CN108670276A - Study attention evaluation system based on EEG signals - Google Patents

Abstract

The study attention evaluation system based on EEG signals that the invention discloses a kind of, including：Brain wave acquisition subsystem is pre-processed for acquiring EEG signals, removes Hz noise and baseline, and EEG signals are transmitted to brain electricity analytical subsystem；Brain electricity analytical subsystem extracts the feature vector that temporal signatures and frequency domain character are classified as attention, feature vector is inputted attention grader, obtains attention grade for analyzing collected EEG signals；Learning state display alarm subsystem feeds back to the attention grade that brain electricity analytical subsystem obtains at student and teacher；It realizes the real-time monitoring of attention of student, solves the technical issues of student's learning state is difficult to objective evaluation.

Description

Learning attention evaluation system based on EEG signals

technical field

The invention belongs to the technical field of attention detection, and in particular relates to a learning attention evaluation system based on electroencephalogram signals.

Background technique

The detection of classroom learning status is an important link in the teaching and learning process. Traditional classroom testing methods are often based on classroom questions, quizzes, and observing student expressions. Teachers often cannot take into account the learning status of each student, and the judgment results are highly random and subjective. Therefore, it is particularly important to realize the objective and effective detection of students' learning status. In recent years, with the rapid development of brain-computer interface technology, it has become possible to capture human mental activities only through the collection and analysis of EEG signals. As a kind of psychological activity, attention plays an important role in students' learning activities. The degree of concentration of students' classroom attention reflects the quality of students' lectures and learning status to a large extent. A large number of literatures point out that the activity of each rhythmic wave of the EEG signal is closely related to the attention state of the human body.

Contents of the invention

In order to solve the above problems, the present invention proposes a learning attention evaluation system based on EEG signals, which realizes real-time monitoring of students' attention and solves the technical problem that it is difficult to objectively evaluate students' learning status.

The present invention adopts the following technical scheme, equips each student with a portable EEG acquisition device, uses the wearable EEG acquisition device to collect the students' EEG data, and analyzes the EEG data of the students in the class to evaluate the students' attention in real time Status, real-time reflection of students' learning and teacher's teaching situation, and feedback the results to students and teachers via mobile APP or PC program, so as to achieve the effect of real-time feedback of learning attention, a learning attention based on EEG signals The power evaluation system specifically includes: EEG acquisition subsystem, EEG analysis subsystem and learning status display reminder subsystem, wherein,

The EEG acquisition subsystem is used to collect EEG signals, perform preprocessing, remove power frequency interference and baseline, and transmit EEG signals to the EEG analysis subsystem;

The EEG analysis subsystem is used to analyze the collected EEG signals, extract time domain features and frequency domain features as feature vectors for attention classification, and input feature vectors into attention classifiers to obtain attention levels;

The learning status display reminder subsystem feeds back the attention level obtained by the EEG analysis subsystem to the students and teachers.

Preferably, the EEG acquisition subsystem uses TI's ADS1298 chip as an analog front end to collect the EEG signals of the parietal lobe P7, P3, Pz, P4 and P8 leads; uses TI's CC2652R microprocessor to perform EEG Signal preprocessing and EEG signal transmission.

Preferably, the pre-processing of the EEG acquisition subsystem includes band-pass filtering and removing the baseline, constructing a digital band-pass filter to remove 50Hz power frequency interference; using the least squares method to fit the baseline, and converting the baseline from the original EEG signal removed.

Preferably, the communication between the EEG acquisition subsystem and the EEG analysis subsystem is realized through a Zigbee star network, and the EEG analysis subsystem acts as a host to request data from the EEG acquisition subsystem in a polling manner. After receiving the instruction, the electrical acquisition subsystem returns the EEG signal data of a specified length.

Preferably, the EEG acquisition subsystem includes a fixed top beam, including a power supply and data transmission processing unit, the data transmission processing unit is used to receive EEG signal data, and preprocess and transmit the EEG signal; The dry electrode, the dry electrode for collection is connected to the top beam through elastic components; the ear reference electrode, the dry electrode for collection and the ear reference electrode transmit the collected signal to the data transmission and processing unit.

Preferably, the frequency domain feature extracted by the EEG analysis subsystem is the power value of the four frequency bands of delta wave, theta wave, alpha wave and beta wave, and the time domain feature is sample entropy and standard deviation, first-order difference mean, second-order difference The average value of the order difference is used to construct the feature vector, and the attention classifier used is a probabilistic neural network PNN, which divides the attention level into three categories: high, medium, and low.

Preferably, the training samples of the probabilistic neural network PNN are public databases and experimental data, wherein the public database is the DEAP database, and the experimental data are the self-assessment level data of the testee's attention obtained from the attention experiment of the testee.

Preferably, the EEG analysis subsystem and the learning state display reminder subsystem adopt a combination of a PC terminal program and a mobile APP to feed back attention levels to teachers and students, and each mobile terminal and PC terminal are connected wirelessly .

Preferably, the learning status display reminder subsystem includes a teacher system and a student system, wherein,

The faculty system includes:

The student's attention display module is used to display the current attention level of all students and the time-varying curve of the attention level of individual students;

Low attention warning module, used to judge and display students whose current attention level is lower than the set value;

Student systems include:

The student's attention display module is used to display the current student's attention level;

The attention warning module is used to remind students whose attention level is lower than the set value for a set time through sound and light alarms.

Beneficial effects achieved by the invention: the present invention is a learning attention evaluation system based on EEG signals, which realizes real-time monitoring of students' attention and solves the technical problem that students' learning status is difficult to objectively evaluate; the present invention adopts a wearable EEG signal The acquisition subsystem uses a microprocessor integrated with a Zigbee module, which reduces the volume and weight of the EEG acquisition subsystem, reduces costs, is easy to wear, and has good portability and practicability; the PNN algorithm is used to classify EEG signals , has the advantages of simple learning process, fast training speed, accurate classification, and good fault tolerance; the learning status detection based on human physiological signals uses external equipment to detect the learning status of each student, which greatly reduces the teaching pressure of teachers, and the evaluation results are objective Fairness; in addition, after the teaching is over, the statistical analysis of the performance of the students is carried out to assist the teachers to objectively evaluate the learning effect of the students, and to continuously improve the teaching methods.

Description of drawings

Fig. 1 is a system architecture diagram of an embodiment of the present invention;

2 is a structural diagram of the EEG acquisition subsystem;

Fig. 3 is the graphical interface under the connection state of the teacher's PC;

Fig. 4 is the attention display interface of the teacher's PC end;

Figure 5 shows the interface diagram of the teacher's mobile APP;

Figure 6 shows the APP interface diagram of the student mobile phone terminal.

Reference signs: 1-top beam for fixing, 2-dry electrode for collection, 3-ear reference electrode, 4-serial port number of computer equipment, 5-connection status of students, 6-confirmation key, 7-students' attention Overall status display window, 8-single student's attention curve window, 9-low attention warning window, 10-students' overall attention situation statistics window, 11-low attention list display window, 12-student attention evaluation report window , 13-personal attention situation display window.

Detailed ways

The technical solutions of the present invention will be further elaborated below according to the drawings and in conjunction with the embodiments.

As shown in Figure 1, the following technical scheme adopted by the present invention is a learning attention evaluation system based on EEG signals, which specifically includes: EEG acquisition subsystem, EEG analysis subsystem and learning status display reminder subsystem, wherein ,

The EEG acquisition subsystem is used to collect EEG signals, preprocess the analog signals, remove power frequency interference and baseline, and transmit the EEG signals to the EEG analysis subsystem;

The EEG analysis subsystem is used to analyze the collected EEG signals, extract time domain features and frequency domain features as feature vectors for attention classification, and input feature vectors into attention classifiers to obtain attention levels;

The learning status display reminder subsystem feeds back the attention level obtained by the EEG analysis subsystem to the students and teachers.

Preferably, the EEG acquisition subsystem uses TI's ADS1298 chip as an analog front end to collect the EEG signals of the parietal lobe P7, P3, Pz, P4 and P8 leads; uses TI's CC2652R microprocessor to perform EEG Signal preprocessing and EEG signal transmission.

Preferably, the pre-processing of the EEG acquisition subsystem includes band-pass filtering and removing the baseline, constructing a digital band-pass filter to remove 50Hz power frequency interference; using the least squares method to fit the baseline, and converting the baseline from the original EEG signal removed.

Preferably, the communication between the EEG acquisition subsystem and the EEG analysis subsystem is realized through a Zigbee star network, and the EEG analysis subsystem acts as a host to request data from the EEG acquisition subsystem in a polling manner. After receiving the instruction, the electrical acquisition subsystem returns the EEG signal data of a specified length.

Preferably, the EEG acquisition subsystem includes a fixed top beam, including a power supply and data transmission processing unit, the data transmission processing unit is used to receive EEG signal data, and preprocess and transmit the EEG signal; The dry electrode, the dry electrode for collection is connected to the top beam through elastic components; the ear reference electrode, the dry electrode for collection and the ear reference electrode transmit the collected signal to the data transmission and processing unit.

Preferably, the frequency domain feature extracted by the EEG analysis subsystem is the power value of the four frequency bands of delta wave, theta wave, alpha wave and beta wave, and the time domain feature is sample entropy, standard deviation, first-order difference mean and second-order difference. The average value of the order difference is used to construct the feature vector, and the attention classifier used is a probabilistic neural network PNN, which divides the attention level into three categories: high, medium, and low.

Preferably, the training samples of the probabilistic neural network PNN are public databases and experimental data, wherein the public database is the DEAP database, and the experimental data are the self-assessment level data of the testee's attention obtained from the attention experiment of the testee.

Preferably, the EEG analysis subsystem and the learning status display reminder subsystem use a combination of PC client and mobile APP to feed back the attention level to teachers and students, and the mobile phone and PC are connected wirelessly.

Preferably, the learning status display reminder subsystem includes a teacher system and a student system, wherein,

The faculty system includes:

The student's attention display module is used to display the current attention level of all students and the time-varying curve of the attention level of individual students;

Low attention warning module, used to judge and display students whose current attention level is lower than the set value;

Student systems include:

The student's attention display module is used to display the current student's attention level;

The attention warning module is used to remind students whose attention level is lower than the set value for a set time through sound and light alarms.

The main implementation process of the present invention is as follows:

Step 1: The EEG acquisition subsystem collects the students' EEG signal data, and after preprocessing, sends the EEG signal data to the EEG analysis subsystem in a wireless manner.

In this embodiment, the EEG acquisition subsystem is specifically shown in Figure 2. The EEG acquisition subsystem can collect the parietal lobe of the student. When in use, the dry electrode is close to the head, and the reference electrode is clamped on the earlobe.

In order to reduce the volume of the EEG acquisition subsystem, the analog front-end adopts the ADS1298 chip produced by TI Company, which has high sampling accuracy and stability. The sampling rate in this embodiment is 250Hz.

The preprocessing of the EEG signal includes band-pass filtering and removing the baseline, constructing a 0.1-40Hz digital band-pass filter to remove 50Hz power frequency interference; using the least square method to fit the baseline, and removing the baseline from the original EEG signal . In this embodiment, the CC2652R microprocessor produced by TI Company is used to realize signal preprocessing and transmission. The processor also has an ARM Cortex M4 core and a Zigbee communication module, which reduces costs while meeting technical requirements.

Step 2: Data interaction between the EEG acquisition subsystem and the EEG analysis system is performed wirelessly. The EEG analysis subsystem first obtains the EEG signal data by reading the serial port buffer area, and obtains the EEG signals of each lead through protocol analysis. The signal is displayed as a time-domain waveform; the time-domain feature and frequency-domain feature are extracted as the feature vector of attention classification, and the feature vector is input into the attention classifier to obtain the attention level and reflect the learning status of students.

In this embodiment, the EEG analysis subsystem is set on the PC side, and the graphical interface of the teacher's PC client connection state is shown in FIG. 3 . In Figure 3, COM shows the connected device, which is used to select the serial port number of the computer device. The student's connection status 5 shows the student's connection status. If it is not connected, it will be red, and if it is connected, it will be green. When all students complete the connection, they can press OK Press key 6 "Next" to enter the formal detection mode, that is, to enter the attention display interface.

In order to reduce the energy consumption of the system and adapt to the real-time requirements of classroom attention detection, the polling method is used to obtain the EEG signal data of each student. After the EEG analysis subsystem and the EEG acquisition subsystem are connected, the coordinator The Zigbee module sends data requests to each EEG acquisition subsystem sequentially at fixed time intervals; the EEG acquisition subsystem is in a dormant state when it does not receive the request instruction to reduce overall power consumption, and it will be awakened immediately after receiving the request instruction. After the data transmission is completed, it will return to the dormant state again.

The selection of classification features is the key to attention classification. The present invention uses the combination of time domain and frequency domain features to complete feature extraction. The fast Fourier transform (FFT) algorithm is used to extract the power value of the EEG signal in the four frequency bands of delta wave, theta wave, alpha wave and beta wave, which has the characteristics of fast operation speed. Using sample entropy, standard deviation, and first- and second-order difference means as time-domain features effectively reflects the time complexity of EEG signals.

After the feature extraction is completed, the feature vectors composed of each feature are sent to the attention classifier, and the probabilistic neural network (PNN) is used as the attention classifier, and the attention level is divided into three categories: high, medium and low, and the public database and experimental data are used. Combined way to train the PNN network. Among them, the experimental data part requires the subjects to watch the clips of the online class, and the self-evaluation level of attention after the class is used as input; the public database uses the DEAP database.

Step 3: The learning status display reminder subsystem feeds back the attention level obtained by the EEG analysis subsystem to the students and teachers. The learning status display reminder subsystem is set with the PC terminal. The level is fed back to the teacher. Figure 4 shows the attention display interface of the teacher's PC client. The upper left corner will display the class currently in class. Click the start button to count the student's attention level; click the get out of class button to stop the statistics.

The overall display window 7 of the student's attention situation shows the current overall student attention level, the high level of attention concentration is green, the medium is yellow, and the low grade is red. Clicking on someone's seat in the overall display window 7 of the student's attention situation can Show this person's attention curve in the attention curve window 8 of single student, the low attention early warning module of teacher system judges and counts the low student list of current attention level, shows in the low attention early warning window 9, uses Useful for alerting teachers to look at students whose current attention levels are too low.

Figure 5 shows the interface diagram of the teacher's mobile APP. The student attention display module of the teacher system classifies the students according to the attention level, high school and low level and counts the proportion of students in the three levels. The statistics window 10 of the overall student attention situation shows the attention level of all students, and the student attention evaluation report Window 12 shows the student proportions of the three grades of attention. The low-attention warning module of the teacher system judges and counts the list of students with low attention levels at present, and displays them in the low-attention list display window 11 of the mobile APP, reminding the teacher to check the students with low current attention levels

Figure 6 shows the APP interface diagram of the student mobile phone terminal. The student's attention display module of the student system displays the current student's attention level in the individual attention situation display window 13.

The attention warning module of the student system reminds the students who are in the low attention level setting time through the sound and light alarm.

In this embodiment, the connection between the mobile phone terminal and the PC terminal is completed in a wireless manner, and the data transmission is completed in the form of building a WiFi local area network.

In addition, after the teaching is over, statistical analysis and evaluation sharing of students' performance is carried out to assist teachers to objectively evaluate students' learning effects and continuously improve teaching methods.

The above schematically describes the present invention and its implementation, which is not restrictive, and what is shown in the drawings is only one of the implementations of the present invention, and the actual structure is not limited thereto. Therefore, if a person of ordinary skill in the art is inspired by it, and without departing from the purpose of the invention, without creatively designing a structure and an embodiment similar to the technical solution, it shall fall within the scope of protection of this patent.

Claims (9)

1. a kind of study attention evaluation system based on EEG signals, which is characterized in that including：Brain wave acquisition subsystem, brain Electroanalysis subsystem and learning state display alarm subsystem, wherein

Brain wave acquisition subsystem is pre-processed for acquiring EEG signals, removes Hz noise and baseline, and by EEG signals It is transmitted to brain electricity analytical subsystem；

Brain electricity analytical subsystem extracts temporal signatures and frequency domain character as attention point for analyzing collected EEG signals Feature vector is inputted attention grader, obtains attention grade by the feature vector of class；

Learning state display alarm subsystem feeds back to the attention grade that brain electricity analytical subsystem obtains at student and teacher.

2. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the brain electricity Acquisition subsystem uses the ADS1298 chips of TI companies as AFE(analog front end), acquires the brain of top P7, P3, Pz, P4 and P8 lead Electric signal；The pretreatment of EEG signals is carried out using the CC2652R microprocessors of TI companies and EEG signals transmit.

3. the study attention evaluation system according to claim 1 or 2 based on EEG signals, which is characterized in that described It includes bandpass filtering and removal baseline that brain wave acquisition subsystem, which carries out pretreatment, constructs digital band-pass filter, removes 50Hz works Frequency interferes；It is removed from original EEG signals using least square fitting baseline, and by baseline.

4. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the brain electricity Communication between acquisition subsystem and brain electricity analytical subsystem passes through the star-like networking realizations of Zigbee, brain electricity analytical subsystem conduct For host to brain wave acquisition subsystem request data in a manner of poll, brain wave acquisition subsystem returns specified length after receiving instruction The EEG signals data of degree.

5. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the brain electricity Acquisition subsystem includes fixation top beam, includes power supply and data transmission and processing unit, data transmission and processing unit is for receiving EEG signals data, and EEG signals are pre-processed and transmitted；Dry electrode is acquired, acquisition passes through elastic portion with dry electrode Part is connected on top beam；Ear's reference electrode acquires dry electrode and ear's reference electrode by the signal transmission of acquisition to number According to transmission processing unit.

6. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the brain electricity Analyzing subsystem extraction frequency domain character be δ waves, four θ waves, α waves and β waves frequency ranges performance number, temporal signatures be Sample Entropy and Standard deviation, first-order difference mean value, second differnce mean value, construction feature vector, the attention grader used is probabilistic neural net Attention grade is divided into high, medium and low three classes by network PNN.

7. the study attention evaluation system according to claim 6 based on EEG signals, which is characterized in that the probability The training sample of neural network PNN is public database and experimental data, and wherein public database is DEAP databases, tests number According to carry out the self-appraisal attention level data of testee that attention is tested to testee.

8. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the brain electricity Between analyzing subsystem and learning state display alarm subsystem using pc client in such a way that cell phone application is combined to teacher With Students ' Feedback attention grade, mobile phone terminal is wirelessly connect with the ends PC.

9. the study attention evaluation system according to claim 1 based on EEG signals, which is characterized in that the study Status display reminding subsystem includes teacher's system and student system, wherein

Teacher's system includes：

Attention of student display module, the attention grade of attention grade and individual student for showing current all students Change over time curve；

Low attention warning module, for judging and showing that current attention grade is less than the student of setting value；

Student system includes：

Attention of student display module, the attention grade for showing current student；

Attention warning module, for pointing out the student that power grade is less than setting value setting time by sound-light alarm.