A fine-grained brain load determination method based on clustering algorithm and related device

Abstract

The present invention provides a method and related device for determining fine-grained brain load based on a clustering algorithm, belonging to the technical field of ergonomics. The method of the present invention combines the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal to obtain a combined EEG feature; clusters the combined EEG feature to obtain several cluster centers; performs cosine similarity scoring on the several cluster centers to obtain several cosine similarity score results; and interpolates the several cosine similarity score results to obtain the brain load value for the entire time period. The present invention solves the problem of low accuracy in determining fine-grained brain load in the prior art.

Description

Technical Field

[0001] The present invention belongs to the technical field of ergonomics, and in particular relates to a fine-grained brain load determination method based on a clustering algorithm and a related device. Background Art

[0002] Most existing fine-grained brain load determination methods are based on experimental paradigms, which make it easier to categorize brain load levels. They first label EEG data, then use deep learning or machine learning to perform classification. Existing fine-grained brain load determination methods also directly calculate brain load by extracting EEG features (power spectral density ratios).

[0003] Methods based on experimental paradigms require labeling of brain load, but data in real-world scenarios is difficult to label. Furthermore, methods based on experimental paradigms use coarse-grained brain load calculations and cannot reflect detailed trends in load changes. Methods based on experimental paradigms are not practical because they are not based on real-world task scenarios. Methods based on experimental paradigms score load levels and are not specific to the individual experimenter (the same task may have different load levels for different individuals), making it difficult to implement applications such as individual load monitoring and reminders.

[0004] Brain load is directly calculated by extracting EEG features (power spectral density value ratio). This method only uses the power spectral density of EEG features, and the information utilization is not comprehensive, which leads to insufficient load calculation, resulting in low accuracy in determining fine-grained brain load. Summary of the Invention

[0005] The purpose of the present invention is to provide a fine-grained brain load determination method and related devices based on a clustering algorithm, so as to solve the problem of low accuracy in determining fine-grained brain load in the prior art.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a fine-grained brain load determination method based on a clustering algorithm, comprising the following steps:

[0008] Obtaining EEG signals, and performing filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals;

[0009] Perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition;

[0010] Calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition;

[0011] Performing dimensionality reduction on the obtained EEG signal to obtain a dimensionality-reduced EEG signal;

[0012] The power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal are merged to obtain the merged EEG features;

[0013] Cluster the merged EEG features to obtain several cluster centers;

[0014] Perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results;

[0015] Interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

[0016] A further improvement of the present invention is that, in the steps of obtaining EEG signals and filtering, whole-brain averaging, independent component analysis and artifact removal preprocessing the obtained EEG signals, the EEG signals are obtained specifically through an EEG device or a 32-channel EEG cap, and the obtained EEG signals are filtered, whole-brain averaged, independent component analysis and artifact removal preprocessing.

[0017] A further improvement of the present invention is that, in the step of performing wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition, the MATLAB wavelet packet decomposition toolkit is specifically used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

[0018] A further improvement of the present invention is that, in the step of reducing the dimension of the obtained EEG signal to obtain the EEG signal after dimensionality reduction, principal component analysis is specifically used to reduce the dimension of the obtained EEG signal to obtain the EEG signal after dimensionality reduction.

[0019] A further improvement of the present invention is that, in the step of clustering the merged EEG features to obtain a plurality of cluster centers, a k-means clustering algorithm is specifically used to cluster the merged EEG features to obtain a plurality of cluster centers.

[0020] A further improvement of the present invention is that, in the step of interpolating a plurality of cosine similarity score results to obtain the brain load value for the entire time period, a weighted summation method is specifically used to interpolate a plurality of cosine similarity score results to obtain the brain load value for the entire time period.

[0021] In a second aspect, the present invention provides a fine-grained brain load determination system based on a clustering algorithm, comprising a data acquisition module, a data preprocessing module, a wavelet packet decomposition module, a power spectrum density acquisition module, a data dimensionality reduction module, a data merging module, a data clustering module, a cosine similarity scoring module, and a data interpolation module;

[0022] The data acquisition module is used to acquire EEG signals;

[0023] The data preprocessing module is used to perform filtering, whole-brain averaging, independent component analysis and artifact removal preprocessing on the obtained EEG signals;

[0024] The wavelet packet decomposition module is used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition;

[0025] The power spectrum density acquisition module is used to calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition;

[0026] The data dimensionality reduction module is used to reduce the dimensionality of the obtained EEG signal to obtain the EEG signal after dimensionality reduction;

[0027] The data merging module is used to merge the power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal to obtain a merged EEG feature;

[0028] The data clustering module is used to cluster the combined EEG features to obtain several cluster centers;

[0029] The cosine similarity scoring module is used to perform cosine similarity scoring on a plurality of cluster centers to obtain a plurality of cosine similarity scoring results;

[0030] The data interpolation module is used to interpolate a plurality of cosine similarity scoring results to obtain the brain load value of the entire time period.

[0031] A further improvement of the present invention is that the data dimension reduction module uses principal component analysis to reduce the dimension of the obtained EEG signal to obtain a signal after dimension reduction.

[0032] In a third aspect, the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the above-described method for determining fine-grained brain load based on a clustering algorithm when executing the computer program.

[0033] In a fourth aspect, the present invention provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the above-mentioned method for determining fine-grained brain load based on a clustering algorithm.

[0034] Compared with the prior art, the present invention has the following beneficial effects:

[0035] This invention is an improved invention. Compared with existing fine-grained brain load determination methods based on clustering algorithms, on the one hand, the present invention combines the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal to obtain a combined EEG feature. It can be seen that the present invention takes more comprehensive factors into consideration, improving the accuracy of the fine-grained brain load determined later. On the other hand, the present invention clusters the combined EEG features to obtain several cluster centers. The clustering algorithm can fully and effectively utilize the multi-dimensional information of the EEG signal, making subsequent analysis and processing (interpolation to obtain brain load) more efficient and accurate. In addition, interpolating several cosine similarity score results to obtain brain load values ​​for the entire time period can ensure data integrity. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Flowchart of the fine-grained brain load determination method based on clustering algorithm of the present invention;

[0037] Figure 2 Schematic diagram of a fine-grained brain workload determination system based on clustering algorithm according to the present invention;

[0038] Figure 3 The brain load curves of different human bodies in the present invention are Figure 1 ;

[0039] Figure 4 The brain load curves of different human bodies in the present invention are Figure 2 ;

[0040] Figure 5 Schematic diagram of the structure of the electronic device of the present invention. DETAILED DESCRIPTION

[0041] In order to further understand the content of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0042] The present invention proposes a clustering-based fine-grained brain load determination method that combines the power spectral density of different frequency bands, the dimensionality-reduced EEG signal, and the EEG signal's onset time to obtain a combined EEG feature. The combined EEG feature is then clustered to obtain several cluster centers. These cluster centers are then scored using cosine similarity to obtain several cosine similarity scores. These cosine similarity scores are then interpolated to obtain a brain load value for the entire time period. Compared to existing technologies, this invention effectively addresses the low accuracy of fine-grained brain load determination in existing technologies.

[0043] Example 1:

[0044] The flowchart of the fine-grained brain load determination method based on clustering algorithm of the present invention is as follows: Figure 1As shown, the fine-grained brain load determination method based on clustering algorithm of the present invention includes the following steps:

[0045] S1. Obtain EEG signals, and perform filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals.

[0046] S2. Perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

[0047] S3. Calculate the power spectral density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition.

[0048] S4. Perform dimensionality reduction on the obtained EEG signal to obtain a dimensionality-reduced EEG signal.

[0049] S5. Merge the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal to obtain a merged EEG feature.

[0050] S6. Cluster the merged EEG features to obtain several cluster centers.

[0051] S7. Perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results.

[0052] S8. Interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

[0053] Example 2:

[0054] The schematic diagram of the fine-grained brain load determination system based on clustering algorithm of the present invention is as follows: Figure 2 As shown, the fine-grained brain load determination system based on clustering algorithm of the present invention includes a data acquisition module, a data preprocessing module, a wavelet packet decomposition module, a power spectrum density acquisition module, a data dimension reduction module, a data merging module, a data clustering module, a cosine similarity scoring module and a data interpolation module.

[0055] The data acquisition module is used to obtain EEG signals.

[0056] The data preprocessing module is used to filter, average the whole brain, perform independent component analysis and remove artifacts on the obtained EEG signals.

[0057] The wavelet packet decomposition module is used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

[0058] The power spectrum density acquisition module is used to calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition.

[0059] The data dimension reduction module is used to reduce the dimension of the obtained EEG signal to obtain the EEG signal after dimension reduction.

[0060] The data merging module is used to merge the power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal to obtain the merged EEG features.

[0061] The data clustering module is used to cluster the merged EEG features to obtain several cluster centers.

[0062] The cosine similarity scoring module is used to perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results.

[0063] The data interpolation module is used to interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

[0064] Example 3:

[0065] S1. Obtain EEG signals, and perform filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals.

[0066] First, EEG signals are acquired through an EEG device or a 32-channel EEG cap, and then the obtained EEG signals are filtered, averaged across the entire brain, subjected to independent component analysis, and preprocessed to remove artifacts.

[0067] Specifically, the EEG signal is first filtered with a notch filter to eliminate 50Hz power line interference. A Butterworth filter is then used to bandpass filter the EEG signal within the 1 to 30Hz frequency range. The EEG signal is then averaged across the entire brain and subjected to independent component analysis (ICA) to extract 20 independent components. Finally, artifacts such as electrocardiogram (ECG) and electrooculogram (EOG) interference are automatically identified and removed.

[0068] S2. Perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

[0069] In this step, the MATLAB wavelet packet decomposition toolkit (its function is sym6) is used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

[0070] S3. Calculate the power spectral density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition.

[0071] In this step, the power spectrum density of different frequency bands is calculated based on the continuous time-frequency information after wavelet packet decomposition (specifically, the power spectrum density of different frequency bands is obtained by calling the wpspectrum function in MATLAB).

[0072] The calculation formula of wavelet packet decomposition is:

[0073]

[0074] in n =0,1,2, indicating the wavelet packet determined by the orthogonal scaling function. and are two sets of coefficients, usually used to define filters or scaling functions and wavelet functions in wavelet packet decomposition. For the n The wavelet packet function of level , k Represents an index, and its domain is the integer domain. t is the independent variable of the wavelet packet function, representing the time domain. Specifically, this embodiment adopts 4-level wavelet packet decomposition, which can easily decompose different frequency bands.

[0075] S4. Perform dimensionality reduction on the obtained EEG signal to obtain a dimensionality-reduced EEG signal.

[0076] In this step, principal component analysis is specifically used to reduce the dimension of the obtained EEG signal to obtain the EEG signal after dimension reduction.

[0077] S5. Merge the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal to obtain a merged EEG feature.

[0078] In this step, FEATURE is specifically used to represent the process of merging the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal to obtain the merged EEG feature.

[0079] The process of merging the power spectral density of different frequency bands, the EEG signal after dimensionality reduction, and the start time of the EEG signal using FEATURE to obtain the merged EEG feature is specifically expressed as follows:

[0080] FEATURE=[the start time of the EEG signal, the EEG signal after dimensionality reduction, the power spectrum density ratio of different frequency bands, and the power corresponding to different frequency bands].

[0081] The power spectral density ratios of different frequency bands include three cases: theta / alpha, theta / beta and beta / (alpha+theta).

[0082] The powers corresponding to different frequency bands are alpha, beta, theta, delta, and gamma.

[0083] S6. Cluster the merged EEG features to obtain several cluster centers.

[0084] In this step, the k-means clustering algorithm is used to cluster the merged EEG features to obtain several cluster centers.

[0085] S7. Perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results.

[0086] The calculation formula for cosine similarity is:

[0087]

[0088] in, similarity is the cosine similarity, A is the cluster center feature vector, B are different scoring criteria vectors.

[0089] S8. Interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

[0090] In this step, a weighted sum method is used to interpolate the cosine similarity score results to obtain the brain load value of the entire time period. The weights of the interpolation of different cluster centers are The calculation formula is as follows:

[0091]

[0092] in, is the influence degree of different cluster centers on the load value in this time period ( Proportional to the weights of different cluster centers ), The calculation of is based on the exponential decrease of e, where the coefficient of decrease is proportional to the distance between the point and the cluster center (the farther the distance, the smaller the impact). represents the weight, The influence of all cluster centers is normalized (the weights are converted to between 0 and 1). is the distance between the point (the feature vector of the time segment, one time segment every three seconds) and the cluster center (Euclidean distance in the feature vector space), For all the cluster centers listed , is the MWL discrete degree of all points in a certain class in the cluster, k The value of is [1-7], k For an index, k Indicates that the weight values ​​of all cluster centers are calculated.

[0093] In order to verify the effectiveness of the fine-grained brain load determination method based on clustering algorithm proposed in this invention, this embodiment observes the trend of brain load curves of different human bodies to determine whether it can match the difficulty of the actual task (flight task). The results of the two brain load curves of different human bodies are as follows: Figure 3 and Figure 4 shown. Figure 3 and Figure 4 The horizontal axis represents time and the vertical axis represents brain load. Figure 3 The curve data shows that the brain load of different human bodies first decreases and then increases, which is consistent with the difficulty of the actual mission (flight mission). Figure 4 and Figure 3 It can be seen that the changes in brain load of different individual groups (also called a collection of many individuals) are similar, reflecting that the method of the present invention can reflect the actual situation of brain load, thereby verifying the effectiveness of the method of the present invention.

[0094] Example 4:

[0095] See also Figure 5 As shown, the present invention also provides an electronic device 100 for a fine-grained brain load determination method based on a clustering algorithm; the electronic device 100 includes a memory 101, at least one processor 102, a computer program 103 stored in the memory 101 and executable on the at least one processor 102, and at least one communication bus 104.

[0096] The memory 101 can be used to store the computer program 103. The processor 102 implements the steps of the clustering algorithm-based fine-grained brain load determination method described in Example 1 by running or executing the computer program stored in the memory 101 and calling the data stored in the memory 101. The memory 101 may mainly include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as a sound playback function, an image playback function, etc.); the data storage area may store data (such as audio data) generated based on the use of the electronic device 100. In addition, the memory 101 may include non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device.

[0097] The at least one processor 102 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The processor 102 may be a microprocessor or any conventional processor, etc. The processor 102 is the control center of the electronic device 100 and connects various parts of the entire electronic device 100 using various interfaces and lines.

[0098] The memory 101 in the electronic device 100 stores a plurality of instructions to implement a fine-grained brain load determination method based on a clustering algorithm. The processor 102 may execute the plurality of instructions to implement:

[0099] Obtaining EEG signals, and performing filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals;

[0100] Perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition;

[0101] Calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition;

[0102] Performing dimensionality reduction on the obtained EEG signal to obtain a dimensionality-reduced EEG signal;

[0103] The power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal are merged to obtain the merged EEG features;

[0104] Cluster the merged EEG features to obtain several cluster centers;

[0105] Perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results;

[0106] Interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

[0107] Example 5:

[0108] If the module / unit integrated in the electronic device 100 is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form, etc. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory and read-only memory (ROM, Read-Only Memory).

[0109] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0110] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0111] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0112] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0113] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A fine-grained brain load determination method based on clustering algorithm, characterized in that: The following steps are involved: Obtaining EEG signals, and performing filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals; Perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition; Calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition; Performing dimensionality reduction on the obtained EEG signal to obtain a dimensionality-reduced EEG signal; The power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal are merged to obtain the merged EEG features; Cluster the merged EEG features to obtain several cluster centers; Perform cosine similarity scoring on several cluster centers to obtain several cosine similarity scoring results; Interpolate several cosine similarity scoring results to obtain the brain load value of the entire time period.

2. The fine-grained brain load determination method based on clustering algorithm according to claim 1 is characterized in that: In the steps of obtaining EEG signals and performing filtering, whole-brain averaging, independent component analysis, and artifact removal preprocessing on the obtained EEG signals, the EEG signals are specifically obtained through an EEG device or a 32-channel EEG cap, and the obtained EEG signals are filtered, whole-brain averaged, independent component analysis, and artifact removal preprocessing.

3. The fine-grained brain load determination method based on clustering algorithm according to claim 1, characterized in that: In the step of performing wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition, the MATLAB wavelet packet decomposition toolkit is specifically used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition.

4. The fine-grained brain load determination method based on clustering algorithm according to claim 1, characterized in that: In the step of reducing the dimension of the obtained EEG signal to obtain the EEG signal after dimension reduction, principal component analysis is specifically used to reduce the dimension of the obtained EEG signal to obtain the EEG signal after dimension reduction.

5. The fine-grained brain load determination method based on clustering algorithm according to claim 1, characterized in that: In the step of clustering the merged EEG features to obtain a plurality of cluster centers, a k-means clustering algorithm is specifically used to cluster the merged EEG features to obtain a plurality of cluster centers.

6. The fine-grained brain load determination method based on clustering algorithm according to claim 1, characterized in that: In the step of interpolating a plurality of cosine similarity scoring results to obtain the brain load value for the entire time period, a weighted sum method is specifically used to interpolate a plurality of cosine similarity scoring results to obtain the brain load value for the entire time period.

7. A fine-grained brain load determination system based on clustering algorithm, characterized in that: It includes data acquisition module, data preprocessing module, wavelet packet decomposition module, power spectrum density acquisition module, data dimension reduction module, data merging module, data clustering module, cosine similarity scoring module and data interpolation module; The data acquisition module is used to acquire EEG signals; The data preprocessing module is used to perform filtering, whole-brain averaging, independent component analysis and artifact removal preprocessing on the obtained EEG signals; The wavelet packet decomposition module is used to perform wavelet packet decomposition on the preprocessed EEG signal to obtain continuous time-frequency information after wavelet packet decomposition; The power spectrum density acquisition module is used to calculate the power spectrum density of different frequency bands based on the continuous time-frequency information after wavelet packet decomposition; The data dimensionality reduction module is used to reduce the dimensionality of the obtained EEG signal to obtain the EEG signal after dimensionality reduction; The data merging module is used to merge the power spectrum density of different frequency bands, the EEG signal after dimensionality reduction and the start time of the EEG signal to obtain a merged EEG feature; The data clustering module is used to cluster the combined EEG features to obtain several cluster centers; The cosine similarity scoring module is used to perform cosine similarity scoring on a plurality of cluster centers to obtain a plurality of cosine similarity scoring results; The data interpolation module is used to interpolate a plurality of cosine similarity scoring results to obtain the brain load value of the entire time period.

8. The fine-grained brain load determination system based on clustering algorithm according to claim 7, characterized in that: The data dimension reduction module uses principal component analysis to reduce the dimension of the obtained EEG signal to obtain the EEG signal after dimension reduction.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the processor implements the steps of the fine-grained brain workload determination method based on a clustering algorithm according to any one of claims 1 to 6.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the computer program implements the steps of the fine-grained brain workload determination method based on a clustering algorithm according to any one of claims 1 to 6.

Images