mother function of the wavelet transform;
x (w) is a wavelet transform result sequence of the original brain wave discrete sequence;
WTx is the time domain signal sequence after inverse transformation, i.e. the final result sequence;
for electroencephalogram vector group
bi(t)＝[x1 x2 x3 … xm-2 xm-1 xm]
i: number of brain wave leads, m: number of vector elements, x: electroencephalogram data, t: the real-time calculation processing of the time points is carried out, and wavelet basis functions (Wb (2^ j, w) under each scale window are decomposed by a multi-scale filter bank algorithmj(x)))j∈z(Wb(2^0,w0(x)))，(Wb(2^1,w1(x)))…(Wb(2^N,wN(x)))wj(x) Representing a set of resulting data sequences after wavelet transform, x representing a sequence number in the data sequence, z: a time domain space;
j: wavelet basis number (dimension number)
A group of time domain reconstruction functions are obtained by inverse transformation of wavelet basis functions and data of all scales as follows:
fi(t)＝∑Wb(2^i,wi)*Ψ2^i(x)
i: dimension
Ψ 2^0(x), Ψ 2^1 (x.. Ψ 2^ N (x): mother function scale wavelet data points
N: the order;
each reconstruction function represents the performance of brain waves, eye movement waves and muscle waves under different scales; the scale also corresponds to the frequency component of the signal, distributed in the conventional rhythm and high-frequency rhythm of the electroencephalogram; for each sequence data of the decomposed reconstruction function, extracting the characteristic points of the data by adopting a waveform recognition algorithm in a pattern recognition algorithm:
Tj(x)∈z；
t: eigenvalue vector, j: dimension, x: discrete data representing waveform characteristics; z: a time domain space; the feature data in the vector tj (x) includes the results of the computation of the characteristic points, amplitudes, variations, slopes, areas, auto-correlation, and cross-correlation, and the computation is from the basic algorithm:
y(t，i)＝(fi(j)-fi(j-1))/Δt
j: discrete data subscript, i: dimension;
obtaining the maximum value in the sequence y (t, i) to obtain one of the characteristic indexes, wherein the positive and negative inversion points are special points, and the number of the special points is represented by the value of t;
to fi(x) The sequence data applies an iterative differential algorithm:
d(j，k)＝∑(fi(j+k)-fi(j+k-1)/(Δt+k))
k: delta of Δ t, from 1.. N, j: numerical serial numbers; i: dimension
For each vector in the matrix d (j, k), sorting and adding data points in the vector, and selecting the maximum sum in each vector as a slope and an amplitude;
for each scale vector f in the reconstruction functioni(x)＝[x1 x2 x3 … xm-2 xm-1 xm]Generating mode types omega 1, omega 2, … and omega c from the processed result vectors y (t, i) and d (j, k), and then calculating the distance of each reconstruction function by using the distance function between the modes; obtaining a variation, self-phase of the reconstruction functionCorrelation values;
calculate the integral of the reconstruction function:
Si(x)＝∫fi(x)*Δx
obtaining the area value of each function;
Tj(x) The vector expresses the waveform characteristics and rules of the reconstruction function; for the acquired eye movement electric signals and frontal muscle electric signals in the brain waves, the eye movement electric signals and the frontal muscle electric signals are distributed in reconstruction functions with specific scales, and for the reconstruction functions, first, a first derivative of the reconstruction functions is obtained:
Di(x)＝(fi(x)-fi(x+m))/Δx
x: abscissa, Δ x: an abscissa increment;
to Di(x) Sorting to obtain maximum and minimum values, setting threshold Q (setting constant), and obtaining
Di(x) Obtaining a group of extreme point vectors by the positive and negative polarity change points:
Mi(j) (ii) a A high point;
mi(j) (ii) a A low point;
j: the number of extreme points;
and (3) adopting an integral algorithm for the correlation reconstruction function:
E＝∫fi(t)^2*Δt
obtaining the power values of myoelectricity and eye movement; aiming at the result data, through a normalization combination algorithm:
Sq＝exp(a*abs(Mi-mi)+b*E)；
a, b: weighting coefficients determined by the signal expression ranges;
quantitative expression Sq of signal intensity indicating interference can be obtained as a real-time identification condition of open eyes and closed eyes;
and (3) extracting the complexity characteristic value of the waveform power of each scale aiming at the reconstruction functions under different scales:
Figure GDA0002665955190000181
i, j: dimension
Sum of complexity:
Ai＝{Pi(j)}；
j: serial number of characteristic value
Eliminating the brain wave wavelet reconstruction component of eye movement and myoelectricity in the reconstruction function, adopting a power spectrum algorithm:
Figure GDA0002665955190000182
fi(x) The method comprises the following steps Brain wave wavelet reconstruction function, x (w): the magnitude of the spectral power;
various components of the power spectrum in brain waves, including the values of α β δ θ band, sef, mef edge frequency, dominant frequency value:
Fi＝{α，β，δ，θ…sef,mef}；
combining the characteristic vectors obtained by waveform identification to form a group of data vectors covering time domain, frequency domain and nonlinear complexity results:
Gi(x)＝{Tij，Fij}；
i: dimension; j: the serial number of the characteristic value.
Data vector Gi(x) A set, which is a primary processing result of the brain wave wavelet reconstruction function, named as a metadata set of brain wave primary processing, may be used as basic data of secondary calculation;
for a data sequence, the following calculation formula is obtained through data weighting:
E＝{c，d}*{Ai，Gi}；
c, d: a weighting coefficient;
applying a normalization calculation to the E data:
memory scale (exp (e)) x 100
Obtaining real-time objective quantitative characteristic indexes of the preparation state before memory in the brain waves.

Claims (3)

1. An automatic knowledge memory traction method based on brain memory scale and induction capture extracts quantitative scales of brain for memory based on collected brain waves, and draws induction and captures the optimal learning state of the brain to realize automatic memory of knowledge, and is characterized in that: the brain wave is used as an original signal to be processed, the brain wave is discretely collected into a computer or a server, wavelet, pattern recognition and power spectrum mathematical calculation analysis are applied, the brain wave is decomposed into components with low frequency, medium frequency, high frequency and ultrahigh frequency under different scales, characteristic components related to concentration and memory are extracted, the components are expressed into 0-100 dimensionless data through normalization processing, quantitative scales of the concentration and memory of the brain are reflected in real time, game change or toy actions output by the computer are controlled by the size of the scales of the concentration and memory are used as stimulus signals, a game presentation mode comprises audio frequency and video, the stimulus signals are fed back to sensory organs of the brain in real time, the brain is induced to be dragged to rapidly enter a learning state, and when a concentration and memory state with a specific degree is captured, the computer presents knowledge contents by sound and video, and the two states output by the computer are alternated, a closed-loop mode is formed, when the learning state is not reached, the relaxation traction and induction are carried out, when the memory state is absorbed, the knowledge is learned and memorized, and the automatic memory of the knowledge is realized; the brain wave sensing electrodes are worn on the forehead and ears of the head, lead potential signals of a plurality of parts of the brain are collected noninvasively, the signal bandwidth comprises ultrahigh frequency components exceeding 30Hz in brain waves, the brain wave signals of each lead enter a single chip microcomputer through preamplification and analog-to-digital conversion, and are directly transmitted to a computer system in a wired mode or are packaged and transmitted to an internet data server through a TCP/IP protocol of a wireless communication control unit, and a calculation control software system in the computer or the server processes, displays, stores and forwards the received brain wave signals in real time, wherein a calculation part mainly adopts wavelet analysis and calculation, combines a mode identification algorithm and a power spectrum algorithm, decomposes waveform dimensions in the brain waves, extracts waveform extreme points, the slope of the waveform, brain electrical frequency spectrum, eye electrical and muscle electrical frequency spectrum, and extracts characteristic indexes of a preparation state before memory of the brain in real time, through normalization processing, obtaining 0-100 dimensionless data as a scale for objectively and quantitatively reflecting the concentration memory ability of the brain in real time, expressing the scale by a software system through game change or toy action, feeding back the scale to an auditory and visual sensory channel of the brain in real time, and controlling the expression of the game and the toy by the concentration memory scale; setting a threshold value of the memory scale, playing the knowledge content to be memorized immediately once the software system catches a pre-memory preparation state exceeding the threshold value, continuously outputting the content of the game or toy action once the software system is lower than the threshold value level, automatically and circularly repeating the process, editing the played knowledge content in a standard format, and automatically replacing the played knowledge content according to the setting of the repetition times; in the wavelet processing calculation, two bioelectric signal sequences of eye movement electric waves and myoelectric waves are decomposed and used as the signal intensity and the marking characteristics of interference signals;
adopting a wavelet formula:
Figure DEST_PATH_IMAGE002
Figure DEST_PATH_IMAGE004
respectively, a scaling factor and a translation parameter of the wavelet transform;
Figure DEST_PATH_IMAGE006
is the angular frequency of the wavelet transform;
Figure DEST_PATH_IMAGE008
is the mother function of the wavelet transform;
Figure DEST_PATH_IMAGE010
is the wavelet transform result sequence of the original brain wave discrete sequence;
Figure DEST_PATH_IMAGE012
is a time domain signal sequence after inverse transformation;
for electroencephalogram vector group
bi1（t）= [ y'1 y'2 y'3 … y'm’-2 y'm’-1 y'm’ ]
i 1: the number of brain wave leads;
m': the number of brain wave data;
y': a brain wave data sequence;
t: a point in time;
real-time computing, decomposing wavelet basis function under each scale window by multi-scale filter bank algorithm
(Wb(2^i,wi（r）)) i∈z
(Wb(2^0, w0（r）))，(Wb(2^1, w1（r）))…(Wb(2^N, wN（r）))
wi(r): represents a set of resulting data sequences after wavelet transformation;
r: representing a sequence number in the data sequence;
z: a time domain space;
i: dimension;
a group of time domain reconstruction functions are obtained by inverse transformation of wavelet basis functions and data of all scales as follows:
ｆi(t)=∑Wb(2^i,wi)*Ψ2^i(u)
i: dimension
Wb: a wavelet basis function;
wi: the result data after wavelet transform of each scale;
Ψ2^0(u), Ψ2^1(u)... Ψ2^N(u): a mother function scale wavelet data point;
n: the order;
u: a data sequence number of a scale wavelet data point;
t: a point in time;
each wavelet reconstruction function represents the performances of brain waves, eye movement electric waves and muscle electric waves under different scales; the scale also corresponds to the frequency component of the signal, distributed in the conventional rhythm and high-frequency rhythm of the electroencephalogram; for each sequence data of the decomposed wavelet reconstruction function, extracting the characteristic points of the data by adopting a waveform recognition algorithm in a pattern recognition algorithm:
Ti（h）∈ z ；
t: a vector of eigenvalues;
i: dimension;
h: a characteristic value data sequence number;
z: a time domain space;
vector Ti(h) The characteristic data in (2) comprises calculation results of specific points, amplitude, variation, slope, area, autocorrelation and cross-correlation, and the calculation is from a basic algorithm:
calculating a data sequence:
y（j'，i）= （fi（j'）-fi（j'-1））/ Δt'
fi(j') is fi(t) the discrete-time data sequence;
j': time discrete data subscripts;
i: dimension;
Δ t': is the time interval between discrete points j '-1 and j';
acquiring the maximum value in the sequence y (j ', i), wherein the positive and negative inversion points are special points, and the number of the special points is determined by the value of delta t';
to fi(j') applying an iterative differentiation algorithm to the sequence data:
Figure DEST_PATH_IMAGE014
Δt1': is a discrete point j2+ k-1 and j2A time interval between + k;
k：Δt1the increment of 'from 1 to N'; n' is a constant;
j2: numerical serial numbers of time points;
i: dimension;
for matrix d (j)2Respective vector in k)Sorting and adding the data points in the vectors, and selecting the maximum value and the sum of the data points in each vector as the slope and the amplitude;
aiming at the reconstruction function, the processed result vectors y (j', i) and d (j) are processed2K) generating mode types omega 1, omega 2, …, and omega c, and then calculating the distance of each wavelet reconstruction function by using the distance function between the modes; obtaining the variation, autocorrelation and cross-correlation values of the wavelet reconstruction function;
calculating a wavelet reconstruction function fiIntegration of (t):
Si= ∫fi（t）*dt
t: time;
i: dimension;
obtaining the area value of each function;
Ti(h) the vector expresses the waveform characteristics and rules of the wavelet reconstruction function; distributing eye movement electric signals and frontal muscle electric signals in collected brain waves in wavelet reconstruction functions of specific scales, and aiming at the wavelet reconstruction functions f'i(t), first, the first derivative is found:
Di（j2'）=（f'i（j2'）-f'i（j2'+Δt2））/Δt2
f'i（j2') is f'i(t) the discrete-time data sequence;
i: dimension;
j2': time discrete data subscripts;
Δt2: is a discrete point j2' -1 and j2The time interval of';
to Di（j2') sorting to obtain maximum and minimum values, setting threshold Q, and obtaining Di（j2') to obtain a set of extreme point vectors:
Mi(t 1): a high point;
mi(t 2): a low point;
i: dimension;
t1, t 2: the time point of the extreme value;
adopting an integral algorithm for wavelet reconstruction functions related to eye movement and myoelectricity:
SS =∫f'i（t）^2 * dt
obtaining the power values of myoelectricity and eye movement; aiming at the result data, through a normalization combination algorithm:
Sq=exp(a’*abs（Mi-mi）+b’*SS)；
a 'and b' are weighting coefficients which are constants and are determined by the signal expression range;
sq indicates quantitative expression of signal intensity of interference as one of display results;
and (3) extracting the complexity characteristic value of waveform power of each scale aiming at the wavelet reconstruction function under different scales:
Figure DEST_PATH_IMAGE016
fi(t): a wavelet reconstruction function;
i: dimension
j4: a sequence number of the complexity data;
t: a point in time;
t3, t 5: a time window starting point;
t4, t 6: a time window end point;
sum of complexity:
Figure DEST_PATH_IMAGE018
i: dimension;
j4: a sequence number of the complexity data;
wavelet reconstruction function with eye movement and myoelectricity eliminated from reconstruction function
Figure DEST_PATH_IMAGE020
And adopting a power spectrum algorithm:
Figure DEST_PATH_IMAGE022
t: a point in time;
Figure DEST_PATH_IMAGE024
: angular frequency of the power spectrum;
i: dimension;
obtaining the components of the power spectrum in the reconstructed brain waves:
Fi={α，β，δ，θ，sef，mef}；
i: dimension;
α, β, δ, θ: power percentage of each band;
sef: edge frequencies of the power spectrum calculation;
mef: center frequency of power spectrum calculation;
combining the feature vectors obtained by waveform identification to form a group of feature data covering time domain and frequency domain:
Gi（j3）={Ti，Fi}；
i: dimension;
j3: a characteristic data sequence number;
data vector Gi（j3) A group, which is a primary processing result of the brain wave wavelet reconstruction function, is named as a metadata group of brain wave primary processing, and is used as basic data of secondary calculation;
for a data sequence, the following calculation formula is obtained through data weighting:
E={c，d }*{Ai，Gi}；
c and d are weighting coefficients and are constants;
Ai: the sum of the complexities;
Gi: characteristic data;
applying a normalization calculation to the E data:
memory scale = (exp (e)) × 100
Obtaining real-time objective quantitative characteristic indexes of the preparation state before memory in the brain waves.
2. The brain-based memory scale and induction capture automatic knowledge memory distraction method of claim 1, which is a method for extracting real-time continuous quantitative scales reflecting the concentration memory state of the brain through the collection and calculation processing of brain wave signals, wherein the scales are used as real-time feedback control signals to drive the processes of induction, distraction and knowledge playback, and is characterized in that: the method is characterized in that a digital sequence of a section of brain wave signal in a fixed-length or variable-time window acquired in real time is utilized, various mathematical calculation algorithms are applied to the digital sequence, wavelet analysis and calculation are mainly used, a pattern recognition algorithm and a power spectrum algorithm are combined, waveform dimensions in brain waves are decomposed, waveform extreme points, the slope of a waveform, an electroencephalogram frequency spectrum, an electrooculogram and an electromyogram frequency spectrum are extracted, the current change characteristics of a real-time continuous brain memory state are extracted and used as scale expression of brain memory, the change precision and speed of scale values meet the requirements of real-time continuous control signals, and the algorithm covers wavelet analysis, power spectrum analysis, pattern recognition nonlinearity and linear numerical signal processing algorithms.
3. The brain-based memory scale and inducement capture automatic knowledge memory distraction method of claim 1, wherein the system software triggers playing of the predetermined knowledge content when capturing the pre-memory readiness state that exceeds a set threshold, and continues outputting the entertainment content of the game or the action performance of the toy below the threshold, characterized in that: the memory of knowledge is generated in a two-way resonance mechanism of knowledge playing and brain focusing on a memory state, system software calculates memory scales of the brain in real time, compares the memory scales with a preset threshold value of the memory state, switches and plays prestored knowledge contents when the memory scales exceed the threshold value, switches and outputs actions of game contents or toys when the memory scales are lower than the threshold value, and the memory scales and the knowledge playing actions occur when the memory scales are lower than the threshold value, and the knowledge playing actions circulate to form resonance; the preset memory scale threshold is invariable or self-adaptively changed, the memory state of the brain is automatically adapted to, the self-adaptive change of the threshold is determined by too long time below the threshold condition and the deviation of the memory scale below the threshold, the knowledge is presented according to a certain rule, and the knowledge content is edited and generated by a database or a character system of a computer or a server; the presentation of the knowledge comprises auditory and visual expressions, the switching of the auditory and the visual expressions or the simultaneous expression is determined by opening eyes and closing eyes, the knowledge is played according to the items, the playing of each item sets the repetition times, the repetition times are not changed, or the repetition times are changed when the scale threshold is adaptively changed or the sleep state appears; the functions of eye opening and closing and the sleep state recognized by the system software are used as options, with or without, as shown in whether the repetition times of the knowledge content playing are fixed or variable, and whether the audio-visual expression of the knowledge presentation is switchable; the playing of visual knowledge is matched with the flickering of a screen, so that the differentiated acceptance of a visual nervous system is triggered, and the visual fatigue is relieved.
CN201710290775.0A 2017-04-28 2017-04-28 Automatic knowledge memory traction method based on memory scale and induction capture of brain Active CN107184205B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710290775.0A CN107184205B (en) 2017-04-28 2017-04-28 Automatic knowledge memory traction method based on memory scale and induction capture of brain

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710290775.0A CN107184205B (en) 2017-04-28 2017-04-28 Automatic knowledge memory traction method based on memory scale and induction capture of brain

Publications (2)

Publication Number Publication Date
CN107184205A CN107184205A (en) 2017-09-22
CN107184205B true CN107184205B (en) 2021-01-22

Family

ID=59873095

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710290775.0A Active CN107184205B (en) 2017-04-28 2017-04-28 Automatic knowledge memory traction method based on memory scale and induction capture of brain

Country Status (1)

Country Link
CN (1) CN107184205B (en)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109171716A (en) * 2018-10-26 2019-01-11 湖北大学 A kind of brain electricity pathology big data disease towards intelligent medical treatment participates in alert cloud system
CN111543986B (en) * 2020-05-12 2021-03-02 清华大学 Electroencephalogram event synchronization method without hardware connection
CN112842328A (en) * 2021-01-25 2021-05-28 北京易飞华通科技开发有限公司 Integrated gait, distance, relative weight and brain wave synchronous sensing analysis system based on Internet of things
CN116312972B (en) * 2023-05-19 2023-08-11 安徽星辰智跃科技有限责任公司 Sleep memory emotion tension adjustment method, system and device based on eye stimulation

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201060561Y (en) * 2007-06-11 2008-05-14 北京易飞华通科技开发有限公司 Biofeedback mental state regulation and auxiliary learning machine
CN103876708A (en) * 2013-11-25 2014-06-25 北京大学人民医院 Wireless mobile information exchange method for pain stress and subjective feeling and pain treatment

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN201060561Y (en) * 2007-06-11 2008-05-14 北京易飞华通科技开发有限公司 Biofeedback mental state regulation and auxiliary learning machine
CN103876708A (en) * 2013-11-25 2014-06-25 北京大学人民医院 Wireless mobile information exchange method for pain stress and subjective feeling and pain treatment

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
脑电波中疼痛、焦虑、抑郁等定量指数组的提取;吴一兵;《中国医学工程》;20170425;第25卷(第4期);1-7 *

Also Published As

Publication number Publication date
CN107184205A (en) 2017-09-22

Similar Documents

Publication Publication Date Title
CN106725462B (en) Acousto-optic Sleep intervention system and method based on EEG signals
CN107184205B (en) Automatic knowledge memory traction method based on memory scale and induction capture of brain
CN109585021B (en) Mental state evaluation method based on holographic projection technology
US9211078B2 (en) Process and device for brain computer interface
CN109224242B (en) Psychological relaxation system and method based on VR interaction
JP2023521187A (en) Acoustic electrical stimulation neuromodulation method and apparatus based on measurement, analysis and control of brain waves
CN109620257B (en) Mental state intervention and regulation system based on biofeedback and working method thereof
Schaaff et al. Towards an EEG-based emotion recognizer for humanoid robots
CN113398422B (en) Rehabilitation training system and method based on motor imagery-brain-computer interface and virtual reality
JP2019528104A (en) In-ear sensing system and method for monitoring biological signals
CN106377252A (en) Biologic information feedback system based on virtual reality
CN106407733A (en) Depression risk screening system and method based on virtual reality scene electroencephalogram signal
CN108310587A (en) A kind of sleep control device and method
CN110946576A (en) Visual evoked potential emotion recognition method based on width learning
Mohamed et al. Comparison of EEG signal decomposition methods in classification of motor-imagery BCI
CN111297379A (en) Brain-computer combination system and method based on sensory transmission
CN112717253A (en) Acousto-optic combined awakening device based on brain wave monitoring
KR20080107961A (en) User adaptative pattern clinical diagnosis/medical system and method using brain waves and the sense infomation treatment techniques
CN113713333A (en) Dynamic virtual induction method and system for lower limb rehabilitation full training process
CN115640827B (en) Intelligent closed-loop feedback network method and system for processing electrical stimulation data
Smitha et al. Voice familiarity detection using EEG-based Brain-Computer Interface
CN116700495A (en) Brain-computer interaction method and equipment based on steady-state visual evoked potential and motor imagery
Sourina et al. Fractal-based brain state recognition from EEG in human computer interaction
Chen et al. The use of a brain computer interface remote control to navigate a recreational device
CN115581456A (en) Method, device and system for generating artwork

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