A group brain-computer interface collaborative control system and method based on SSVEP

Abstract

The present invention discloses a group brain-computer interface collaborative control system and method based on SSVEP. The control system includes: a brain-controlled game stimulation module, a data receiving module, a process control module, a data parsing module, a pattern recognition module, and a message communication module. The real-time EEG data of the group EEG acquisition device is sent to the control system. The control system parses and encapsulates it into 8-channel signals, obtains the high-order cumulants of the group multi-dimensional EEG signals through the pattern recognition module, and extracts features using FBCCA to achieve collaborative operation of the group brain-controlled game. This control system is flexibly applicable to various scenarios, can stimulate the rehabilitation enthusiasm of patients, and improve the rehabilitation effect. This application uses collaborative technology to process EEG signals, has good stability and real-time performance, high data processing efficiency, the system can efficiently extract the features of high-dimensional EEG signals, the accuracy of synchronous decoding of group EEG signals is ≥95%, the response speed is ≤3s, and the recognition accuracy and speed are comparable to those of a single BCI.

Description

Technical Field

[0001] The present invention relates to the technical field of bioelectrical signal processing, and particularly to a group brain-computer interface collaborative control system and method based on SSVEP. Background Art

[0002] Human society determines that an individual cannot act alone without a group, which reflects the social nature of people. The inevitable result of individual development is to move closer to the group. As a new way for humans to interact with the external environment, research on the brain-computer interface (Brain Computer Interface, BCI) has become relatively mature. Common BCIs include: motor imagery (MI), motion visual evoked potentials (mVEP), and steady-state visual evoked potential (SSVEP). However, most current brain-computer interfaces are implemented based on a single system, focusing on the interaction between a single subject and the machine, which deviates from the essential attribute of humans' participation in group competition and cooperation behaviors. Group BCI refers to using relatively mature BCI technology to simultaneously collect and process the electroencephalogram (EEG) signals of each subject within the same group brain-computer interface system, and then using collaborative technology to achieve real-time interactive control of multiple subjects on the same external device, thereby realizing the group control system of BCI. The final link in the development of BCI theory is to be used in various complex environments, and the collaborative cooperation between multiple people will be more efficient. In medical rehabilitation, the rehabilitation training of a single patient may have little effect. By taking advantage of the social attribute that people cannot be separated from the group, certain rules are formulated to establish a competitive or cooperative relationship between the group system of patients and others, increasing the entertainment of the rehabilitation process, and thus accelerating the process of rehabilitation and reconstruction of motor function. Summary of the Invention

[0003] The main purpose of the embodiments of the present invention is to propose a group brain-computer interface collaborative control system and method based on SSVEP, aiming to design a system suitable for various group brain-computer interactions and collaborative controls.

[0004] The technical solution for the present invention to solve the above technical problems is to provide a group brain-computer interface collaborative control system based on SSVEP, including:

[0005] A brain-controlled game stimulation module, configured to display a flashing block with a visual stimulation function and to display the scores of competitors;

[0006] A data receiving module, configured to collect the EEG data of multiple EEG acquisition devices through the TCP / IP service protocol;

[0007] A process control module, configured to receive the EEG data from the data receiving module and send the return message to the brain-controlled game stimulation module for interaction;

[0008] A data parsing module, configured to process the EEG data from the process control module;

[0009] A pattern recognition module, configured to perform filtering and singular value decomposition preprocessing on the EEG data processed by the data parsing module, decode the EEG signal through the FBCCA algorithm, and return the result to the process control module;

[0010] A message communication module, configured to implement the binding of the UDP socket and the host port number, and set the status flag bit of the brain-controlled game stimulation module according to the process status.

[0011] Further, the TCP / IP service protocol includes 8 EEG data channel bits, 1 flag bit, 1 power bit, and 1 redundant channel bit.

[0012] Further, the data parsing module includes: a filter unit and an FBCCA recognition algorithm unit.

[0013] To solve the above technical problems, the present invention also proposes a group brain-computer interface control method based on SSVEP, including the following steps:

[0014] (1) Establish a TCP / IP connection with multiple EEG acquisition devices, collect the EEG data of the multiple EEG acquisition devices in real time, and send it to the group BCI control system through the TCP / IP communication protocol;

[0015] (2) Select different game modes in the brain-controlled game stimulation module selection interface, determine the stimulation mode, and send the start flag bit and end flag bit information to the process control module;

[0016] (3) When the process control module detects that the start flag bit information sent by the brain-controlled game stimulation module is "1" and the end flag bit information is "0", send a START signal, first empty the recognition result and the EEG data, and start a multi-process to receive the 11-channel EEG data of different EEG acquisition devices;

[0017] (4) Encapsulate into an EEG signal through the data parsing module;

[0018] (5) Write the encapsulated original EEG signal into a stack buffer with a depth of 11 bits;

[0019] (6) Send the original data in the stack buffer to the filter for preprocessing of the EEG signal;

[0020] (7) The preprocessed data is decoded through the FBCCA recognition algorithm to obtain the recognition results of different subjects;

[0021] (8) The recognition result is forwarded to the process control module, and the EEG signal recognition results of different subjects are printed on the console for debugging;

[0022] (9) While in step (7), the recognition result is forwarded to the brain-controlled game stimulation module, and the score is calculated based on the binary classification accuracy and the display is updated;

[0023] (10) Set the start flag bit information received in step (3) to "0", and wait again for the flag bit information sent by the brain-controlled game stimulation module in step (3);

[0024] (11) When the brain-controlled game stimulation module sends the end flag bit information "1", the process control module closes the communication interface.

[0025] Further, the step of encapsulating into EEG signals by the data parsing module includes:

[0026] When the data parsing module receives the START signal from the process control module, it sends a START instruction to the TCP communication module to start receiving the message queues transmitted by different EEG acquisition devices;

[0027] When the message queue content is not empty and the current time is in the task state, the data parsing module will verify and remove the frame headers and trailers of the data frames to ensure the data is valid. Each data frame is 33 bytes in size and contains 11-bit depth information, consisting of 8 EEG data channel bits, a tag bit, a power bit, and a redundancy bit. The 8-channel EEG information in each frame is added to the stack buffer variable, and the remaining information is also cached;

[0028] When the time delay of the process control module reaches 3 seconds, it sends an end-of-task command, and the loop for receiving EEG signal data frames stops.

[0029] Further, the step of sending the original data in the stack buffer area to the filter for EEG signal preprocessing includes:

[0030] Set the cut-off frequencies of the Butterworth filter to f0 = 50Hz, f s = 1KHz, and the quality factor Q = 35 to determine the bandwidth of the filter. The formula is:

[0031]

[0032] Calculate the forward transfer coefficient b k , and the formula is:

[0033]

[0034] Calculate the reverse transfer coefficient a m , and the formula is:

[0035]

[0036] The size of the input raw buffer data for each channel is 3000 sampling points, and the size of the input matrix is 8 * 3000. The data passes through a Butterworth forward filter to obtain the output of the same dimension. The specific formula is as follows:

[0037]

[0038] The output of the Butterworth forward filter is used as the input of the Butterworth reverse filter. The specific formula is as follows:

[0039]

[0040] The results of the forward transfer filter and the reverse transfer filter are combined so that the final output is consistent with the original data in phase and dimension. The formula is:

[0041] y filtfilt (n) = y forward (n) + y backward (n)

[0042] Delete the axis with dimension 1 in the output result, reduce the matrix dimension, and add it to the preprocessing temporary cache variable.

[0043] Furthermore, the steps of decoding the preprocessed data through the FBCCA recognition algorithm to obtain the recognition results of different subjects include:

[0044] Create a multi-band Chebyshev type I filter, return the filter coefficients, and perform frequency feature extraction and noise reduction processing on the data;

[0045] Perform canonical correlation analysis on the EEG template data and the preprocessed EEG signals. Calculate the vector product of the preprocessed data and the template signal to obtain the CCA input data. The calculation formula is:

[0046] data svd = data processed ×Q template

[0047] The output is used as the input of singular value decomposition, and the left singular matrix, singular value matrix, and right singular matrix are output. Based on singular value decomposition, select the three largest singular values and calculate the CCA index ρ. The formula is:

[0048] ρ = 1.25×S0 + 0.67×S1 + 0.5×S2

[0049] Through the feedback weight W fb Perform weighted processing on the index ρ to adjust the weighting degree in the i-th frequency range. The formula is:

[0050]

[0051] Determine the frequency range according to the index P, obtain an identification result and return it.

[0052] Compared with the prior art, the present application has the following advantages:

[0053] (1) The present application can flexibly select different game stimulation modes, which are applicable to scenarios of multi-group brain-computer interaction and cooperative control. This group-based rehabilitation training method can stimulate the enthusiasm of patients and accelerate the rehabilitation and reconstruction process of motor functions during rehabilitation, and has a better training effect than single-body BCI.

[0054] (2) The present application uses collaborative technology to process multi-dimensional EEG signals, adopts a group interaction framework in the client-server mode, and a communication mechanism in the message queue mode to achieve the cooperative control of the group BCI system, improves the data processing efficiency and real-time performance of the system, and ensures the stability and reliability of the system in complex environments. This enables more efficient group interaction and cooperative control in games, training, or other applications where cooperation or competition occurs among multiple people;

[0055] (3) The present application has the ability to efficiently extract the features of high-dimensional EEG signals, can synchronously decode the EEG signals of group subjects, and the recognition accuracy and stability are the same as those of single-body BCI. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on the structures shown in these drawings.

[0057] Figure 1 It is the system framework diagram of the group brain-computer interface cooperative control system based on SSVEP described in the present invention;

[0058] Figure 2 It is the TCP / IP data protocol format of the present invention;

[0059] Figure 3 It is the UDP message communication process of the present invention;

[0060] Figure 4 It is the EEG signal preprocessing process of the present invention;

[0061] Figure 5 It is the FBCCA EEG signal decoding process of the present invention. Detailed implementation manners

[0062] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without making creative efforts belong to the scope of protection of the present invention.

[0063] It should be noted that all directional indications (such as up, down, left, right, front, back,...) in the embodiments of the present invention are only used to explain the relative positional relationship and movement conditions between components in a specific posture (as shown in the accompanying drawings). If the specific posture changes, the directional indications will also change accordingly.

[0064] In addition, in the present invention, descriptions such as "first" and "second" are only for descriptive purposes, and cannot be understood as indicating or implying their relative importance or implicitly indicating the quantity of the indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, the meanings of "several" and "multiple" are at least two, such as two, three, etc., unless otherwise clearly and specifically defined.

[0065] In the present invention, unless otherwise clearly defined and limited, terms such as "connection" and "fixation" should be understood in a broad sense. For example, "fixation" may be a fixed connection, a detachable connection, or integrated; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and may be the communication inside two elements or the interaction relationship between two elements, unless otherwise clearly limited. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0066] In addition, the technical solutions between various embodiments of the present invention can be combined with each other, but it must be based on the fact that those of ordinary skill in the art can implement it. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present invention.

[0067] The present invention proposes a group brain-computer interface collaborative control system and method based on SSVEP, aiming to...

[0068] Next, the specific structure of the group brain-computer interface collaborative control system based on SSVEP proposed by the present invention will be described in specific embodiments:

[0069] In the technical solution of this embodiment, as Figure 1As shown in the figure, a collaborative control system for a group brain-computer interface based on SSVEP includes:

[0070] A brain-controlled game stimulation module, which is used to display a flashing block with visual stimulation function and to display the scores of competitors;

[0071] A data receiving module, which is used to collect electroencephalogram (EEG) data of multiple EEG acquisition devices through the TCP / IP service protocol;

[0072] Specifically, the data receiving module is TCP / IP communication-connected with multiple EEG acquisition devices, and collects EEG data in real time and sends it to the process control module, and distributes the data to the data parsing module. The TCP / IP service protocol includes 8 EEG data channel bits, 1 flag bit, 1 power bit, and 1 redundant channel bit.

[0073] A process control module, which is used to receive the EEG data of the data receiving module and send the return message to the brain-controlled game stimulation module for interaction;

[0074] Specifically, the process control module calls the data parsing module interface to implement data splitting and encapsulation. Different processes of the process control module receive and process EEG data. The received original EEG signal is preprocessed by the pattern recognition module. The preprocessing process is as Figure 3 shown in the figure, to obtain the filtered EEG signal, and then use it as the input of the FBCCA algorithm for decoding, and distribute the return message to the brain-controlled game stimulation module for interaction.

[0075] A data parsing module, which is used to process the EEG data of the process control module;

[0076] Specifically, it parses the EEG data of multiple devices, organizes and encapsulates them into 8-channel data packets, verifies and removes the frame headers and frame tails of the data frames to ensure the validity of the data. The size of each data frame is 33 bytes, including 11-bit depth information, which is composed of 8 EEG data channel bits, a tag bit, a power bit, and a redundant bit. The 8-channel EEG information in each frame is taken and encapsulated into an EEG signal.

[0077] A pattern recognition module, which is used to perform filtering and singular value decomposition preprocessing on the EEG data processed by the data parsing module, and decode the EEG signal through the FBCCA algorithm, and return the result to the process control module;

[0078] Specifically, the original EEG signal preprocessing uses a Butterworth filter for forward filtering and backward filtering, and performs combined dimensionality reduction processing. The final result is stored in a preprocessing temporary cache variable, as Figure 4As shown in the figure. The pattern recognition module filters and preprocesses the EEG signals cached in the preprocessing link through singular value decomposition, and decodes the EEG signals through the FBCCA algorithm, and returns the results to the process control module. The brain-controlled game stimulation module is responsible for generating visual stimuli and feedback on game operations, such as firing to hit the target and displaying the score.

[0079] The process of decoding EEG signals by the FBCCA algorithm is as follows: First, a multi-band Chebyshev type I filter is created, and the system returns and saves the filter characteristic parameters to variables for frequency feature extraction and noise reduction processing. Then, canonical correlation analysis (CCA) is performed on the EEG template data and the preprocessed EEG signals in each frequency range. After obtaining the CCA index, the frequency range is determined according to weighted processing, and finally the recognition result is obtained and returned, as Figure 5 shown. This process ensures accurate identification and classification of signals in different frequency ranges.

[0080] The message communication module is used to implement the binding of the UDP socket and the host port number, and set the status flag bit of the brain-controlled game stimulation module according to the process status.

[0081] Specifically, the process control module and the brain-controlled game stimulation module are connected through the message communication module, which is implemented through the UDP message communication module. The message communication process is as Figure 3 shown. When the task state is activated, the process enters the flashing stimulation stage, and all game interfaces start to flash the stimulation block. The process control module will start multiple processes to receive data from different EEG acquisition devices. Each frame of EEG data consists of 8 EEG data channel bits, a label bit, a power bit, and 1 redundant channel, as Figure 2 shown. After being encapsulated by the data parsing module, it enters the stack buffer area with a depth of 11 bits.

[0082] It can be understood that this application adopts multi-process, TCP / IP, and UDP communication technologies. Based on a single BCI, through a group interaction framework in the client-server mode, it synchronously decodes and sends the real-time EEG data of multiple clients to the server, and finally realizes the competitive operations of multiplayer cooperative confrontation games on the server. The system can collect and process EEG data from multiple devices in real time in parallel. The accuracy rate of the group EEG decoding algorithm is ≥95%, and the response speed is ≤3s. Through the above system architecture and data flow, it effectively solves the problem that the single BCI system ignores group competition and cooperation behaviors, and provides reliable technical support for application scenarios such as multiplayer cooperative confrontation games.

[0083] Furthermore, the TCP / IP service protocol includes 8 EEG data channel bits, 1 flag bit, 1 power bit, and 1 redundant channel bit.

[0084] Further, the data parsing module includes: a filter unit and an FBCCA recognition algorithm unit.

[0085] Embodiment 2

[0086] A method for controlling a group brain-computer interface based on SSVEP includes the following steps:

[0087] (1) Establish TCP / IP connections with multiple electroencephalogram (EEG) acquisition devices, collect EEG data of the multiple EEG acquisition devices in real time, and send them to the group BCI control system through the TCP / IP communication protocol.

[0088] (2) Select different game modes in the brain control game stimulation module selection interface, determine the stimulation mode, and send start flag bit and end flag bit information to the process control module.

[0089] (3) When the process control module detects that the start flag bit information sent by the brain control game stimulation module is "1" and the end flag bit information is "0", send a START signal, first empty the recognition result and EEG data, and start multiple processes to receive 11-channel EEG data of different EEG acquisition devices.

[0090] (4) Package into EEG signals through the data parsing module.

[0091] (5) Write the packaged original EEG signals into a stack buffer with a depth of 11 bits.

[0092] (6) Send the original data in the stack buffer to a filter for preprocessing of EEG signals.

[0093] (7) Through the FBCCA recognition algorithm, decode the preprocessed data to obtain the recognition results of different subjects.

[0094] (8) Forward the recognition results to the process control module, and the console prints the recognition results of the EEG signals of different subjects for debugging.

[0095] (9) While in step (7), forward the recognition results to the brain control game stimulation module, calculate scores according to the binary classification accuracy rate, and update the display.

[0096] (10) Set the start flag bit information received in step (3) to "0", and wait again for the flag bit information sent by the brain control game stimulation module in step (3).

[0097] (11) When the brain control game stimulation module sends the end flag bit information "1", the process control module closes the communication interface.

[0098] Further, the step of packaging into EEG signals through the data parsing module includes:

[0099] When the data parsing module receives the START signal from the process control module, it sends a START instruction to the TCP communication module to start receiving the message queues transmitted by different EEG acquisition devices;

[0100] When the content of the message queue is not empty and the current time is in the task state, the data parsing module will verify and remove the frame headers and trailers of the data frames to ensure the validity of the data. The size of each data frame is 33 bytes, including 11-bit depth information, which consists of 8 EEG data channel bits, a tag bit, a power bit, and a redundancy bit. The 8-channel EEG information in each frame is taken and added to the stack buffer variable, and the remaining information is also cached;

[0101] When the time delay of the process control module reaches 3 seconds, it sends an end-of-task command, and the loop to receive EEG signal data frames stops.

[0102] Further, the step of sending the original data in the stack buffer area to the filter for EEG signal preprocessing includes:

[0103] Set the cut-off frequency of the Butterworth filter f0 = 50Hz, f s = 1KHz, and the quality factor Q = 35 to determine the bandwidth of the filter. The formula is:

[0104]

[0105] Calculate the forward transfer coefficient b k The formula is:

[0106]

[0107] Calculate the reverse transfer coefficient a m The formula is:

[0108]

[0109] The size of the input original cache data for each channel is 3000 sampling points, and the input matrix size is 8 * 3000. The data passes through a Butterworth forward filter to obtain an output of the same dimension. The specific formula is as follows:

[0110]

[0111] The output of the Butterworth forward filter is used as the input of the Butterworth reverse filter. The specific formula is as follows:

[0112]

[0113] Merge the results of the forward transfer filter and the reverse transfer filter to make the final output have the same phase and dimension as the original data. The formula is:

[0114] yfiltfilt y = (n) forward y + (n) backward (n)

[0115] Delete the axis with dimension 1 in the output result, reduce the matrix dimension, and add it to the preprocessing temporary cache variable.

[0116] Further, the steps of decoding the preprocessed data through the FBCCA recognition algorithm to obtain the recognition results of different subjects include:

[0117] Create a multi-band Chebyshev type I filter, return the filter coefficients, and perform frequency feature extraction and noise reduction processing on the data;

[0118] Perform canonical correlation analysis on the EEG template data and the preprocessed EEG signals, calculate the vector product of the preprocessed data and the template signal to obtain the CCA input data, and the calculation formula is:

[0119] data svd = data processed × Q template

[0120] Output as the input of singular value decomposition, output the left singular matrix, singular value matrix and right singular matrix. Based on the singular value decomposition, select the three largest singular values and calculate the CCA index ρ, and the formula is:

[0121] ρ = 1.25 × S0 + 0.67 × S1 + 0.5 × S2

[0122] Through the feedback weight Perform weighted processing on the index ρ to adjust the weighting degree in the i-th frequency range, and the formula is:

[0123]

[0124] Determine the frequency range according to the index P, obtain a recognition result and return it.

[0125] As mentioned above, only the preferred specific embodiments of the present invention are described, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims

1. A group brain-computer interface collaborative control system based on SSVEP, characterized in that Including: A brain-controlled game stimulation module, which is used to display a flashing block with visual stimulation function and to display the scores of competitors; A data receiving module, which is used to collect electroencephalogram (EEG) data of multiple EEG acquisition devices through the TCP / IP service protocol; A process control module, which is used to receive the EEG data of the data receiving module and send a return message to the brain-controlled game stimulation module for interaction; A data parsing module, which is used to process the EEG data of the process control module; A pattern recognition module, which is used to perform filtering and singular value decomposition preprocessing on the EEG data processed by the data parsing module, decode the EEG signal through the FBCCA algorithm, and return the result to the process control module; A message communication module, which is used to implement the binding of the UDP socket and the host port number, and set the status flag bit of the brain-controlled game stimulation module according to the process status; The process control module calls the data parsing module interface to implement data splitting and encapsulation; different processes of the process control module receive and process the EEG data, and the received original EEG signal is preprocessed by the pattern recognition module; the preprocessing includes obtaining the filtered EEG signal, and then using it as the input of the FBCCA algorithm for decoding, and distributing the return message to the brain-controlled game stimulation module for interaction; The data parsing module will parse the EEG data of multiple devices, organize and encapsulate them into 8-channel data packets, verify and remove the frame headers and frame tails of the data frames to ensure data validity. The size of each data frame is 33 bytes, including 11-bit depth information, which consists of 8-bit EEG data, a tag bit, a power bit, and a redundant bit; The process of decoding the EEG signal by the FBCCA algorithm is as follows: First, create a multi-band Chebyshev type I filter, and the system returns and saves the filter characteristic parameters to variables for frequency feature extraction and noise reduction processing. Then, perform canonical correlation analysis on the EEG template data and the preprocessed EEG signal in each frequency range. After obtaining the CCA index, determine the frequency range according to weighted processing, and finally obtain the recognition result and return it.

2. The SSVEP-based group brain-computer interface collaborative control system according to claim 1, wherein The data parsing module includes: a filter unit and an FBCCA recognition algorithm unit.

3. A group brain-computer interface control method based on SSVEP, characterized in that Including the following steps: (1) Establish a TCP / IP connection with multiple EEG acquisition devices, collect the EEG data of multiple EEG acquisition devices in real time, and send it to the group BCI control system through the TCP / IP communication protocol; (2) Select different game modes in the brain-controlled game stimulation module selection interface, determine the stimulation mode, and send start flag bit and end flag bit information to the process control module; (3) When the process control module detects that the start flag bit information sent by the brain-controlled game stimulation module is "1" and the end flag bit information is "0", send a START signal, first empty the recognition result and EEG data, and start multiple processes to receive 11-channel EEG data of different EEG acquisition devices; (4) Encapsulate into an EEG signal through the data parsing module; (5) Write the encapsulated original EEG signal into a stack buffer area with a depth of 11 bits; (6) Send the original data in the stack buffer area to the filter for EEG signal preprocessing; (7) The preprocessed data is decoded through the FBCCA recognition algorithm to obtain the recognition results of different subjects; (8) The recognition results are forwarded to the process control module, and the console prints the EEG signal recognition results of different subjects for debugging; (9) At the same time as step (7), the recognition results are forwarded to the brain-controlled game stimulation module, and scores are calculated based on the binary classification accuracy and the display is updated; (10) Set the start flag bit information received in step (3) to "0", and wait again for the brain-controlled game stimulation module in step (3) to send the flag bit information; (11) When the brain-controlled game stimulation module sends the end flag bit information "1", the process control module closes the communication interface.

4. The group brain-computer interface collaborative control method based on SSVEP according to claim 3, characterized in that The steps of encapsulating the raw data into EEG signals by the data parsing module include: When the data parsing module receives the START signal from the process control module, it sends a START instruction to the TCP communication module to start receiving the message queues transmitted by different EEG acquisition devices; When the message queue content is not empty and the current moment is in the task state, the data parsing module will verify and remove the frame headers and trailers of the data frames to ensure the data is valid. Each data frame is 33 bytes in size and contains 11-bit depth information, which consists of 8-bit EEG data, a tag bit, a power bit, and a redundant bit. The 8-channel EEG information in each frame is added to the stack buffer variable, and the rest of the information is also cached; When the time delay of the process control module reaches 3 seconds, it sends an end command for the task state, and the loop to receive EEG signal data frames stops.

5. The method for collaborative control of a group brain-computer interface based on SSVEP according to claim 3, wherein The steps of sending the raw data in the stack buffer area to the filter for EEG signal preprocessing include: Set the cut-off frequency of the Butterworth filter \(f_0 = 50Hz\), \(f\) s \(= 1KHz\), quality factor \(Q = 35\), determine the bandwidth of the filter, the formula is: Calculate the forward propagation coefficient b k , the formula is: Calculate the reverse transmission coefficient a m , the formula is: The size of the input raw buffer data for each channel is 3000 sampling points, and the input matrix size is 8 * 3000. The data passes through a Butterworth forward filter to obtain an output of the same dimension. The specific formula is as follows: The output of the Butterworth forward filter is used as the input of the Butterworth reverse filter. The specific formula is as follows: The results of the forward transfer filter and the reverse transfer filter are combined so that the final output is consistent in phase and dimension with the original data. The formula is: y filtfilt y(n) = y forward y(n) + y backward y(n) Delete the axis with dimension 1 in the output result to reduce the matrix dimension and add it to the preprocessing temporary buffer variable.

6. The group brain-computer interface collaborative control method based on SSVEP according to claim 3, wherein The steps of decoding the preprocessed data through the FBCCA recognition algorithm to obtain the recognition results of different subjects include: Create a multi-band Chebyshev type I filter and return the filter coefficients to perform frequency feature extraction and noise reduction processing on the data; Perform canonical correlation analysis on the EEG template data and the preprocessed EEG signals. Calculate the vector product of the preprocessed data and the template signal to obtain the CCA input data. The calculation formula is: data svd = data processed ×Q template The output is used as the input of the singular value decomposition, and the left singular matrix, the singular value matrix, and the right singular matrix are output. Based on the singular value decomposition, select the three largest singular values and calculate the CCA index ρ. The formula is: ρ = 1.25×S0 + 0.67×S1 + 0.5×S2 By feedback weights The index ρ is weighted, and the weighting degree in the i-th frequency range is adjusted. The formula is as follows: Determine the frequency range according to the index P, obtain a recognition result and return it.

Images