Self-adaptive BCI interactive robot system and method for children with hyperactivity

Abstract

The invention relates to the technical field of interactive robots, in particular to a self-adaptive BCI interactive robot system for children with hyperactivity, which comprises an LD3322 voice recognition module, a TGAM brain wave sensor module and a main control board, voice features in the voice electric signals are extracted; the TGAM brain wave sensor module adopts a brain-computer interface technology to collect brain electrical signals; the main control panel comprises a main chip STM32 and an electroencephalogram signal processing chip, and the main chip STM32 receives voice features, presets a keyword library, calls out a corresponding control instruction A in the keyword library and outputs the control instruction A to the steering engine control panel; the electroencephalogram signal processing chip receives the electroencephalogram signal, calculates the concentration degree, sets an index threshold value, calls out a voice prompt signal and a control instruction B corresponding to the duration of the concentration degree in a preset voice library if the concentration degree is larger than or equal to the index threshold value, plays the voice signal through a loudspeaker, and outputs the control instruction B to the steering engine control panel.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of interactive robots, in particular to an adaptive BCI interactive robot system for children with hyperactivity disorder and a method thereof. BACKGROUND

[0002] Attention Deficit Hyperactivity Disorder (ADHD) is a common psychological and behavioral problem in childhood, with core features of inattention, distractibility, hyperactivity, and impulsivity, which can significantly affect children's brain cognition and executive function. Current interventions for ADHD children mainly rely on daily cognitive training and home collaborative intervention, which has obvious shortcomings. The application of brain-computer interface (BCI) technology in concentration training is limited in scope, mostly confined to auxiliary treatment in brain hospitals or professional institutions, with low penetration rate in family settings. Traditional training methods mainly focus on static tasks, lacking game design, and children's willingness to participate is low, limiting the training effect. Data shows that the traditional static training has an improvement rate of about 10% per time, and about 30% after 5 times, with a large room for improvement. At the same time, existing solutions cannot dynamically adjust intervention strategies according to real-time attention states of children, making it impossible to achieve precise and adaptive training. In addition, national policies emphasize the prevention and intervention of children's psychological and behavioral problems, including the identification and early intervention of ADHD and other neurodevelopmental disorders. As a cutting-edge technology, BCI has a wide range of applications, and policy support continues to increase. The academic and clinical supervision in this field is progressing rapidly, with some areas reaching the global top level. This provides a good policy environment and technical foundation for the research and application of related technologies. In view of this, we propose an adaptive BCI interactive robot system for children with hyperactivity disorder and a method thereof. SUMMARY

[0003] The present application aims to solve the problem that current interventions for ADHD children rely on daily cognitive training and home collaborative intervention, which has obvious shortcomings: BCI technology has a narrow application in concentration training, mostly confined to professional institutions, with low penetration rate in family settings; traditional training mainly focuses on static tasks, lacking game design, and children's willingness to participate is low, with an improvement rate of about 10% per time and about 30% after 5 times, with room for improvement; and it is difficult to dynamically adjust strategies according to real-time states, making it impossible to achieve precise and adaptive training.

[0004] To achieve the above-mentioned purpose, the present application provides an adaptive BCI interactive robot system for children with hyperactivity disorder, which comprises an LD3322 voice recognition module, a TGAM brain wave sensor module, and a main control board, wherein: The LD3322 voice recognition module adopts a microphone to collect voice electrical signals in the external environment and extracts voice features in the voice electrical signals; the TGAM brain wave sensor module adopts brain-computer interface technology to collect brain electrical signals; the main control panel includes a main chip STM32 and a brain electrical signal processing chip, the main chip STM32 receives voice features, a preset keyword library, calls out corresponding control instructions A in the keyword library, and then outputs to a rudder control panel to drive a robot; the brain electrical signal processing chip receives brain electrical signals to calculate concentration, sets an index threshold, if the concentration is greater than or equal to the index threshold, calls out a voice prompt signal corresponding to the concentration duration in a preset voice library and a control instruction B, plays the voice signal through a loudspeaker, and outputs the control instruction B to the rudder control panel.

[0005] As a further improvement of the technical solution, the microphone in the LD3322 voice recognition module can vibrate the diaphragm and the fixed electrode to form a capacitor, meet the capacitance formula, transmit the external voice sound wave to the diaphragm, change the sound wave air pressure to push the diaphragm to vibrate, periodically change the distance between the diaphragm and the fixed electrode, and then make the capacitance change reversely and nonlinearly, connect the microphone to a direct current bias voltage and a high resistance load resistor to form a closed loop, cause charging / discharging current due to the capacitance change, and finally generate voice electrical signals on the high resistance load resistor.

[0006] As a further improvement of the technical solution, the voice recognition module LD3322 voice recognition module performs analog conversion on the input voice electrical signals to obtain time domain discrete signals, converts the time domain discrete signals into frequency domain discrete signals to complete time-frequency orthogonal decomposition; then, high-pass filter compensation is performed on the time domain discrete signals to compensate for high frequency attenuation, the signals are divided into short frames and a Hamming window is applied to reduce spectral leakage; FFT transformation is performed on the windowed frame signals and power spectrum is calculated, the power spectrum is distributed through a center frequency according to a mel frequency scale, a mel filter bank is applied, logarithmic transformation is performed on the output of each filter, DCT transformation is performed to extract low frequency coefficients, MFCC coefficients are obtained, and a plurality of MFCC coefficients constitute an MFCC feature vector corresponding to the voice of the external environment, i.e. voice features.

[0007] The above-mentioned further scheme has the beneficial effects that analog conversion and time-frequency decomposition ensure the accuracy of basic signal processing, high-pass filter compensation makes the signal more complete, frame division and windowing combined with the Hamming window effectively reduce spectral leakage, making subsequent frequency analysis more reliable, FFT transformation and power spectrum calculation provide frequency domain basis for feature extraction, the mel filter bank conforms to human ear hearing characteristics, logarithmic transformation and DCT transformation highlight key low frequency features, and the finally obtained MFCC feature vector can accurately depict the essence of the voice, improve the accuracy and robustness of voice recognition, and make the system more stable and accurate in understanding voice content in scenes such as recognizing voice instructions of children with hyperactivity disorder, and lay a high-quality feature foundation for the interactive robot to accurately execute instructions.

[0008] On the basis of the above technical scheme, the application can also be improved as follows: the brain-computer interface in the TGAM brain wave sensor module comprises a plurality of electrodes, and differential acquisition is used to reduce common mode interference during detection: first, the difference between the target electrode and the reference electrode electric signal is calculated, and then the difference signal is amplified, and the amplified signal is the collected brain electric signal.

[0009] The beneficial effect of the above further scheme is that through the cooperation of multiple electrodes and the differential acquisition mode, common mode interference (such as power frequency 50Hz and electromyographic noise) can be targetedly reduced, and through the process of first calculating the difference between the target electrode and the reference electrode electric signal and then amplifying the difference, the useless signals caused by environmental noise and body interference can be effectively eliminated, and purer brain electric signals are retained, providing a high-quality data basis for subsequent concentration detection based on brain electric signals, state analysis of hyperactive children, etc., making the brain state reflected by the brain electric signals more real and accurate, improving the reliability of the entire system (such as an interactive robot system combined with brain electric signals) for judging the state of children, and assisting precise intervention and training.

[0010] On the basis of the above technical scheme, the application can also be improved as follows: the main chip STM32 receives voice features, a preset keyword library, extracts a control instruction A corresponding to the input voice features, cross-calculates the cumulative distances of a plurality of voice features in the keyword library and the input voice features, calls out a control instruction corresponding to the input voice features with the smallest cumulative distance in the keyword library, and finally outputs the control instruction to the steering engine control board.

[0011] As a further improvement of the technical scheme, the main chip STM32 presets a standard feature template of each control instruction in the keyword library: the keyword library comprises two core associations: standard feature template ↔ control instruction, the standard feature template is a standard MFCC feature vector, the LD3322 voice recognition module is used to collect the voice output by the child multiple times, extract the MFCC feature vector in the voice, and each MFCC feature vector corresponds to a known control instruction; and the control instruction is an action code of the robot.

[0012] The beneficial effect of the above further scheme is that the keyword library of the main chip STM32 uses the MFCC feature vectors extracted by collecting the voice of the child multiple times as the standard feature template, and each standard feature template corresponds to a known control instruction (robot action code), and the template is constructed based on the actual voice of the child, which can make the standard feature template more suitable for the voice characteristics of the child, improve the accuracy of voice feature matching, make the control instruction recognition more accurate, when the voice of the child is input, the corresponding control instruction is called out through the cumulative distance calculation, the robot can perform actions more accurately, the adaptability and fluency of human-machine voice interaction are enhanced, and especially in the interactive scene for hyperactive children, the voice intention of the child can be better understood, and the intervention and training based on voice control can be effectively carried out.

[0013] On the basis of the above technical scheme, the EEG signal processing chip can also be improved as follows: the EEG signal processing chip receives and samples the EEG signal according to a sampling frequency, constructs a transfer function to remove 50Hz power frequency interference, and obtains a filtered signal through recursive calculation; the frequency spectrum of the filtered signal is obtained through FFT, a frequency band associated with relaxation degree is defined, the corresponding energy is calculated, and the mapping relationship is established through clinical calibration to calculate the concentration degree.

[0014] The above further scheme has the beneficial effects that: the EEG signal processing chip collects and samples the EEG signal according to a sampling frequency, constructs a transfer function to remove 50Hz power frequency interference (which is easy to mix into the EEG signal and affect subsequent analysis), and obtains a more pure filtered signal through recursive calculation, then obtains the frequency spectrum through FFT, defines a frequency band associated with relaxation degree, calculates the energy, and establishes a mapping relationship through clinical calibration to calculate the concentration degree; noise can be effectively removed, and the characteristics related to the concentration degree in the EEG signal can be accurately extracted, providing accurate and quantitative concentration data for the intervention of children with hyperactivity disorder, making the training based on EEG feedback (such as robot interaction and motivation) more targeted, helping to improve the training effect, while ensuring the scientificity and reliability of the detection, and supporting real-time and dynamic monitoring of the concentration state of children.

[0015] On the basis of the above technical scheme, the EEG signal processing chip can also be improved as follows: the EEG signal processing chip can also be improved as follows: the EEG signal processing chip receives and samples the EEG signal according to a sampling frequency, constructs a transfer function to remove 50Hz power frequency interference, and obtains a filtered signal through recursive calculation; the frequency spectrum of the filtered signal is obtained through FFT, a frequency band associated with relaxation degree is defined, the corresponding energy is calculated, and the mapping relationship is established through clinical calibration to calculate the concentration degree.

[0016] As a further improvement of the technical scheme, the rudder control board is specifically an STM32103 rudder control board, which controls 17 rudders to realize walking and dancing of the robot.

[0017] A self-adaptive BCI interactive robot method for children with hyperactivity disorder, comprising the following steps: Step one: using a microphone to collect voice electrical signals in the external environment, and extracting voice features in the voice electrical signals; Step two: collecting EEG signals using brain-computer interface technology; Step three: presetting a keyword library, calling corresponding control instruction A in the keyword library, and then outputting to a rudder control board to drive a robot; Step four: calculating the concentration degree, setting an index threshold, judging whether the concentration degree meets the standard, calling the corresponding voice prompt signal and control instruction B in the preset voice library when the concentration degree meets the standard, playing the voice signal using a loudspeaker, and outputting the control instruction B to the rudder control board.

[0018] In addition to the purposes, features and advantages described above, the present application has other purposes, features and advantages. The present application will be described in further detail below with reference to the drawings. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 The overall module schematic diagram of the present application; Figure 2 The overall module schematic diagram of the present application; Figure 3 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 4 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 5 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figures 6-8 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 9 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 10 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 11 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 12 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 13 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; Figure 14 The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application;

[0020] The brain-computer interface hardware schematic diagram in the TGAM brain wave sensor module of the present application; 100, LD3322 speech recognition module; 200, TGAM brain wave sensor module; 300, main control board; 310, main chip STM32310; 320, electroencephalogram signal processing chip 320. DETAILED DESCRIPTION

[0021] The technical solutions in the embodiments of the present application will be described in detail below with reference to the drawings of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0022] REFERENCE Figures 1-14 The adaptive BCI interactive robot system for children with hyperactivity disorder includes an LD3322 speech recognition module 100, a TGAM brain wave sensor module 200 and a main control board 300, wherein: The LD3322 voice recognition module 100 adopts a microphone to collect voice electric signals in the external environment The microphone includes a vibrating diaphragm and a fixed electrode to form a capacitor, satisfying the basic formula of capacitance: Wherein C is the capacitance value, ε is the dielectric constant of the medium between the plates, A is the overlapping area of the diaphragm and the fixed electrode, D is the distance between the diaphragm and the fixed electrode, and the dynamic change amount ; When the sound wave generated by the voice in the external environment is transmitted to the diaphragm, the pressure change of the sound wave pushes the diaphragm to vibrate, periodically changing the distance between the diaphragm and the fixed electrode , resulting in a reverse nonlinear change of the capacitance ; ; The microphone is connected to a direct current bias voltage and a high resistance load resistor , forming a closed loop. The capacitance change will cause charging / discharging current, and finally generate a voice electric signal on the high resistance load resistor : Using the mature principle of capacitance change, the change of sound wave pressure can be effectively converted into the change of electric signal, realizing stable and sensitive collection of voice signals. Based on the clear physical mechanism, the accuracy and reliability of voice electric signal collection are guaranteed.

[0023] The voice recognition module LD3322 voice recognition module 100 performs analog-to-digital conversion on the voice electric signal , and obtains a time domain discrete signal , wherein fs is the sampling rate, , is the number of sampling points; Then, the time domain discrete signal is converted into a frequency domain discrete signal , wherein , is the imaginary unit; is the complex exponential base function, realizing time domain→frequency domain orthogonal decomposition; The time domain discrete signal is high-pass filtered to compensate for high-frequency attenuation, and a pre-emphasized time domain discrete signal is obtained; The time domain discrete signal is cut into short frames, and a Hamming window is applied to the time domain discrete signal to reduce spectral leakage, and a windowed frame signal is obtained, wherein​ For window functions; Perform an FFT transform on the windowed frame signal to obtain its frequency domain representation, and then calculate the power spectrum. ,in This is the result after FFT transformation. Frame length; power spectrum A set of Mel filters is used, with the center frequencies of the Mel filters distributed according to the Mel frequency scale; the first... The output of the Mel filter ,in The number of points after FFT transformation. For the first Frequency response of a Mel filter; Taking the logarithm of the output of each Mel filter, we get The logarithmic result is then subjected to a DCT transform to extract low-frequency coefficients, yielding the MFCC coefficients. The discrete cosine transform formula is: ,in ,in The number of Mel filters. To extract the number of MFCC coefficients, The MFCC coefficients are used to construct the MFCC feature vector corresponding to the external environmental speech, which is the speech feature. .

[0024] The LD3322 voice recognition module 100 uses the UNV6288 chip and outputs voice features via serial port. To the main control board 300.

[0025] like Figure 3 As shown, the TGAM EEG sensor module 200 uses brain-computer interface technology to collect EEG signals. The specific working principle is as follows: When neurons in the cerebral cortex (approximately 14 billion) transmit information, they generate synchronous discharges. These synchronous discharges create a weak potential difference on the scalp surface. Brain-computer interface technology includes multiple electrodes. When detecting EEG signals, differential acquisition is used to reduce common-mode interference (such as 50Hz power frequency and electromyographic noise). This involves multiple electrodes separately acquiring electrical signals from target electrodes (such as the prefrontal cortex) and reference electrodes (such as the earlobe and occipital bone). and Calculate the electrical signal difference between the two electrical signals: And the difference in electrical signals Enlarge: ,in The amplification factor is the difference in the amplified electrical signal. This refers to electroencephalogram (EEG) signals. ; The brain-computer interface in the TGAM brain wave sensor module 200 adopts a mind technology chip, efficiently processes brain wave data, detects concentration in real time, and transmits brain wave signals to the main control board 300 through NRF24L01 wireless communication. The TGAM brain wave sensor module 200 accurately collects brain wave signals , provides a reliable data source for subsequent analysis based on brain wave signals (such as concentration detection of children with hyperactivity), and effectively reduces interference and enhances weak brain wave signals through differential acquisition and amplification processing. At the same time, the mind technology chip ensures efficient data processing and real-time concentration detection, and wireless communication enables convenient signal transmission, helping to build a complete brain wave signal acquisition-transmission-analysis application system and support subsequent precise judgment of children's brain state (such as concentration) and implementation of intervention strategies.

[0026] As shown in Figure 4 , the main control board 300 includes a main chip STM32310 and a brain wave signal processing chip 320, wherein the main chip STM32310 receives a voice feature , a preset keyword library , is a standard feature template of the th control instruction in the keyword library, is the th feature value in the standard feature template; extract the control instruction A corresponding to the voice feature in the keyword library: cross-compute the cumulative distance between multiple voice features and the voice feature : , wherein is a path weight (0 or 1, ensuring path continuity), , specifically the inter-frame Euclidean distance; call out the control instruction corresponding to the minimum cumulative distance of the voice feature in the keyword library: control instruction ; and output the control instruction corresponding to the minimum cumulative distance to the steering engine control board to drive the robot; The keyword library contains two core associations: standard feature template ↔ control instruction. The standard feature template is a standard MFCC feature vector, and the standard feature template is obtained by collecting the child's output voice multiple times through the LD3322 voice recognition module 100, extracting the MFCC feature vector in the voice, and each MFCC feature vector corresponds to a known control instruction. The control instruction is an action code (such as 0x01 = forward, 0x02 = left turn) executed by the robot. ​The main chip STM32310 accurately recognizes voice commands and converts them into control signals that the robot can execute, enabling voice control of the robot. Based on the mature DTW algorithm and a comprehensive keyword library, it can accurately match voice features with control commands, ensuring the accuracy and reliability of command recognition. Combined with the feature templates collected and constructed by the LD3322 voice recognition module 100, it is adapted to children's voice scenarios, improving the recognition effect of children's voice commands. This allows the robot to accurately execute actions based on voice, enhancing the fluency and intelligence of human-computer interaction, and providing stable and efficient technical support for the voice control link in the adaptive BCI interactive robot system for children with ADHD.

[0027] refer to Figure 9 As shown, the EEG signal processing chip 320 receives EEG signals. According to sampling frequency ( =250Hz) for EEG signals sampling: ,in, For sampling signals, The sampling period; For the sampled signal Construct the transfer function Remove brain signals 50Hz power frequency interference, transfer function middle Normalized angular frequency ( For interference frequency, =250Hz is the sampling rate). The damping coefficient; The sampled signal is calculated recursively. Obtain the filtered signal : ; Filtered signal (length ), calculate the spectrum using FFT , ; Define relaxation correlation Wavelength (8-13Hz), related to attention span (13-30Hz), calculate the corresponding energy: , By establishing a linear mapping relationship through clinical trials, attention levels can be calculated. The EEG signal processing chip 320 can also display focus level via OLED; Set the index threshold, if the concentration ≥ index threshold, then call out the preset voice library concentration duration corresponding voice prompt signal, control instruction B; the voice signal is played by the loudspeaker, and the control instruction B is output to the steering wheel control board to drive the robot, and the steering wheel control board is specifically an STM32103 steering wheel control board, 17 steering wheels are controlled, the robot walks and dances are realized; The voice signal in the electroencephalogram signal processing chip 320 is used to output encouragement through the loudspeaker, such as "children, please concentrate ~" "children, come and catch me!" "The end is approaching, children, please cheer up!" and the like; when the control instruction B is output to the steering wheel control board to drive the robot, the robot outputs dance actions such as shaking left and right, waving hands, and swinging hands. For example, according to the real-time loudspeaker playing of the voice signal and the driving of the robot action according to the attention state of the child: when the concentration lasts for 0-10s, the voice signal outputs "children, please concentrate ~", and the robot outputs shaking left and right; when the concentration lasts for 10-30s, the voice signal outputs "children, come and catch me!", and the robot outputs waving hands; when the concentration lasts for 30-60s, the voice signal outputs "The end is approaching, children, please cheer up!", and the robot outputs swinging hands, so that the concentration time of the child is continuously increased through training.

[0028] The working principle of the loudspeaker in the electroencephalogram signal processing chip 320 is based on the synergistic effect of electromagnetic induction and mechanical vibration sound generation, and the core structure includes a permanent magnet, a voice coil and a diaphragm. When an audio electric signal is input, the voice coil as a current-carrying conductor is placed in the magnetic field generated by the permanent magnet. According to the Ampere force formula, the voice coil will be subjected to periodic changes in Ampere force due to the size and direction of the current, and then drive the diaphragm connected thereto to produce mechanical vibration of the same frequency. The diaphragm vibration compresses and stretches the surrounding air to form alternating sound waves, and finally converts the electric signal into an air transmissible sound signal, realizing the playing of voice and other electric signals; The loudspeaker is used to stably output encouraging voice, enhance interactivity and interest, and the steering wheel is used to control the robot action to enrich the feedback form, so as to help improve the concentration time and participation of children with hyperactivity syndrome in the training process, and promote the effective development of intervention training.

[0029] A self-adaptive BCI interactive robot method for children with hyperactivity syndrome, comprising the following steps: Step one: use a microphone to collect voice electric signals in the external environment, and extract voice features in the voice electric signals; Step two: collect electroencephalogram signals by using brain-computer interface technology; Step three: preset a keyword library, call out the corresponding control instruction A in the keyword library, and then output to the steering wheel control board to drive the robot; Step four: calculate the concentration, set the index threshold, judge whether the concentration meets the standard, and play the corresponding voice prompt signal and control instruction B in the preset voice library when the concentration meets the standard, adopt the loudspeaker to play the voice signal, and output the control instruction B to the steering wheel control board.

[0030] The above shows and describes the basic principles, main features and advantages of the present application. Those skilled in the art should understand that the present application is not limited to the above-mentioned embodiments, and the above-mentioned embodiments and descriptions in the specification are only preferred examples of the present application and are not intended to limit the present application. Without departing from the spirit and scope of the present application, various changes and improvements can be made to the present application, and these changes and improvements all fall within the scope of the claimed present application. The scope of protection of the present application is defined by the appended claims and their equivalents.

Claims

1. An adaptive BCI interactive robot system for children with hyperactivity disorder, characterized in that, It comprises an LD3322 voice recognition module (100), a TGAM brain wave sensor module (200) and a main control board (300), wherein: The LD3322 voice recognition module (100) collects voice electrical signals in the external environment through a microphone and extracts voice features in the voice electrical signals; the TGAM brain wave sensor module (200) collects brain electrical signals through brain-computer interface technology; and the main control board (300) comprises a main chip STM32 (310) and a brain electrical signal processing chip (320), wherein the main chip STM32 (310) receives voice features, presets a keyword library, calls out a corresponding control instruction A in the keyword library, and then outputs to a rudder control board to drive a robot; the brain electrical signal processing chip (320) receives brain electrical signals to calculate concentration, sets an index threshold, and if the concentration is greater than or equal to the index threshold, calls out a voice prompt signal corresponding to the concentration duration in a preset voice library and a control instruction B, plays the voice signal through a loudspeaker, and outputs the control instruction B to the rudder control board.

2. The adaptive BCI interactive robot system for children with hyperactivity disorder according to claim 1, characterized in that: The microphone in the LD3322 voice recognition module (100) can vibrate a diaphragm and a fixed electrode to form a capacitor, satisfy a capacitance formula, and transfer external voice sound waves to the diaphragm, so that the sound wave air pressure changes to push the diaphragm to vibrate, periodically change the distance between the diaphragm and the fixed electrode, and then cause the capacitor to change in a reverse non-linear manner. The microphone is connected to a direct current bias voltage and a high resistance load resistor to form a closed loop, the capacitance change causes charging / discharging current, and finally generates a voice electrical signal on the high resistance load resistor.

3. The self-adapting BCI interactive robot system for children with hyperactivity according to claim 2, wherein: The voice recognition module LD3322 voice recognition module (100) performs analog conversion on the input voice electrical signal, obtains a time domain discrete signal, and then converts it into a frequency domain discrete signal to complete time-frequency orthogonal decomposition; then, the time domain discrete signal is high-pass filtered to compensate for high frequency attenuation, is divided into short frames, and a Hamming window is applied to reduce spectral leakage; the windowed frame signal is subjected to FFT transformation and power spectrum calculation, the power spectrum is distributed through a center frequency according to a mel frequency scale through a mel filter bank, the logarithm of the output of each filter is taken, and DCT transformation is performed to extract low frequency coefficients to obtain MFCC coefficients. A certain number of MFCC coefficients constitute an MFCC-feature vector corresponding to the voice of the external environment, i.e. voice features.

4. The self-adapting BCI interactive robot system for children with hyperactivity disorder according to claim 1, wherein: The brain-computer interface in the TGAM brain wave sensor module (200) comprises a plurality of electrodes, detects, and uses differential acquisition to reduce common mode interference: first, the difference between the target electrode and the reference electrode electrical signal is calculated, and then the difference signal is amplified. The amplified signal is the collected brain electrical signal.

5. The self-adapting BCI interactive robot system for children with hyperactivity disorder according to claim 1, wherein: The main chip STM32 (310) receives voice features, presets a keyword library, extracts a control instruction A corresponding to the input voice features, cross-calculates the cumulative distances of a plurality of voice features in the keyword library and the input voice features, calls out a control instruction corresponding to the minimum cumulative distance of the input voice features in the keyword library, and finally outputs the control instruction to the rudder control board.

6. The self-adapting BCI interactive robot system for children with hyperactivity according to claim 5, wherein: The main chip STM32 (310) includes a standard feature template of each control instruction in the preset keyword library: the keyword library contains two core associations: standard feature template ↔ control instruction, the standard feature template is a standard MFCC feature vector, the standard feature template collects the voice output by the child multiple times through the LD3322 voice recognition module (100), extracts the MFCC feature vector in the voice, and each MFCC feature vector corresponds to a known control instruction; the control instruction is an action code executed by the robot.

7. The adaptive BCI interaction robot system for children with hyperactivity according to claim 4, characterized in that: The EEG signal processing chip (320) receives and samples the EEG signal according to the sampling frequency, constructs a transfer function to remove 50Hz power frequency interference, and obtains a filtered signal through recursive calculation; The FFT is performed on the filtered signal to obtain a frequency spectrum, a frequency band associated with relaxation is defined, the corresponding energy is calculated, and the mapping relationship is established through clinical calibration to calculate the concentration.

8. The self-adapting BCI interactive robot system for children with hyperactivity according to claim 7, wherein: Set an index threshold, if the concentration ≥ the index threshold, the voice prompt signal corresponding to the concentration duration in the preset voice library is called out, the control instruction B is output to the steering wheel control board to drive the robot.

9. The self-adapting BCI interactive robot system for children with hyperactivity according to claim 8, wherein: The steering wheel control board is specifically an STM32103 steering wheel control board, which controls 17 steering wheels to realize the walking and action dance of the robot.

10. An adaptive BCI interactive robot method for hyperactive children, applied to the adaptive BCI interactive robot system for hyperactive children according to any one of claims 1-9, characterized in that, The method comprises the following steps: Step one: use a microphone to collect voice electrical signals in the external environment, and extract voice features in the voice electrical signals; Step two: collect EEG signals using brain-computer interface technology; Step three: preset a keyword library, call out the corresponding control instruction A in the keyword library, and then output to the steering wheel control board to drive the robot; Step four: calculate the concentration, set an index threshold, and determine whether the concentration meets the standard, when the concentration meets the standard, call out the voice prompt signal and control instruction B corresponding to the duration in the preset voice library, use a loudspeaker to play the voice signal, and output the control instruction B to the steering wheel control board.

Images