CN107657868A - A kind of teaching tracking accessory system based on brain wave - Google Patents
A kind of teaching tracking accessory system based on brain wave Download PDFInfo
- Publication number
- CN107657868A CN107657868A CN201710979904.7A CN201710979904A CN107657868A CN 107657868 A CN107657868 A CN 107657868A CN 201710979904 A CN201710979904 A CN 201710979904A CN 107657868 A CN107657868 A CN 107657868A
- Authority
- CN
- China
- Prior art keywords
- brain wave
- teaching
- brain
- rhythm
- waking
- 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.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
- G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
- G09B19/00—Teaching not covered by other main groups of this subclass
-
- 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
- A61B5/68—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient
- A61B5/6801—Arrangements of detecting, measuring or recording means, e.g. sensors, in relation to patient specially adapted to be attached to or worn on the body surface
- A61B5/6802—Sensor mounted on worn items
- A61B5/6803—Head-worn items, e.g. helmets, masks, headphones or goggles
-
- 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
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- 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
- A61B5/7225—Details of analog processing, e.g. isolation amplifier, gain or sensitivity adjustment, filtering, baseline or drift compensation
-
- 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
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Signal Processing (AREA)
- Animal Behavior & Ethology (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Heart & Thoracic Surgery (AREA)
- Biomedical Technology (AREA)
- Pathology (AREA)
- Biophysics (AREA)
- Psychiatry (AREA)
- Physiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Business, Economics & Management (AREA)
- Educational Administration (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Technology (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Psychology (AREA)
- Power Engineering (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention provides a kind of teaching based on brain wave to track accessory system,It passes through brain electro-detection headgear,Communication gate and system server constructing system framework,Collection and the noise suppression preprocessing of EEG signals are carried out using brain electro-detection headgear,Eeg signal regain consciousness by system server/tired state recognition handles,And the corresponding relation between the curricula information of period where keeping track of corresponding clear-headed/tired state and the eeg signal,Clear-headed/tired state of the identical student for various teaching course can also further be counted,And clear-headed/tired state of variant student and the teacher corresponding to corresponding curricula in identical curricula,The subjective judgement for avoiding relying on teacher holds state of attention and teaching affairs and subjective assessment the problem of judging teachers ' teaching quality and easily producing deviation for relying on course auditor of student,It can be analyzed for teaching efficiency and objective data auxiliary support is provided.
Description
Technical Field
The invention relates to the technical field of teaching quality analysis and evaluation, in particular to a teaching tracking auxiliary system based on brain waves.
Background
Teaching effect analysis hiccup assessment is one of the basic content of education metrology. The analysis and evaluation of the teaching effect are closely related to a plurality of factors, such as the teaching quality of each course in the previous period, the mutual cooperation of each teaching link of the course, the teaching effect of a teacher, the quality of students and the learning state. The teaching effect of the teacher and the learning state of the student are important factors directly influencing the teaching effect, so that how to or the relevant data of the two factors in the teaching activity is an important direction for analyzing and researching the teaching effect.
The attention state condition of the students in the teaching course learning can better reflect the learning state of the students, and meanwhile, the attention state condition of most of the students in the teaching course of the same teacher can reflect the teaching effect of teaching of the teacher laterally. In the actual classroom teaching, the teacher often can observe student's reaction and attention condition when lecturing to adjust the teaching mode of oneself according to actual conditions, with the teaching effect that promotes the course. How to observe the reaction and attention condition of the student, a teacher with rich teaching experience can roughly judge through the facial expression, the limb action and the like of the student. However, this method is only a subjective judgment, and is affected by individual differences between teachers and students, and in addition, the number of students in a teaching classroom is generally large, and the teaching level of teachers is different, so it is difficult to accurately grasp the attention state and teaching condition of each student in the classroom in time. In addition, at present, the teaching evaluation of teachers is mainly judged by listening to classes, and because the subjective evaluation is also based on the listeners in the classes, the learning states of the students are influenced by the listeners in the classes, and the judgment mode has certain deviation from the real teaching effect inevitably.
Therefore, how to objectively and quantitatively obtain the attention state condition of the student in the teaching course learning to be used as auxiliary information for teaching effect analysis and provide important objective data support for the teaching effect analysis is a technical problem to be solved in teaching research.
Disclosure of Invention
In view of the above-mentioned shortcomings in the prior art, an object of the present invention is to provide a teaching tracking assistance system based on brain waves, which can detect and track and record the corresponding relationship between the waking/sleepy state of the detected object and the teaching course, so as to provide objective data assistance support for the teaching effect analysis.
In order to solve the technical problems, the invention adopts the following technical means:
a teaching tracking auxiliary system based on brain waves comprises a brain wave detection headgear, a communication gateway and a system server, wherein the communication gateway is in data communication with the system server through an internet, and the brain wave detection headgear is in data communication connection with the communication gateway so as to be in data communication with the system server through the communication gateway;
the electroencephalogram detection headgear comprises an electroencephalogram acquisition module for acquiring electroencephalogram signals, an electroencephalogram preprocessing module for performing denoising preprocessing on the acquired electroencephalogram signals, and a data sending module for sending external data to the preprocessed electroencephalogram signals;
the communication gateway is used for establishing data communication connection between the electroencephalogram detection headset and the system server;
the system server records teaching course information corresponding to different time periods, is used for receiving the preprocessed brain wave signals from the brain wave detection headgear through the communication gateway, analyzing and processing the received brain wave signals, identifying the waking/sleepy state, tracking and recording the corresponding relation between the waking/sleepy state and the teaching course information of the time period where the brain wave signals are located, and is used as auxiliary information for analyzing the learning state of the teaching course of a detection object of the brain wave detection headgear.
In the teaching tracking auxiliary system based on brain waves, furthermore, a plurality of electroencephalogram detection head sleeves are provided, and each electroencephalogram detection head sleeve is correspondingly provided with an electroencephalogram ID; the number of the communication gateways is at least one, and each electroencephalogram detection headgear can establish data communication connection with the at least one communication gateway;
the preprocessed brain wave signals sent out by the brain wave detection headgear carry the brain wave ID and time information corresponding to the brain wave detection headgear;
the brain wave detection system is characterized in that student information corresponding to different brain wave IDs is recorded in the system server, the system server receives a preprocessed brain wave signal from a brain wave detection headset through a communication gateway, then determines student information and teaching course information corresponding to the brain wave signal according to the brain wave ID and time information carried by the brain wave signal, and analyzes and processes the brain wave signal and identifies the waking/sleepy state, so that the corresponding relation between the waking/sleepy state of the student corresponding to the brain wave signal and the teaching course information is tracked and recorded, and the student information is used as auxiliary information for analyzing the learning state of the teaching course for the student.
In the above-mentioned brain wave-based teaching tracking assistance system, the system server is further configured to count the waking/sleepy states of the same student for different teaching courses according to the correspondence between the waking/sleepy states of the student and the teaching course information recorded by tracking, and to use the counted waking/sleepy states as the assistance information for analyzing the learning states of the corresponding student for the different teaching courses.
In the above-mentioned brain wave-based teaching tracking assistance system, further, the system server further records teacher information corresponding to different teaching courses, and is configured to count the waking/drowsy states of different students in the same teaching course and a teacher corresponding to a corresponding teaching course according to the correspondence between the waking/drowsy states of the students recorded by tracking and the teaching course information, and is configured to serve as assistance information for analyzing the influence status of the teaching courses of the corresponding teachers on the learning states of the students.
In the teaching tracking auxiliary system based on brain waves, as a preferred scheme, the brain wave preprocessing module performs denoising preprocessing on the collected brain wave signals by adopting a dual-tree complex wavelet transform method.
In the above-mentioned brain wave-based teaching tracking assistance system, as a preferred scheme, the brain wave preprocessing module employs a dual-tree complex wavelet transform method to perform a denoising threshold function f (W) for denoising and preprocessing the collected brain wave signal h,l ) Comprises the following steps:
wherein, W h,l Represents the wavelet coefficient on the l scale of the h-th layer dual-tree complex wavelet transform,representing wavelet coefficients W h,l Corresponding wavelet coefficient estimated value obtained by denoising threshold function, wherein lambda represents a preselected threshold; sgn (·) is the return function sign, exp (·) is the exponent function sign.
In the above teaching tracking assistance system based on brain waves, as a preferred scheme, the method for analyzing and processing the received brain wave signals and identifying the waking/drowsy states in the system server specifically includes:
acquiring a sample entropy estimation value of the brain wave signal, extracting theta rhythm wave, alpha rhythm wave and beta rhythm wave in the brain wave signal respectively, calculating average power spectral density G (theta) of the theta rhythm wave, average power spectral density G (alpha) of the alpha rhythm wave and average power spectral density G (beta) of the beta rhythm wave respectively, and further calculating to obtain average power ratio F of the theta rhythm wave and the alpha rhythm wave relative to the beta rhythm wave (θ+α)/β = (G (θ) + G (α))/G (β) and the average power ratio F of θ rhythm wave to β rhythm wave θ/β = G (θ)/G (β), thereby calculating the mean power ratio F as the sample entropy estimation value of the brain wave signal (θ+α)/β And the average power ratio F θ/β Forming an identification feature vector of the brain wave signal;
when the human brain is in a waking state or a sleepy state, the sample entropy estimation value of the brain wave signal and the average power ratio F (θ+α)/β And the average power ratio F θ/β There is a difference, so the sample entropy estimated values of the respective brain wave signals when the human brain is in the awake state and the sleepy state and the average power ratio F are counted in advance (θ+α)/β And the average power ratio F θ/β And training the classifier of the support vector machine after marking and storing the classifier in a system server, wherein the system server utilizes the classifier pair of the support vector machine obtained by trainingThe identification feature vector of the received brain wave signal is used for carrying out classified identification on the waking state and the sleepy state, so that the waking/sleepy state identification result is obtained.
In the above-described teaching follow-up support system based on brain waves, specifically, the average power spectral density G (θ) of θ rhythm waves, the average power spectral density G (α) of α rhythm waves, and the average power spectral density G (β) of β rhythm waves are calculated in the following manner:
wherein f is θh And f θl Respectively representing the upper limit value and the lower limit value of the frequency band of theta rhythm waves in the brain wave signal; f. of αh And f αl Respectively representing the upper limit value and the lower limit value of the frequency band of alpha rhythm waves in the brain wave signal; f. of βh And f βl Respectively representing the upper limit value and the lower limit value of the frequency band of the beta rhythm wave in the brain wave signal; p is a radical of X (f) Which represents the average power spectrum of the brain wave signal at frequency f.
In the above teaching tracking assistance system based on brain waves, specifically, the manner of obtaining the sample entropy estimation value of the brain wave signal is as follows:
1) Presetting a value of a distance tolerance threshold value r and a value of a vector dimension m, wherein r is greater than 0, and the vector dimension m is a positive integer greater than 1;
2) Processing the brain wave signal U to form a m-dimensional brain wave data vector X m (i)：
X m (i)＝[u(i),u(i+1),...,u(i+m-1)]，i∈{1,2,…,N-m+1}；
Wherein U (i) represents the value of the ith sampling point in the brain wave signal U, and N is the total number of the sampling points of the brain wave signal U;
3) For each different value of i, respectively calculating a brain wave data vector X corresponding to the ith sampling point in the brain wave signal U m (i) Brain wave data vector X corresponding to the rest of sampling points m (j) Distance d [ X ] between m (i),X m (j)]：
d[X m (i),X m (j)]＝max{|u(i+k)-u(j+k)||k＝1,2,…,m-1}；
Wherein j belongs to {1,2, …, N-m +1}, and j is not equal to i;
4) For each value of i, respectively counting the corresponding distance value d [ X ] m (i),X m (j)]The number of distances less than the distance tolerance threshold r is recorded asAnd recording the ratio of the number to the total distance number N-m asNamely:
5) The ratio corresponding to all the values of i when the dimension of the vector is m is solvedAverage ratio ofNamely:
6) Let the vector dimension be m +1, repeat steps 2) -5) to obtain the corresponding average ratio B when the vector dimension is m +1 m+1 (r)；
7) Calculating a sample entropy estimation value SampEn (m, r, N) of the brain wave signal by the following formula:
SampEn(m,r,N)＝-ln[B m+1 (r)/B m (r)]。
compared with the prior art, the invention has the following beneficial effects:
the brain wave-based teaching tracking auxiliary system provided by the invention constructs a system framework through the brain wave detection headgear, the communication gateway and the system server, utilizes the brain wave detection headgear to collect brain electric signals and carry out denoising pretreatment, utilizes the system server to carry out waking/sleepy state identification treatment on the brain wave signals, and tracks and records the corresponding relation between the corresponding waking/sleepy state and the teaching course information of the time period in which the brain wave signals are positioned.
Drawings
Fig. 1 is a schematic structural diagram of the brain wave-based teaching tracking assistance system of the present invention.
Fig. 2 is a schematic diagram of a dual-tree complex wavelet transform decomposition and reconstruction process.
Fig. 3 is a schematic diagram of a dual-tree complex wavelet transform using a parity filter bank.
Fig. 4 is a schematic diagram of a dual-tree complex wavelet transform using a Q-shift filter.
FIG. 5 is a graph of a function of a hard threshold function and a soft threshold function.
FIG. 6 is a functional graph of an improved denoising threshold function compared to a hard threshold function and a soft threshold function.
Fig. 7 is a comparison diagram of the effect of denoising brain wave signals using wavelet transform and dual-tree complex wavelet transform.
Fig. 8 is a comparison graph of the effect of denoising brain wave signals by a dual-tree complex wavelet transform method using a hard threshold function, a soft threshold function, and an improved denoising threshold function.
Fig. 9 is a graph showing the change of the average power of three rhythm waves, i.e., theta wave, alpha wave and beta wave, with time.
FIG. 10 is a graph showing the change in the average power of α/β, θ/β, (α + θ)/(α + β) with time.
FIG. 11 is a sample entropy scatter plot of electroencephalograms in awake and sleepy states.
Detailed Description
Research shows that the movement of cerebral cortex and each sensory area are closely related to each rhythm wave of electroencephalogram signals, so the electroencephalogram signals are always regarded as the 'gold standard' which can reflect the fatigue state of human bodies most, and the 'gold standard' reflects the forming process of the electroencephalogram signals. The electroencephalogram signal mainly consists of several rhythm waves of Delta (Delta), theta (Theta), alpha (Alpha), beta (Beta) and Gamma (Gamma), and only four basic rhythms of Alpha wave, beta wave, theta wave and Delta wave are usually involved in practical related application research. The delta wave appears on the frontal lobe part of the brain obviously, the frequency is between 0.5 and 4Hz, and the amplitude is about 20 to 200 MuV; very small amounts of delta waves occur when an adult is extremely fatigued and lethargy, but not when the adult is in a normal awake state. The theta wave appears in the temporal lobe part of the brain remarkably, the frequency is between 4 and 8Hz, and the amplitude range is between 100 and 150 muV; large fluctuations in the theta wave occur when adults are in a tired, drowsy state, on the other hand showing that the cerebral cortex is somewhat inhibited. The alpha wave appears most remarkably in the occipital lobe area of the brain, is concentrated between 8Hz and 13Hz, and has the amplitude of about 50 to 100 MuV; the main manifestation of electrical activity in the cerebral cortex in a conscious and quiescent state is a large increase in alpha rhythm waves. The beta rhythm waves are usually found in the frontal lobe, the parietal lobe and the central region, the frequency is 14-30 Hz, and the amplitude is about 5-20 muV; when a person becomes emotional stress or highly concentrated, the beta wave increases significantly. The attention state condition of the student in teaching course learning can be objectively analyzed and judged through brain wave signal detection.
Based on the research thought, the invention provides a teaching tracking auxiliary system based on brain waves, as shown in fig. 1, the system comprises a brain electrical detection headgear 1, a communication gateway 2 and a system server 3, the communication gateway 2 is in data communication with the system server 3 through an internet, and the brain electrical detection headgear 1 is in data communication connection with the communication gateway 2, so as to be in data communication with the system server 3 through the communication gateway 2. The electroencephalogram detection headgear comprises an electroencephalogram acquisition module, an electroencephalogram preprocessing module and a data sending module, wherein the electroencephalogram acquisition module is used for acquiring electroencephalogram signals, the electroencephalogram preprocessing module is used for denoising the acquired electroencephalogram signals, and the data sending module is used for sending external data to the preprocessed electroencephalogram signals. The communication gateway 2 is used for establishing data communication connection between the electroencephalogram detection headset 1 and the system server 3, and the communication gateway can be wired or wireless; if the wireless communication gateway is adopted, the electroencephalogram detection headset can be in wireless data communication connection with the communication gateway. Teaching course information corresponding to different time intervals is recorded in the system server 3, and is used for receiving the preprocessed brain wave signals from the brain wave detection headgear 1 through the communication gateway 2, analyzing and processing the received brain wave signals, identifying the waking/sleepy state, tracking and recording the corresponding relation between the waking/sleepy state and the teaching course information of the time interval where the brain wave signals are located, and being used as auxiliary information for analyzing the learning state of the teaching course of a detection object of the brain wave detection headgear.
By utilizing the brain wave-based teaching tracking auxiliary system, brain wave signals of a student can be collected by utilizing the brain wave detection head and are transmitted to the system server through the communication gateway, and the system server identifies the waking/sleepy state of the brain wave signals through analysis processing, so that the corresponding relation between the corresponding waking/sleepy state and the teaching course information of the time period in which the brain wave signals are located is tracked and recorded, and the brain wave signal tracking auxiliary system is used as auxiliary information for analyzing the learning state of the student aiming at the teaching course. Therefore, the problem that deviation is easy to occur depending on subjective judgment of teachers on the attention states and teaching conditions of students is solved, and objective data auxiliary support can be provided for teaching effect analysis by using data information obtained by the teaching tracking auxiliary system.
In order to detect the brain waves of teaching tracking of a plurality of trainees, in the brain wave-based teaching tracking auxiliary system, as shown in fig. 1, a plurality of brain electricity detection head sleeves 1 are provided, and each brain electricity detection head sleeve is correspondingly provided with a brain electricity ID used as a data identity identification code of the corresponding brain electricity detection head sleeve; one or more communication gateways 2 may be provided, and the number of the communication gateways 2 is determined according to the communication range of the system arrangement, as long as each electroencephalogram detection headset 1 is ensured to be capable of establishing data communication connection with at least one communication gateway 2. Under the design scheme, the electroencephalogram detection headset 1 needs to carry electroencephalogram ID and time information corresponding to the electroencephalogram detection headset in the preprocessed electroencephalogram signals sent outside; correspondingly, student information corresponding to different electroencephalogram IDs also needs to be recorded in the system server 3, after the system server 3 receives the preprocessed electroencephalogram signals from the electroencephalogram detection headgear 1 through the communication gateway 2, student information and teaching course information corresponding to the electroencephalogram signals are respectively determined according to the electroencephalogram IDs carried by the electroencephalogram signals and time information, the electroencephalogram signals are analyzed and processed, and the waking/sleepy state is identified, so that the corresponding relation between the waking/sleepy state of the students corresponding to the electroencephalogram signals and the teaching course information is tracked and recorded, and the corresponding relation is used as auxiliary information for analyzing the learning state of the teaching courses for the students. Therefore, a plurality of electroencephalogram detection heads in the system can be used for synchronously identifying and tracking the waking state/sleepy state of multiple persons, even the whole students in a teaching course.
On the basis of the system framework, the system server can be further designed to calculate the waking/sleepy states of the same student for different teaching courses according to the corresponding relationship between the waking/sleepy states of the student and the teaching course information recorded by tracking, and the waking/sleepy states of the same student for different teaching courses can be used as auxiliary information for analyzing the learning states of the corresponding student for different teaching courses. The statistical data can be used for tracking and recording the learning states of different teaching courses of a single student, so that whether the corresponding student has the problems of improper learning method or weak learning ability and the like or not can be conveniently analyzed and found, and therefore targeted guidance is performed, and the learning effect of the student is improved.
Similarly, on the basis of the system framework, the system server may further be designed to record teacher information corresponding to different teaching courses, so as to count the waking/drowsiness states of different students in the same teaching course and the teachers corresponding to the corresponding teaching courses according to the correspondence between the waking/drowsiness states of the students recorded by tracking and the teaching course information, and to serve as auxiliary information for analyzing the influence of the teaching courses of the corresponding teachers on the learning states of the students. The statistical data can be used for tracking and recording the influence condition of the teaching course of the delayed teacher on the learning state of the student, so that whether the corresponding teacher has the problems of improper teaching method or the problem that the teaching skill needs to be improved or the like can be conveniently analyzed and found, and therefore targeted training or teaching adjustment is performed to help improve the teaching effect. Therefore, the problem that deviation is easily generated when the teaching quality of the teacher is judged by relying on subjective evaluation of the audience of the lesson is avoided, and objective quantitative evaluation on the teaching effect of the teacher is facilitated.
In the teaching tracking auxiliary system based on the brain wave, the denoising pretreatment of the brain wave signal and the identification of the waking/sleepy state are the technical key points of system design.
A. And (4) denoising and preprocessing the electroencephalogram signals.
In the teaching tracking auxiliary system, the electroencephalogram detection headset is provided with a electroencephalogram preprocessing module for denoising and preprocessing the acquired electroencephalogram signals, and the electroencephalogram signals have extremely weak characteristics, so that the acquired electroencephalogram signals are easily submerged by noise due to internal factors of a human body and external environment changes in the signal acquisition process, and therefore, for analysis and processing of the electroencephalogram signals, the electroencephalogram signals need to be denoised and purified, and further subsequent analysis can be performed.
The main methods for denoising and preprocessing the electroencephalogram signals comprise artifact subtraction, a principal component analysis method, an independent component analysis method, a wavelet analysis method and the like. The artifact subtraction is to respectively measure the original electroencephalogram signal and the artifact signal through actual measurement, and simultaneously, if the real electroencephalogram signal in the originally acquired electroencephalogram signal is not related to the artifact, the measured artifact is subtracted from the originally measured electroencephalogram signal, so that the obtained signal is used as the real electroencephalogram signal. The principal component analysis method mainly utilizes the orthogonality principle to convert a group of variables with correlation into mutually independent variables, then selects some more important components from the separated independent variables for analysis, discards other less important components, and finally adopts the least square method to the selected principal components to realize the estimation of the relevant parameters of the model; according to the distribution condition of the electroencephalogram acquisition electrodes, PCA decomposition is carried out on the acquired electroencephalogram signals to obtain a plurality of components with mutual independence, artifact components are selected and removed, and finally the remaining signals are used for realizing reconstruction of the electroencephalogram signals, so that noise is removed to the maximum extent. The independent component analysis method mainly adopts an optimal objective function to process signals so that components obtained by separation can approach source signals to the maximum extent. The wavelet analysis method is mainly used for preprocessing and denoising electroencephalogram signals and extracting relevant features in the application field of the electroencephalogram signals, so that the wavelet analysis method is generally applied to electroencephalogram signal processing.
Theory of wavelet transform.
1) Continuous wavelet transform and discrete wavelet transform:
let Ψ (t) ∈ L 2 (R) if Ψ (ω) satisfiesThe formula is as follows:
L 2 (R) represents L2 norm space, psi (t) is called as mother wavelet or base wavelet function, and the wavelet function psi (t) can be obtained by performing expansion and shift operation on the mother wavelet psi (t) a,b (t)：
a and b are both constants, and a > 0. Then by a continuous wavelet transform defining the available signal x (t):
a is a scaling factor; b is a shifting factor.
When a, b are both continuous, due to WT at different points x (a, b) satisfies the reconstruction condition, so that the wavelet basis function Ψ a,b (t) have a correlation therebetween such that WT is caused to pass through wavelet transform x The information obtained after (a, b) has great redundancy, and when a, b are respectively dispersed into m, n, if the wavelet basis function psi obtained after dispersion a,b (t) satisfying the condition of perfect orthogonality, there is no redundancy between coefficients obtained by wavelet transform, which not only greatly compresses data but also reduces the amount of computation, and the discretization of the shift factor and the scale factor is usually realized by dyadic wavelet sampling, in which a is discretized into 2 -j B is discrete as 2 -j k, then the discrete wavelet transform obtained by discretization by the continuous wavelet transform:
j is a discrete scaling factor; k is the discrete translation factor.
Then the wavelet function Ψ j,k (t) can be expressed as:
Ψ j,k (t)＝2 -j/2 Ψ(2 -j t-k),j,k∈Z；
after introducing the scaling function, the signal can be generally represented by a wavelet function and a scaling function together, wherein the scaling function phi j，k (t) is defined as:
φ j，k (t)＝2 -j/2 φ(2 -j t-k),j,k∈Z；
using a scale filter h 0 (n) to denote φ (t), then:
therefore, the finite energy signal f (t) can be decomposed to obtain:
assuming that the wavelet satisfies the perfect orthogonality condition, its profile coefficients and detail coefficients can be expressed as:
c j (k)＝＜f(x),φ j,k (t)＞；
d j (k)＝＜f(x),Ψ j,k (t)＞；
discrete wavelet transform has the capability of multi-resolution analysis while the profile coefficients c j (k) And a detail coefficient d j (k) The following recursion relationship exists:
2) And (3) complex wavelet transformation:
the transformation base of Fourier transform is composed of an oscillation function e jwt (= cos (wt) + jsin (wt))The phase difference between the real part and the imaginary part is just 90 degrees, then the phase difference just can form a Hilbert transform pair, a complex wavelet transform method is provided, the advantages of Fourier transform and wavelet transform are combined at the same time, only the wavelet function and the scale function of the complex wavelet transform are both formed in a complex form, and the complex wavelet function can be expressed as:
Ψ(t)＝Ψ r (t)+iΨ i (t)；
wavelet basis function Ψ of low-pass filter r (t) wavelet basis function Ψ for the high-pass filter i (t) is just able to constitute a hilbert pair, and at the same time satisfies the equation:
then, when the complex wavelet transform is used for signal processing, the translational offset brought by the signal transformation process can be greatly reduced.
3) Dual-tree complex wavelet transform:
the process of decomposing and reconstructing a one-dimensional signal using dual-tree complex wavelet transform is shown in fig. 2. Wherein the number of decomposition layers is 3,h 0 And h 1 Low-pass and high-pass filter banks, g, representing the real part tree used in the decomposition, respectively 0 And g 1 Respectively represent the low-pass and high-pass filters used in the imaginary tree during decomposition, and h 'during signal reconstruction in the same manner' 0 H of' 1 Filter bank, g ', adopted for real part tree' 0 G of' 1 A filter bank is employed for the imaginary tree. There are two main types of filter banks designed at present that can achieve complete signal reconstruction: 1) Parity filterbank, 2), Q-shift filterbank. In which a bank of odd-even filters is used for a dual-tree complex wavelet transformIs shown in fig. 3. The odd-even filter is a filter which alternately uses odd-long and even-long filters when performing signal decomposition and signal reconstruction, that is, a filter which alternately uses odd-long and even-long filters between each layer of a tree a and a tree b, so that relatively good symmetry is shown between the tree a and the tree b, wherein the same odd-long filter is used in the first layer of the tree a and the tree b, while in the second layer structure, the same odd-long filter is used in the tree b as the previous layer, and the even-long filter is used in the tree a, and for the decomposition of the subsequent layers, the odd-even filter is repeatedly and alternately used, so that the average error between the signal obtained by reconstruction and the original signal can be minimized, wherein the design requirement of the odd-even filter group satisfies the following relation:
wherein H 00 The length of the (z) filter is even long, while G 00 The length of the (z) filter is odd long.
The Q-shift filter group is adopted to process electroencephalogram signals, wherein the structure of using the Q-shift filter to perform dual-tree complex wavelet transformation is shown in figure 4. As can be seen from fig. 4, when the Q-shift filter bank is used to perform dual-tree complex wavelet transform, even-long filters are used in the second and subsequent layers of the two trees, but not with strict linear phase characteristics, although the Q-shift filter has 1/4 sampling interval delay, 1/2 sampling period delay needs to be maintained between the two trees, so that by using an even-long filter with 1/4 (+ Q) period delay in tree a and an even-long filter with 3/4 (+ 3Q) period delay in tree b, it can be ensured that the sampling point of tree b just falls in the middle of the sampling point of tree a, and a symmetric sampling manner is achieved, as can be seen from the linear time invariant theory and Z transform, the designed Q-shift filter bank needs to satisfy the following relation:
H L2 (z) is a linear phase filter with a 1/2 sample interval delay and a length of 4n, and H L (z) represents a Q-shift filter of length 2n but with 1/4 sample period delay, by pair H L2 (z) subsampling to obtain H L (z)。
The dual-tree complex wavelet transform is composed of two parallel paths of wavelet transforms, so that the decomposition and reconstruction of signals are realized, wherein the scale coefficient of a real part treeAnd wavelet coefficientsThe following formula:
j is a scale factor; j is the largest dimension and J =1,2, …, J.
Obtaining wavelet coefficients of an imaginary treeAnd scale factorThe following formula:
the relation between wavelet coefficient and scale coefficient can be obtained by combining real part tree and imaginary part tree, and the wavelet coefficient obtained after dual-tree complex wavelet transform decomposition can be knownAnd scale factorRespectively as follows:
wavelet coefficient d obtained by reconstructing real part tree and imaginary part tree j (t) and scale factor c J (t) is represented by the following formula:
the reconstructed signal x (t) is then:
x(t)＝d j (t)+c J (t)。
the dual-tree complex wavelet transform has the following properties: (1) translational invariance; when the input signal generates weak oscillation or generates small delay, each sub-band signal obtained after decomposition and reconstruction does not have large fluctuation or distortion, and has approximate translation and no deformation. (2) Anti-frequency aliasing effects: compared with the traditional discrete wavelet transform, the dual-tree complex wavelet transform can better inhibit the occurrence of frequency aliasing.
The above wavelet transformation theory knowledge is mature knowledge, and the existing relevant documents of wavelet transformation theory knowledge can be specifically referred. Based on the analysis, as a preferred scheme, in the electroencephalogram detection headset of the teaching tracking assistance system, the electroencephalogram preprocessing module preferably performs denoising preprocessing on the acquired electroencephalogram signals by adopting a dual-tree complex wavelet transform method.
In the dual-tree complex wavelet transform denoising preprocessing process, a denoising threshold function needs to be selected. Two threshold functions that are commonly used are: a hard threshold function and a soft threshold function. The expressions for the hard threshold function and the soft threshold function are as follows, respectively.
Hard threshold function:
soft threshold function:
wherein, W h,l Represents wavelet coefficients at the l-th scale of the h-th wavelet decomposition,representing wavelet coefficients W h,l Corresponding wavelet coefficient estimates from a threshold function, λ representing a preselected threshold, sgn (·) being the sign of the return function. Two threshold functions are shown in fig. 5.
In order to achieve a better filtering effect, the invention provides an improved denoising threshold function in the process of denoising and preprocessing the collected brain wave signals by adopting a dual-tree complex wavelet transform method, and the expression of the improved denoising threshold function is as follows:
wherein, f (W) h,l ) Representing the de-noising threshold function, W h,l Represents the wavelet coefficient on the l scale of the h-th layer dual-tree complex wavelet transform,representing wavelet coefficients W h,l Corresponding wavelet coefficient estimated value obtained by denoising threshold function, wherein lambda represents a preselected threshold; sgn (·) is the return function sign, exp (·) is the exponent function sign.
The following improved denoising threshold function f (W) proposed by the present invention j,k ) Analysis and investigation were performed.
1) When W is h,l When the pressure is higher than 0, the pressure is higher,
2) When W is h,l When the ratio is less than 0, the reaction mixture is,
then there is
3) And simultaneously:
from the above derivation, in the two-dimensional plane coordinate system of the x-axis and the y-axis, the improved denoising threshold function is asymptotic with y = x, i.e. the denoising threshold function is asymptotic with y = xAs an asymptote, with W h,l The increase in the number of the first and second,gradual approximation which overcomes the soft threshold function W h,l Andhas the disadvantage of a constant deviation between them, and as it approaches the threshold value gradually, it does not directly zero its wavelet coefficients but approaches zero in a gradual manner, so that the function becomes continuous, providing for its further exploitation, at | W h,l |&And within the wavelet coefficient threshold range of lambda,all zeroes, which are consistent with soft and hard threshold functions, at | W h,l |&And the wavelet coefficient of lambda adopts gradual compression, and the compression is reduced along with the increase of the wavelet coefficient, so that the noise component in the useful signal can be well processed. The improved denoising threshold function is compared to the soft and hard threshold functions as shown in FIG. 6.
Through the above discussion, in order to embody the dual-tree complex wavelet transform denoising and improve the effect of the denoising threshold function, the pair of effects of performing denoising processing on brain wave signals by using the wavelet transform and the dual-tree complex wavelet transform method is shown in fig. 7; meanwhile, under the condition that the dual-tree complex wavelet transform method is adopted to perform denoising processing on brain wave signals, the effect pairs of the hard threshold function, the soft threshold function and the improved denoising threshold function provided by the invention are respectively adopted, as shown in fig. 8. As can be seen from the comparison between fig. 7 and fig. 8, the denoising oscillation is reduced by using the dual-tree complex wavelet transform for denoising, and for quantitatively analyzing the denoising effect of the dual-tree complex wavelet transform and the discrete wavelet transform, the root mean square error and the signal-to-noise ratio are quantitatively analyzed, wherein the calculation formulas of the root mean square error RMSE and the signal-to-noise ratio SNR are as follows:
wherein x is i And respectively representRespectively representing the signal values before and after denoising, wherein N is the total number of sampling points. The comparison of the different threshold function denoising effect evaluation indexes is shown in table 1:
TABLE 1
As can be seen from table 1, when the brain wave signals are denoised by using the improved denoising threshold function, the SNR is improved, and the RMSE of the signals is also reduced, which indicates that the improved denoising threshold function has a better denoising effect and can well retain some detail information in the signals compared with the hard threshold function and the soft threshold function.
B. The brain wave signal is identified as being awake/sleepy.
The distinction of the waking state and the sleepy state is realized, the EEG signals in different states are selected, and the characteristics which can reflect the change of the mental state most are selected to form a group of characteristic quantities so as to be convenient for subsequent identification and classification.
Through research, the sample entropy estimation value of the brain wave signal and the average power ratio F of the brain wave signal are found when the human brain is in an awake state or a sleepy state (θ+α)/β And the average power ratio F θ/β There are differences. Therefore, the sample entropy estimation value of the brain wave signal and the average power ratio F may be employed (θ+α)/β And the average power ratio F θ/β The recognition feature vector of the electroencephalogram signal is formed, and the awake/sleepy state of the electroencephalogram signal is recognized.
1. Average power spectrum ratio.
And performing power spectrum analysis on different rhythm waves in the extracted electroencephalogram signal by adopting a power spectrum estimation method. For a certain sampleThis signal x T (t) for sample signals, fast fourier transforming:
wherein the power spectral density is defined as:
where f denotes frequency, T denotes signal period, and T denotes time. After fast Fourier transform, x is obtained n (e jf ) For its frequency spectrum X N (f) The square of the mode is solved, and finally, a calculation formula of the average power spectrum is obtained:
wherein, N represents the total number of sampling points of the signal, N represents the nth sampling point, and N belongs to {1,2, …, N }. From the above, the average power spectral density of a signal in a certain frequency band is:
wherein f is h Is the upper limit of the frequency band h, f l Lower limit of the frequency band h, p X (f) Is the average power spectral density of the signal.
Then the frequency band h 1 Sum frequency band h 2 The average power ratio of these two different frequency bands is defined as:
the average power change conditions of theta rhythm wave, alpha rhythm wave and beta rhythm can be obtained respectively.
The above theoretical knowledge about the average power spectrum, i.e. the average power spectral density, is mature knowledge, and can be specifically referred to the existing relevant theoretical knowledge literature.
Specifically, for the teaching tracking assistance system of the present invention, the theta rhythm wave, the alpha rhythm wave, and the beta rhythm wave in the brain wave signal may be extracted, the average power spectral density G (theta) of the theta rhythm wave, the average power spectral density G (alpha) of the alpha rhythm wave, and the average power spectral density G (beta) of the beta rhythm wave may be calculated, and the average power ratio F of the theta rhythm wave and the alpha rhythm wave to the beta rhythm wave may be calculated (θ+α)/β = (G (θ) + G (α))/G (β) and the average power ratio F of θ rhythm wave to β rhythm wave θ/β = G (θ)/G (β). Wherein, the mean power spectral density G (theta) of theta rhythm waves, the mean power spectral density G (alpha) of alpha rhythm waves and the mean power spectral density G (beta) of beta rhythm waves are calculated in the following way:
wherein f is θh And f θl Respectively representing the upper limit value and the lower limit value of the frequency band of theta rhythm waves in the brain wave signal; f. of αh And f αl Respectively representing the upper limit value and the lower limit value of the frequency band of alpha rhythm waves in the brain wave signal; f. of βh And f βl Respectively representing the upper limit value and the lower limit value of the frequency band of the beta rhythm wave in the brain wave signal; p is a radical of X (f) Which represents the average power spectrum of the brain wave signal at frequency f.
Fig. 9 shows a time-dependent change chart of average power of three rhythm waves, i.e., a θ wave, an α wave, and a β wave, obtained by correlating electroencephalogram signal data obtained from a change from an awake state to a drowsy state. On the basis of solving the power of each rhythm wave, the average power of the related rhythm waves is further constructed, such as: α/β, θ/β, (α + θ)/β, (α + θ)/(α + β), the changes of which are shown in FIG. 10. It can be seen that with the continuous deepening of the sleepiness degree, the average of every rhythm wave of the brain wave rhythm waveThe power change was not significant, but by comparison it was found that F was in the awake and sleepy states (α+θ)/β ,F θ/β The ratio of the two parameters is obviously changed, so the invention selects the two parameters as the identification characteristics for only carrying out the waking state/sleepy state identification on the brain wave signals.
2. Sample entropy.
Sample entropy is also an important feature contained in the brain electrical signal. The physical meaning of the sample entropy is similar to the approximate entropy, and the larger the value of the sample entropy, the more complex the signal sequence is represented. For the electroencephalogram signal, the total number N of sampling points is finite, and therefore only the sample entropy estimation value can be obtained. The acquisition mode of the sample entropy estimation value of the brain wave signal is as follows:
1) Presetting a value of a distance tolerance threshold value r and a value of a vector dimension m, wherein r is greater than 0, and the vector dimension m is a positive integer greater than 1;
2) Processing the brain wave signal U to form a m-dimensional brain wave data vector X m (i)：
X m (i)＝[u(i),u(i+1),...,u(i+m-1)]，i∈{1,2,…,N-m+1}；
Wherein U (i) represents the value of the ith sampling point in the brain wave signal U, and N is the total number of the sampling points of the brain wave signal U;
3) For each different value of i, respectively calculating a brain wave data vector X corresponding to the ith sampling point in the brain wave signal U m (i) Brain wave data vector X corresponding to other sampling points m (j) Distance d [ X ] between m (i),X m (j)]：
d[X m (i),X m (j)]＝max{|u(i+k)-u(j+k)||k＝1,2,…,m-1}；
Wherein j belongs to {1,2, …, N-m +1}, and j is not equal to i;
4) For each value of i, respectively counting the corresponding distance value d [ X ] m (i),X m (j)]The number of smaller distance tolerance thresholds r is recorded asAnd recording the ratio of the number to the total distance number N-m asNamely:
5) The ratio corresponding to the values of all i when the dimension of the vector is m is solvedAverage ratio of B m (r), namely:
6) Let the vector dimension be m +1, repeat steps 2) -5) to obtain the corresponding average ratio B when the vector dimension is m +1 m+1 (r)；
7) Calculating a sample entropy estimation value SampEn (m, r, N) of the brain wave signal by the following formula:
SampEn(m,r,N)＝-ln[B m+1 (r)/B m (r)]。
for the electroencephalogram data in the waking state and the sleepy state, 60 groups of sample data are selected for calculation, and the obtained scatter distribution diagram of the sample entropy value of the electroencephalogram signal is shown in fig. 11.
Through the analysis, the sample entropy estimation value of the brain wave signal and the average power ratio F of the brain wave signal in the waking state or the sleepy state of the human brain can be seen (θ+α)/β And the average power ratio F θ/β There are significant differences. Therefore, as an implementable method for performing analysis processing and waking/drowsy state recognition on received brain wave signals in the system server, the sample entropy estimation values of the respective brain wave signals and the average power ratio F when the human brain is in the waking state and the drowsy state can be counted in advance (θ+α)/β And the average power ratio F θ/β And training a support vector machine classifier after marking and storing the classifier in the system clothesIn the server, the system server uses the trained support vector machine classifier to perform classification recognition of the awake state and the drowsy state on the recognition feature vector of the received brain wave signal, so as to obtain the recognition result of the awake/drowsy state.
In summary, the teaching tracking assistance system based on brain waves provided by the invention constructs a system framework through the brain wave detection headgear, the communication gateway and the system server, utilizes the brain wave detection headgear to collect and de-noise the brain electric signals, utilizes the system server to identify and process the brain wave signals, and tracks and records the corresponding relationship between the corresponding waking/sleepy states and the teaching course information of the time period of the brain wave signals, and can further count the waking/sleepy states of the same student aiming at different teaching courses, the waking/sleepy states of different students in the same teaching course and the teacher corresponding to the corresponding teaching course, so as to be used as auxiliary information for analyzing the learning state of the teaching courses by a detection object of the brain wave detection headgear and used as auxiliary information for analyzing the influence condition of the teaching courses of the corresponding teacher on the learning state of the student, and avoid the problem that the teaching course is easy to generate deviation by subjective judgment of the attention state and teaching state of the student depending on the teaching course of the teacher and provide objective auxiliary teaching course data for teaching effect analysis.
Finally, the above embodiments are only intended to illustrate the technical solution of the present invention and not to limit the same, and although the present invention has been described in detail with reference to the embodiments, it will be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention, which should be covered by the claims of the present invention.
Claims (9)
1. A teaching tracking auxiliary system based on brain waves is characterized by comprising a brain wave detection headgear, a communication gateway and a system server, wherein the communication gateway is in data communication with the system server through an internet, and the brain wave detection headgear is in data communication connection with the communication gateway so as to be in data communication with the system server through the communication gateway;
the electroencephalogram detection headgear comprises an electroencephalogram acquisition module for acquiring electroencephalogram signals, an electroencephalogram preprocessing module for performing denoising preprocessing on the acquired electroencephalogram signals, and a data sending module for sending external data to the preprocessed electroencephalogram signals;
the communication gateway is used for establishing data communication connection between the electroencephalogram detection headset and the system server;
the system server records teaching course information corresponding to different time periods, is used for receiving the preprocessed brain wave signals from the brain wave detection headgear through the communication gateway, analyzing and processing the received brain wave signals, identifying the waking/sleepy state, tracking and recording the corresponding relation between the waking/sleepy state and the teaching course information of the time period where the brain wave signals are located, and is used as auxiliary information for analyzing the learning state of the teaching course of a detection object of the brain wave detection headgear.
2. The brain wave-based teaching tracking assistance system according to claim 1, wherein the electroencephalogram detection head is provided in plurality, and each electroencephalogram detection head is provided with an electroencephalogram ID; the number of the communication gateways is at least one, and each electroencephalogram detection headgear can establish data communication connection with the at least one communication gateway;
the brain wave signals which are sent out by the brain wave detection headgear after the preprocessing carry the brain wave ID and the time information corresponding to the brain wave detection headgear;
the brain wave teaching system is characterized in that student information corresponding to different brain wave IDs is recorded in the system server, the system server receives a preprocessed brain wave signal from a brain wave detection headgear through a communication gateway, determines student information and teaching course information corresponding to the brain wave signal according to the brain wave ID and time information carried by the brain wave signal, analyzes and processes the brain wave signal and identifies the waking/sleepy state, so that the corresponding relation between the waking/sleepy state of the student corresponding to the brain wave signal and the teaching course information is tracked and recorded, and the student information is used as auxiliary information for analyzing the learning state of the teaching course.
3. The brain wave-based teaching tracking assistance system according to claim 2, wherein the system server is further configured to count the waking/sleepy states of the same student for different teaching courses according to the correspondence between the tracked waking/sleepy states of the student and the teaching course information, and use the counted waking/sleepy states as the assistance information for analyzing the learning states of the corresponding students for different teaching courses.
4. The brain wave-based teaching tracking assistance system according to claim 2, wherein the system server further records teacher information corresponding to different teaching courses, and is configured to count the waking/drowsy states of different students in the same teaching course and a teacher corresponding to the corresponding teaching course according to the correspondence between the waking/drowsy states of the students tracked and the teaching course information, and is used as the assistance information for analyzing the influence of the teaching courses of the corresponding teachers on the learning states of the students.
5. The brain wave-based teaching tracking assistance system according to claim 1, wherein the brain wave preprocessing module performs denoising preprocessing on the collected brain wave signals by using a dual-tree complex wavelet transform method.
6. The brain wave-based teaching tracking assistance system according to claim 5, wherein the brain wave preprocessing module employs a denoising threshold function f (Wt) for denoising preprocessing the collected brain wave signals by means of dual-tree complex wavelet transform h,l ) Comprises the following steps:
wherein, W h,l Represents the wavelet coefficient on the l scale of the h-th dual-tree complex wavelet transform,representing wavelet coefficients W h,l Corresponding wavelet coefficient estimated value obtained by denoising threshold function, wherein lambda represents a preselected threshold; sgn (·) is the sign of the return function, exp (·) is the sign of the exponential function.
7. The brain wave-based teaching tracking assistance system according to claim 1, wherein the method for performing analysis processing and waking/sleepy state recognition on the received brain wave signals in the system server is specifically:
acquiring a sample entropy estimation value of the brain wave signal, extracting theta rhythm wave, alpha rhythm wave and beta rhythm wave in the brain wave signal respectively, calculating average power spectral density G (theta) of the theta rhythm wave, average power spectral density G (alpha) of the alpha rhythm wave and average power spectral density G (beta) of the beta rhythm wave respectively, and further calculating to obtain average power ratio F of the theta rhythm wave and the alpha rhythm wave relative to the beta rhythm wave (θ+α)/β = (G (θ) + G (α))/G (β) and the average power ratio F of θ rhythm wave to β rhythm wave θ/β = G (θ)/G (β), thereby calculating the mean power ratio F as the sample entropy estimation value of the brain wave signal (θ+α)/β And the average power ratio F θ/β Forming an identification feature vector of the brain wave signal;
when the human brain is in a waking state or a sleepy state, the sample entropy estimation value of the brain wave signal and the average power ratio F (θ+α)/β And the average power ratio F θ/β The difference exists, so the sample entropy estimated values of the brain wave signals of the human brain in the waking state and the sleepy state and the average power ratio F are counted in advance (θ+α)/β And the average power ratio F θ/β And training a support vector machine classifier after markingAnd the system server performs classified recognition of the waking state and the drowsy state on the recognition feature vector of the received brain wave signal by using the trained support vector machine classifier so as to obtain a waking/drowsy state recognition result.
8. The brain wave-based teaching tracking assistance system according to claim 7, wherein the mean power spectral density G (θ) of θ rhythm waves, the mean power spectral density G (α) of α rhythm waves, and the mean power spectral density G (β) of β rhythm waves are calculated by:
wherein f is θh And f θl Respectively representing the upper limit value and the lower limit value of the frequency band of theta rhythm waves in the brain wave signal; f. of αh And f αl Respectively representing the upper limit value and the lower limit value of the frequency band of alpha rhythm waves in the brain wave signal; f. of βh And f βl Respectively representing the upper limit value and the lower limit value of the frequency band of the beta rhythm wave in the brain wave signal; p is a radical of X (f) Which represents the average power spectrum of the brain wave signal at frequency f.
9. The brain wave-based teaching tracking assistance system according to claim 7, wherein the sample entropy estimation value of the brain wave signal is obtained by:
1) Presetting the value of a distance tolerance threshold value r and the value of a vector dimension m, wherein r is more than 0, and the vector dimension m is a positive integer more than 1;
2) Processing the brain wave signal U to form a m-dimensional brain wave data vector X m (i)：
X m (i)＝[u(i),u(i+1),...,u(i+m-1)]，i∈{1,2,…,N-m+1}；
Wherein U (i) represents the value of the ith sampling point in the brain wave signal U, and N is the total number of the sampling points of the brain wave signal U;
3) For each different fetch of iRespectively calculating the corresponding brain wave data vector X of the ith sampling point in the brain wave signal U m (i) Brain wave data vector X corresponding to other sampling points m (j) Distance d [ X ] between m (i),X m (j)]：
d[X m (i),X m (j)]＝max{|u(i+k)-u(j+k)||k＝1,2,…,m-1}；
Wherein j belongs to {1,2, …, N-m +1}, and j is not equal to i;
4) For each value of i, respectively counting the corresponding distance value d [ X ] m (i),X m (j)]The number of smaller distance tolerance thresholds r is recorded asAnd recording the ratio of the number to the total distance number N-m asNamely:
5) The ratio corresponding to all the values of i when the dimension of the vector is m is solvedAverage ratio of B m (r), namely:
6) Let the vector dimension be m +1, repeat steps 2) -5) to obtain the corresponding average ratio B when the vector dimension is m +1 m+1 (r)；
7) Calculating a sample entropy estimation value SampEn (m, r, N) of the brain wave signal by the following formula:
SampEn(m,r,N)＝-ln[B m+1 (r)/B m (r)]。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979904.7A CN107657868A (en) | 2017-10-19 | 2017-10-19 | A kind of teaching tracking accessory system based on brain wave |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710979904.7A CN107657868A (en) | 2017-10-19 | 2017-10-19 | A kind of teaching tracking accessory system based on brain wave |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107657868A true CN107657868A (en) | 2018-02-02 |
Family
ID=61118028
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710979904.7A Pending CN107657868A (en) | 2017-10-19 | 2017-10-19 | A kind of teaching tracking accessory system based on brain wave |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107657868A (en) |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108335725A (en) * | 2018-03-09 | 2018-07-27 | 新华网股份有限公司 | Generate the method and device of test and evaluation report |
| CN108537704A (en) * | 2018-04-17 | 2018-09-14 | 深圳市心流科技有限公司 | Classroom evaluating method, device and computer readable storage medium |
| CN108542403A (en) * | 2018-03-09 | 2018-09-18 | 新华网股份有限公司 | Determine the method and apparatus of children's degree tired out |
| CN108670276A (en) * | 2018-05-29 | 2018-10-19 | 南京邮电大学 | Study attention evaluation system based on EEG signals |
| CN108830461A (en) * | 2018-05-23 | 2018-11-16 | 深圳市心流科技有限公司 | Instruction analysis method, server and computer readable storage medium |
| CN108888280A (en) * | 2018-05-24 | 2018-11-27 | 吉林大学 | Student based on electroencephalogramsignal signal analyzing listens to the teacher attention evaluation method |
| CN109035538A (en) * | 2018-10-16 | 2018-12-18 | 深圳美特优科技有限公司 | A kind of visiting personnel registration checking device based on recognition of face |
| CN109242799A (en) * | 2018-09-19 | 2019-01-18 | 安徽理工大学 | A kind of Wavelet noise-eliminating method of variable threshold value |
| CN109409281A (en) * | 2018-10-22 | 2019-03-01 | 河南科技大学 | A kind of noise-reduction method based on improved wavelet threshold function |
| CN110070770A (en) * | 2019-03-07 | 2019-07-30 | 成都工业学院 | A kind of education cognitive system based on brain electric installation |
| CN110151199A (en) * | 2019-03-29 | 2019-08-23 | 江苏理工学院 | A kind of private tutor's auxiliary system based on EEG signals |
| WO2019205253A1 (en) * | 2018-04-28 | 2019-10-31 | 深圳市科迈爱康科技有限公司 | Auxiliary learning and teaching method, terminal apparatus, and computer readable storage medium |
| WO2019223543A1 (en) * | 2018-05-23 | 2019-11-28 | 深圳市心流科技有限公司 | Teaching analysis method and server, and computer-readable storage medium |
| CN110658911A (en) * | 2018-06-29 | 2020-01-07 | 深圳市掌网科技股份有限公司 | Virtual reality course adjusting system and method, virtual reality helmet and processing device |
| CN110968950A (en) * | 2019-11-27 | 2020-04-07 | 武汉理工大学 | Human-computer cooperation disassembly sequence planning method based on personnel fatigue |
| CN111402643A (en) * | 2020-04-07 | 2020-07-10 | 符智博 | Teaching method, education system, equipment and medium based on electroencephalogram education system |
| CN111443799A (en) * | 2020-03-24 | 2020-07-24 | 周林文 | Auxiliary learning method based on brain-computer interface, terminal and computer storage medium |
| CN111954290A (en) * | 2018-03-30 | 2020-11-17 | Oppo广东移动通信有限公司 | Electronic device, power adjusting method and related product |
| CN112528853A (en) * | 2020-12-09 | 2021-03-19 | 云南电网有限责任公司昭通供电局 | Improved dual-tree complex wavelet transform denoising method |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139695A (en) * | 2015-09-28 | 2015-12-09 | 南通大学 | EEG collection-based method and system for monitoring classroom teaching process |
| CN106097264A (en) * | 2016-06-07 | 2016-11-09 | 西北工业大学 | Based on dual-tree complex wavelet and morphologic satellite telemetering data filtering method |
| CN106691440A (en) * | 2016-12-07 | 2017-05-24 | 中国民用航空总局第二研究所 | Controller fatigue detection method and system based on BP neural network |
| CN106951835A (en) * | 2017-03-03 | 2017-07-14 | 东华大学 | A kind of EEG signals noise remove method |
| CN206355046U (en) * | 2016-09-06 | 2017-07-28 | 国家电网公司高级培训中心 | One kind is listened to the teacher condition monitoring system |
| CN107184187A (en) * | 2017-07-03 | 2017-09-22 | 重庆大学 | Pulse Wave Signal Denoising processing method based on DTCWT Spline |
-
2017
- 2017-10-19 CN CN201710979904.7A patent/CN107657868A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105139695A (en) * | 2015-09-28 | 2015-12-09 | 南通大学 | EEG collection-based method and system for monitoring classroom teaching process |
| CN106097264A (en) * | 2016-06-07 | 2016-11-09 | 西北工业大学 | Based on dual-tree complex wavelet and morphologic satellite telemetering data filtering method |
| CN206355046U (en) * | 2016-09-06 | 2017-07-28 | 国家电网公司高级培训中心 | One kind is listened to the teacher condition monitoring system |
| CN106691440A (en) * | 2016-12-07 | 2017-05-24 | 中国民用航空总局第二研究所 | Controller fatigue detection method and system based on BP neural network |
| CN106951835A (en) * | 2017-03-03 | 2017-07-14 | 东华大学 | A kind of EEG signals noise remove method |
| CN107184187A (en) * | 2017-07-03 | 2017-09-22 | 重庆大学 | Pulse Wave Signal Denoising processing method based on DTCWT Spline |
Non-Patent Citations (2)
| Title |
|---|
| 郭兴明等: "基于小波变换和样本熵的心音识别研究", 《计算机应用研究》 * |
| 颜松等: "汽车驾驶员瞌睡状态脑电波特征提取的研究", 《中国生物医学工程学报》 * |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108335725A (en) * | 2018-03-09 | 2018-07-27 | 新华网股份有限公司 | Generate the method and device of test and evaluation report |
| CN108542403A (en) * | 2018-03-09 | 2018-09-18 | 新华网股份有限公司 | Determine the method and apparatus of children's degree tired out |
| CN111954290B (en) * | 2018-03-30 | 2023-04-18 | Oppo广东移动通信有限公司 | Electronic device, power adjusting method and related product |
| CN111954290A (en) * | 2018-03-30 | 2020-11-17 | Oppo广东移动通信有限公司 | Electronic device, power adjusting method and related product |
| CN108537704A (en) * | 2018-04-17 | 2018-09-14 | 深圳市心流科技有限公司 | Classroom evaluating method, device and computer readable storage medium |
| WO2019201215A1 (en) * | 2018-04-17 | 2019-10-24 | 深圳市心流科技有限公司 | Class teaching evaluating method and apparatus and computer readable storage medium |
| WO2019205253A1 (en) * | 2018-04-28 | 2019-10-31 | 深圳市科迈爱康科技有限公司 | Auxiliary learning and teaching method, terminal apparatus, and computer readable storage medium |
| CN108830461A (en) * | 2018-05-23 | 2018-11-16 | 深圳市心流科技有限公司 | Instruction analysis method, server and computer readable storage medium |
| WO2019223543A1 (en) * | 2018-05-23 | 2019-11-28 | 深圳市心流科技有限公司 | Teaching analysis method and server, and computer-readable storage medium |
| CN108888280A (en) * | 2018-05-24 | 2018-11-27 | 吉林大学 | Student based on electroencephalogramsignal signal analyzing listens to the teacher attention evaluation method |
| CN108670276A (en) * | 2018-05-29 | 2018-10-19 | 南京邮电大学 | Study attention evaluation system based on EEG signals |
| CN110658911A (en) * | 2018-06-29 | 2020-01-07 | 深圳市掌网科技股份有限公司 | Virtual reality course adjusting system and method, virtual reality helmet and processing device |
| CN109242799A (en) * | 2018-09-19 | 2019-01-18 | 安徽理工大学 | A kind of Wavelet noise-eliminating method of variable threshold value |
| CN109242799B (en) * | 2018-09-19 | 2021-10-12 | 安徽理工大学 | Variable-threshold wavelet denoising method |
| CN109035538A (en) * | 2018-10-16 | 2018-12-18 | 深圳美特优科技有限公司 | A kind of visiting personnel registration checking device based on recognition of face |
| CN109409281A (en) * | 2018-10-22 | 2019-03-01 | 河南科技大学 | A kind of noise-reduction method based on improved wavelet threshold function |
| CN110070770A (en) * | 2019-03-07 | 2019-07-30 | 成都工业学院 | A kind of education cognitive system based on brain electric installation |
| CN110151199A (en) * | 2019-03-29 | 2019-08-23 | 江苏理工学院 | A kind of private tutor's auxiliary system based on EEG signals |
| CN110968950A (en) * | 2019-11-27 | 2020-04-07 | 武汉理工大学 | Human-computer cooperation disassembly sequence planning method based on personnel fatigue |
| CN111443799A (en) * | 2020-03-24 | 2020-07-24 | 周林文 | Auxiliary learning method based on brain-computer interface, terminal and computer storage medium |
| CN111402643A (en) * | 2020-04-07 | 2020-07-10 | 符智博 | Teaching method, education system, equipment and medium based on electroencephalogram education system |
| CN112528853A (en) * | 2020-12-09 | 2021-03-19 | 云南电网有限责任公司昭通供电局 | Improved dual-tree complex wavelet transform denoising method |
| CN112528853B (en) * | 2020-12-09 | 2021-11-02 | 云南电网有限责任公司昭通供电局 | Improved dual-tree complex wavelet transform denoising method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107657868A (en) | A kind of teaching tracking accessory system based on brain wave | |
| Mammone et al. | Automatic artifact rejection from multichannel scalp EEG by wavelet ICA | |
| Cheong et al. | Feature extraction of EEG signal using wavelet transform for autism classification | |
| Xu et al. | Pattern recognition of motor imagery EEG using wavelet transform | |
| Jerritta et al. | Electrocardiogram-based emotion recognition system using empirical mode decomposition and discrete Fourier transform. | |
| Fathima et al. | Wavelet based features for epileptic seizure detection | |
| CN101690659A (en) | Brain wave analysis method | |
| CN103584872A (en) | Psychological stress assessment method based on multi-physiological-parameter integration | |
| CN114533086B (en) | Motor imagery brain electrolysis code method based on airspace characteristic time-frequency transformation | |
| CN105997064A (en) | Method for identifying human lower limb surface EMG signals (electromyographic signals) | |
| CN108309290A (en) | The automatic removal method of Muscle artifacts in single channel EEG signals | |
| CN110946576A (en) | Visual evoked potential emotion recognition method based on width learning | |
| Wang et al. | Classification of EEG signal using convolutional neural networks | |
| Patnaik et al. | Deep RNN learning for EEG based functional brain state inference | |
| CN115211858A (en) | Emotion recognition method and system based on deep learning and storable medium | |
| Feng et al. | Feature extraction algorithm based on csp and wavelet packet for motor imagery eeg signals | |
| Witteveen et al. | Comparison of a pragmatic and regression approach for wearable EEG signal quality assessment | |
| Hossain et al. | A deep convolutional autoencoder for automatic motion artifact removal in electrodermal activity | |
| CN113974557A (en) | Deep neural network anesthesia depth analysis method based on electroencephalogram singular spectrum analysis | |
| Shi et al. | Dynamic time-frequency feature extraction for brain activity recognition | |
| CN117883082A (en) | Abnormal emotion recognition method, system, equipment and medium | |
| Kapfo et al. | LSTM based Synthesis of 12-lead ECG Signal from a Reduced Lead Set | |
| Awang et al. | Analysis of EEG signals by eigenvector methods | |
| Mesin | Non-propagating components of surface electromyogram reflect motor unit firing rates | |
| CN115270847A (en) | Design decision electroencephalogram recognition method based on wavelet packet decomposition and convolutional neural network |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180202 |