Neural signal coding generation method based on micro-vibration and wearable device

Abstract

The present invention relates to the technical field of biomedical engineering and human-computer interaction technology, and more specifically, to a neural signal encoding generation method based on micro-vibration and a wearable device. The method comprises the following steps: receiving input information and extracting feature vectors; mapping the feature vectors to a multidimensional vibration parameter space; generating a micro-vibration signal based on the vibration parameter space; adjusting the micro-vibration signal according to a user's personal neural response model; adaptively adjusting the micro-vibration signal based on current context information; collecting user neural response data; calculating the error between the neural response data and a target neural response pattern; optimizing the vibration parameters based on the error; updating the user's personal neural response model; performing multimodal fusion on the optimized vibration parameters to obtain a final micro-vibration output signal. By constructing a multidimensional vibration parameter space and adopting a sequence generation technology driven by deep learning, the information transmission accuracy of the present invention is as high as 95%.

Description

Technical Field

[0001] The present invention relates to the technical field of biomedical engineering and human-computer interaction technology, and more specifically, to a neural signal coding generation method based on micro-vibration and a wearable device. Background Art

[0002] With the rapid development of human-computer interaction technology, information transmission methods based on tactile feedback have attracted widespread attention. Especially in the field of wearable devices, how to transmit complex information, regulate emotional states, and even enhance cognitive abilities through precise tactile stimulation has become a research hotspot. However, existing technologies still face many challenges in this field.

[0003] Traditional vibration feedback devices can usually only provide vibrations of a single frequency and intensity, and their information transmission capabilities are extremely limited. For example, the vibration reminder function of smartphones and smart bracelets can often only convey simple information such as "a message has arrived", and cannot distinguish details such as the type of message, urgency, etc. This crude information encoding method severely limits the information transmission potential of the tactile channel.

[0004] In recent years, some researchers have tried to develop more complex tactile feedback systems. For example, Luo et al. proposed a speech coding model based on neural oscillations in "Phase patterns of neuronal responses reliably discriminate speech in human auditory cortex" (Neuron, 2007, 54 (6): 1001-1010). Although this model has made some progress in speech recognition, it is still limited to the auditory system and has not been extended to the tactile field. In addition, such systems often use preset fixed vibration patterns and lack the ability to adapt to individual differences in users and environmental changes.

[0005] On the other hand, existing tactile feedback devices also have obvious deficiencies in user experience. Long-term use may cause skin discomfort or fatigue, seriously affecting the use effect. More importantly, these devices hardly take into account the potential impact of tactile stimulation on brain neuroplasticity, and fail to fully utilize the tactile channel for emotional regulation.

[0006] In view of these limitations of existing technologies, there is an urgent need for new tactile feedback methods and devices that can achieve high-precision neural signal encoding, adaptive regulation, promote neural plasticity, and effectively regulate emotional states. Summary of the invention

[0007] The present invention is proposed to solve the above technical problems. The present invention provides a method for generating neural signal coding based on micro-vibration, which realizes a highly intelligent and personalized micro-vibration neural signal coding system through innovative technologies such as multi-dimensional vibration parameter space, personalized neural response model, context-aware adaptation, and deep reinforcement learning optimization.

[0008] The present invention provides a neural signal encoding generation method based on micro-vibration, comprising the following steps: receiving input information and extracting feature vectors; mapping the feature vectors to a multidimensional vibration parameter space; generating a micro-vibration signal based on the vibration parameter space; adjusting the micro-vibration signal according to a user's personal neural response model; adaptively adjusting the micro-vibration signal based on current context information; collecting user neural response data; calculating the error between the neural response data and a target neural response pattern; optimizing the vibration parameters based on the error; updating the user's personal neural response model; and performing multimodal fusion on the optimized vibration parameters to obtain a final micro-vibration output signal.

[0009] Specifically, the multidimensional vibration parameter space includes five dimensions: amplitude, frequency, duration, spatial distribution and rhythmic pattern, wherein: the corresponding vibration parameter vector is obtained by inputting the feature vector into a pre-trained deep neural network; the amplitude range is 0-500μm, the frequency range is 1-500Hz, and the duration range is 10ms-1s; the spatial distribution is represented by a 4x4 matrix, and each element corresponds to the activation intensity of a vibration unit; the rhythmic pattern is represented by the spectral features obtained by Fourier transform.

[0010] Specifically, the steps of constructing the personal neural response model include: collecting the user's response data to a standard micro-vibration sequence; using a support vector regression algorithm, with the standard micro-vibration sequence as input and the response data as output, to train an initial model; and continuously updating the model parameters through an online learning algorithm based on the user's feedback data during daily use.

[0011] Specifically, the adaptive adjustment step includes: collecting environmental vibration and noise data through an acceleration sensor and a microphone; analyzing the spectral characteristics of the environmental data using a fast Fourier transform algorithm; when it is detected that the environmental interference in a specific frequency band exceeds a preset threshold, adjusting the frequency of the micro-vibration signal to avoid the interference frequency band; and dynamically adjusting the intensity of the micro-vibration signal according to the user's activity status, wherein the activity status is determined by the root mean square value of the acceleration sensor.

[0012] Specifically, the neuroplasticity enhancement training steps include: setting training goals, such as improving attention or working memory; designing a micro-vibration sequence with progressive difficulty based on the training goals; monitoring the user's neural activity through electroencephalogram (EEG) or functional near-infrared spectroscopy (FNIRS) in each training stage; calculating a similarity score with the target state based on the neural activity data; and adaptively adjusting the micro-vibration sequence parameters for the next stage based on the similarity score.

[0013] Specifically, the emotional state adjustment step includes: evaluating the user's emotional state in real time through heart rate variability, skin conductance and facial electromyography signals; mapping the emotional state to a two-dimensional emotional space, including two dimensions of pleasure and arousal; selecting a preset micro-vibration mode according to the difference between the target emotional state and the current state; guiding the user's emotional state to transition to the target state by gradually adjusting the micro-vibration parameters; continuously monitoring changes in emotional indicators and dynamically optimizing the adjustment strategy.

[0014] Specifically, the safety control steps include: real-time monitoring of the user's physiological indicators, including heart rate, blood pressure and skin temperature; automatically reducing the micro-vibration intensity or pausing the output when the physiological indicators exceed the safety threshold; accumulating and recording the daily micro-vibration usage time, and when it exceeds the recommended usage time, issuing a reminder and limiting the usage intensity; setting an emergency stop mechanism to allow the user to immediately terminate all micro-vibration outputs through specific operations.

[0015] The wearable device of the method described herein comprises: a multimodal micro-vibration generator array, consisting of 16 piezoelectric ceramic drivers arranged in a 4x4 matrix; a high-precision sensor module, including a 3-axis accelerometer, a 3-axis gyroscope, a temperature sensor, a photoelectric volume pulse wave sensor, and a skin conductance sensor; a microprocessor unit, including a main processor and a neural network coprocessor; a storage unit, including 256MB LPDDR4RAM and 64GB eMMC flash memory; a wireless communication module, supporting Bluetooth 5.2, Wi-Fi 6, and NFC; a power management unit, including a 500mAh lithium polymer battery and an intelligent power management chip.

[0016] Specifically, the specific parameters of the multimodal micro-vibration generator array are: each driver can generate vibrations in three directions of X, Y, and Z, and can rotate; the amplitude range is 0-500μm, and the accuracy is 0.1μm; the frequency range is 1-500Hz, and the accuracy is 0.1Hz; through PWM signal control, 256 levels of amplitude adjustment and 1024 levels of frequency adjustment can be achieved.

[0017] Specifically, the specific parameters of the high-precision sensor module are: 3-axis accelerometer, range of ±16g, sampling rate of 1kHz, 16-bit resolution; 3-axis gyroscope, range of ±2000° / s, sampling rate of 1kHz, 16-bit resolution; temperature sensor, accuracy of ±0.1℃, sampling rate of 10Hz; photoelectric volumetric pulse wave sensor, sampling rate of 200Hz, 12-bit resolution; skin conductance sensor, range of 0-100μS, sampling rate of 100Hz, 14-bit resolution.

[0018] The beneficial effects of the present invention are mainly reflected in the following aspects:

[0019] First, by constructing a multi-dimensional vibration parameter space and adopting deep learning-driven sequence generation technology, the present invention significantly improves the capacity and accuracy of information encoding. Experimental results show that the information transmission accuracy of the present invention is as high as 95%, far exceeding the existing technology. This means that users can receive richer and more complex information through the tactile channel, greatly expanding the possibility of human-computer interaction.

[0020] Secondly, the personalized neural response model and context-aware adaptation mechanism of the present invention greatly improve the user experience. The system can dynamically adjust the micro-vibration parameters according to each user's unique response characteristics and the current environment, ensuring comfort for long-term use. In the test, the user comfort score reached 6.2 (out of 7 points), which is significantly higher than other solutions.

[0021] Furthermore, this invention integrates the neuroplasticity enhancement training function into the tactile feedback device for the first time. By designing targeted micro-vibration sequences and dynamically adjusting according to the user's real-time EEG response, this system successfully promotes the plasticity of the brain and improves the user's cognitive ability. Experiments show that the user's working memory ability has increased by an average of 18%. This breakthrough has opened up a new research direction for cognitive enhancement technology.

[0022] Finally, the present invention has also achieved remarkable results in emotion regulation. The emotion state assessment system based on multi-dimensional physiological indicators, combined with the carefully designed emotion regulation micro-vibration mode, can effectively guide the user's emotions to the target state. The test results show that the emotion regulation effect reaches 85%, far exceeding other solutions.

[0023] In summary, the present invention not only solves the deficiencies of existing tactile feedback technology in information encoding and user experience, but also innovatively integrates the functions of neuroplasticity enhancement and emotion regulation into a unified system. This comprehensive performance advantage makes the present invention have broad application prospects in the fields of wearable devices, human-computer interaction, health management, etc., and is expected to lead a new round of innovation in tactile feedback technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a logic block diagram of the overall core steps of the present invention.

[0025] Figure 2 A logical block diagram optimized for the personal neural response model of the present invention.

[0026] Figure 3 4 is a logic block diagram of the system architecture of the wearable device of the present invention. DETAILED DESCRIPTION

[0027] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation methods, structures, features and effects thereof are described in detail below in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0028] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0029] The present invention relates to a method for generating neural signal coding based on micro-vibration and a wearable device thereof, and belongs to the field of biomedical engineering and human-computer interaction technology. The method generates specific tactile stimulation on the user's skin surface through a precisely controlled micro-vibration sequence, thereby inducing a predefined neural signal pattern in the brain. This innovative method can not only efficiently transmit information, but also accurately adjust the user's physiological and psychological state, and has broad application prospects.

[0030] Please refer to Figure 1-3 In order to achieve the above object, the present invention provides a method for generating neural signal encoding based on micro-vibration, comprising the following steps:

[0031] Step S1: Receive input information and extract feature vectors.

[0032] First, the present invention processes the input information I(t) through a multimodal feature extraction network F to obtain a feature vector set [f1, f2, ..., f n ].

[0033] Preferably, a convolutional neural network is used as the feature extractor, which includes 5 convolutional layers and 3 fully connected layers. Then, the extracted features are normalized using the formula f i ′=(f i -μ i ) / σ i , where μ i and σ iare the mean and standard deviation of the ith feature, respectively. This can eliminate the dimensional differences between different features and improve the stability of subsequent processing.

[0034] Step S2: Map the feature vector to the multidimensional vibration parameter space.

[0035] In this step, the present invention constructs a multidimensional vibration parameter space V, including five dimensions: amplitude (A), frequency (F), duration (D), spatial distribution (S) and rhythmic pattern (R). Preferably, the amplitude range is 0-500μm, the frequency range is 1-500Hz, and the duration range is 10ms-1s. The spatial distribution is represented by a 4×4 matrix, and each element corresponds to the activation intensity of a vibration unit. The rhythmic pattern is represented by the spectral features obtained by Fourier transform. Next, the present invention designs a sequence generation network G to map the standardized features to the vibration parameter space: G(f1′, f2′,…, f n ′)=[v1,v2,…,v m ], where v i ∈V represents a multidimensional vibration parameter vector. The network adopts LSTM (Long Short-Term Memory) structure to capture the temporal dependencies between features.

[0036] Step S3: Generate a micro-vibration signal based on the vibration parameter space.

[0037] In order to convert the parameter vector sequence into a continuous micro-vibration signal, the present invention applies a time encoding function T:T([v1,v2,…,v m ])=s(t), where s(t) represents the micro-vibration signal that changes with time. In the specific implementation, an interpolation algorithm is used to generate a smooth transition between adjacent parameter vectors to ensure the continuity of the signal.

[0038] Step S4: Adjust the micro-vibration signal according to the user's personal neural response model. In this step, the user's personal neural response model M(u) is first constructed: M(u) = {θ1, θ2, ..., θ k}, where θ i Represents the parameters of the model, and u represents a specific user. In order to build this model, we collect the user's response data to the standard micro-vibration sequence and use the support vector regression algorithm to train the initial model. Preferably, a Gaussian kernel function is used, the kernel parameter γ is set to 0.1, and the regularization parameter C is set to 5. Then, the personalized adjustment function P:s′(t)=P(s(t),M(u)) is applied to adjust the micro-vibration signal. This function is actually an adaptive filter whose parameters are determined by the neural response model M(u).

[0039] Step S5: Adaptively adjust the micro-vibration signal based on the current context information.

[0040] First, the environmental vibration and noise data are collected through the acceleration sensor and microphone. Then, the spectrum characteristics of the environmental data are analyzed using the fast Fourier transform algorithm. When the environmental interference in a specific frequency band (such as 50-200Hz) is detected to exceed the preset threshold (such as -40dB), the frequency of the micro-vibration signal is adjusted to avoid the interference frequency band.

[0041] At the same time, the intensity of the micro-vibration signal is dynamically adjusted according to the user's activity status. The activity status is determined by the root mean square value of the acceleration sensor, and the threshold is set to 0.5g. When the activity intensity exceeds the threshold, the micro-vibration intensity is increased by 20% to ensure the perceptibility of the signal.

[0042] Step S6: Collect user neural response data.

[0043] The present invention uses a high-precision sensor module to collect the user's neural response data N(t) in real time. Specifically, it includes 1. measuring heart rate variability through a photoelectric volumetric pulse wave sensor with a sampling rate of 200Hz; 2. using a skin conductance sensor to monitor emotional changes with a range of 0-100μS and a sampling rate of 100Hz; 3. using a forehead patch electroencephalogram (EEG) electrode to collect the frontal lobe theta wave (4-8Hz) and beta wave (13-30Hz) activities with a sampling rate of 250Hz.

[0044] Step S7: Calculate the error between the neural response data and the target neural response pattern. This step calculates the error E(t) between the actual neural response and the target response: E(t) = ∥ N(t) - N * (t)∥, where N * (t) represents the expected neural response pattern. Preferably, the root mean square error (RMSE) is used as the error metric, and the error threshold is set to 0.2, and parameter optimization is triggered when the error exceeds the threshold.

[0045] Step S8: Optimize vibration parameters based on errors. The present invention uses a deep reinforcement learning algorithm RL to optimize vibration parameters: v i ′=RL(v i , E(t)). Specifically, the Double Q-learning algorithm is used, the state space is the current vibration parameters and neural response errors, and the action space is the parameter adjustment amount. The reward function is designed to be the negative value of the error to encourage the system to minimize the error. The learning rate is set to 0.01 and the discount factor is 0.95.

[0046] Step S9: Update the user's personal neural response model.

[0047] Based on the newly collected data, the present invention continuously updates the user's personal neural response model: M′(u)=Update(M(u),N(t),s″(t)). An online learning algorithm is used, and the most recent 1000 data points are used for each update, and the learning rate is set to 0.001. This ensures that the model can adapt to slow changes in user response characteristics.

[0048] Step S10: Perform multimodal fusion on the optimized vibration parameters to obtain the final micro-vibration output signal.

[0049] First, the optimized vibration parameter v′ i Decompose into different modes: v′ i = {v x ,v y ,v z ,v r}, where v x ,v y ,v z Represent the vibration parameters in three spatial directions, v r Then, the multi-modal vibration synthesis function H is applied: s″′(t)=H(v x ,v y ,v z ,v r ) generates the final micro-vibration output signal.

[0050] Preferably, in one embodiment of the present invention, the multi-modal vibration synthesis adopts a weighted superposition method, and the weight coefficient is obtained by optimizing the genetic algorithm. Specifically, we set the population size to 100, the number of iterations to 50, the crossover probability to 0.8, and the mutation probability to 0.1. This method can maximize the user's comfort while ensuring the vibration effect.

[0051] It is worth noting that the present invention also includes some additional functional modules, which further enhance the practicality and adaptability of the system. These modules include neuroplasticity enhancement training, emotional state regulation, and safety control mechanisms.

[0052] First, regarding neuroplasticity enhancement training, the present invention designs a progressive training program. Specifically, we first set a training goal, such as improving attention or enhancing working memory. Then, based on this goal, a micro-vibration sequence with progressive difficulty is designed. In each training stage, the user's neural activity is monitored by electroencephalography (EEG) or functional near-infrared spectroscopy (fNIRS). Based on these neural activity data, we calculate the similarity score with the target state. Finally, based on this similarity score, the parameters of the micro-vibration sequence for the next stage are adaptively adjusted.

[0053] Preferably, in another embodiment of the present invention, the similarity score is calculated using the cosine similarity method, and the threshold is set to 0.8. When the score is lower than the threshold, the system will reduce the difficulty of training; when the score continues to exceed the threshold, the difficulty will be gradually increased. This dynamic adjustment mechanism ensures the effectiveness and personalization of training.

[0054] Secondly, regarding the emotional state regulation function, the present invention adopts a multi-dimensional evaluation and regulation method.

[0055] The specific steps are as follows:

[0056] 1. Real-time assessment of the user's emotional state through heart rate variability, skin conductance and facial electromyography signals.

[0057] 2. Map the emotional state into a two-dimensional emotional space, including the dimensions of pleasure and arousal.

[0058] 3. Select a preset micro-vibration mode based on the difference between the target emotional state and the current state.

[0059] 4. Guide the user's emotional state to transition to the target state by gradually adjusting the micro-vibration parameters.

[0060] 5. Continuously monitor changes in emotional indicators and dynamically optimize adjustment strategies.

[0061] Preferably, in one embodiment of the present invention, the emotional state assessment adopts a fuzzy logic algorithm. We divide the pleasure and arousal into 5 levels each, forming a total of 25 emotional states. The selection of micro-vibration mode is based on the nearest neighbor principle, that is, the preset mode closest to the target state is selected. The progressive adjustment adopts a linear interpolation method, and the parameters are updated every 5 seconds to ensure a smooth transition.

[0062] Finally, in order to ensure the safety and comfort of users, the present invention also designs a complete safety control mechanism. This mechanism mainly includes the following aspects:

[0063] 1. Real-time monitoring of the user's physiological indicators, including heart rate, blood pressure and skin temperature.

[0064] 2. When physiological indicators exceed the safety threshold, the micro-vibration intensity is automatically reduced or the output is suspended.

[0065] 3. Cumulatively record the daily micro-vibration usage time. When the recommended usage time is exceeded, a reminder will be issued and the usage intensity will be limited.

[0066] 4. Set up an emergency stop mechanism to allow the user to immediately terminate all micro-vibration outputs through specific operations.

[0067] Preferably, in one embodiment of the present invention, the safety threshold is set as follows: the heart rate does not exceed 1.5 times the user's resting heart rate, the blood pressure does not exceed 140 / 90 mmHg, and the skin temperature does not exceed 38°C. The recommended daily usage time is 120 minutes. After this time, the system will issue a rest reminder every 30 minutes and reduce the micro-vibration intensity to 50% of the original.

[0068] In general, the present invention realizes a highly intelligent and personalized micro-vibration neural signal encoding system by comprehensively using advanced technologies such as multi-dimensional vibration parameter space, personalized neural response model, context-aware adaptation, and deep reinforcement learning optimization. This system can not only effectively transmit information, but also dynamically adjust according to the individual characteristics of the user and environmental factors, while taking into account security and user experience.

[0069] In order to realize the above-mentioned neural signal encoding generation method based on micro-vibration, the present invention also provides a corresponding wearable device. The device integrates multiple advanced technologies to provide users with a comfortable, accurate and safe micro-vibration experience.

[0070] At the core of the wearable device is a carefully designed multi-modal micro-vibration generator array. This array consists of 16 high-precision piezoelectric ceramic drivers, which are cleverly arranged in a 4x4 matrix structure. Each driver is a miniature "all-rounder" that can generate precise vibrations in the three spatial directions of X, Y, and Z, and can even rotate. This full range of vibration capabilities allows the device to generate a rich and diverse tactile experience.

[0071] It is worth mentioning that the performance parameters of these actuators are impressive. Their amplitude range is from 0 to 500μm with an accuracy of up to 0.1μm, while the frequency range covers 1Hz to 500Hz with an accuracy of 0.1Hz. By adopting pulse width modulation (PWM) signal control, the system achieves 256 levels of amplitude adjustment and 1024 levels of frequency adjustment. This fine control capability lays a solid hardware foundation for creating complex micro-vibration patterns.

[0072] Complementing this powerful array of micro-vibration generators is a set of high-precision sensor modules. This module is like the "sensory nerve" of the device, constantly collecting the user's physiological data and information about the surrounding environment. It includes a 3-axis accelerometer and a 3-axis gyroscope, which accurately capture every movement of the user with an amazing 1kHz sampling rate and 16-bit resolution. In addition, there is a temperature sensor with an accuracy of ±0.1℃, a photoelectric volumetric pulse wave sensor with a sampling rate of 200Hz, and a skin conductance sensor with a range of 0-100μS and a sampling rate of 100Hz. The coordinated work of these sensors enables the device to fully understand the user's physiological state and activities, providing rich and reliable data support for the real-time optimization of micro-vibration parameters.

[0073] Behind these hardware is a powerful computing and storage system. In a preferred embodiment of the present invention, a dual-processor structure is adopted. The main processor uses the ARM Cortex-M7 architecture with a main frequency of up to 400MHz, which is responsible for the control and data processing of the overall system. It is equipped with a neural network coprocessor specifically for deep learning algorithms. This coprocessor supports up to 5TOPS of computing power and provides strong computing support for complex neural signal encoding algorithms. In terms of storage, the device is equipped with 256MB of LPDDR4 RAM and 64GB of eMMC flash memory, which not only ensures the smoothness of the system operation, but also provides sufficient storage space for a large amount of user data and model parameters.

[0074] In order to ensure that the device can work continuously and stably, the present invention has also made great efforts in power management. The device uses a 500mAh lithium polymer battery, which, in conjunction with an intelligent power management chip, can dynamically adjust power consumption according to usage to maximize the device's battery life. In addition, the device also supports wireless charging, adopts the Qi standard, and supports 15W fast charging, providing users with a convenient charging experience.

[0075] In terms of communication, the wearable device is in full bloom. It supports the latest Bluetooth 5.2 technology, which is not only compatible with the traditional Bluetooth low energy (BLE) mode, but also supports high-speed transmission mode, which can quickly exchange data with devices such as smartphones. At the same time, the device also integrates a Wi-Fi 6 module, supporting 2.4GHz and 5GHz dual-band, which makes it possible to transmit large data and synchronize with the cloud. Not only that, the device is also equipped with NFC function, which supports near-field communication and fast pairing, further improving the user's convenience.

[0076] Finally, in order to ensure a good human-computer interaction experience, the wearable device of the present invention is equipped with a 1.4-inch AMOLED touch screen. This screen has a resolution of 400x400 and a refresh rate of 60Hz, which not only has excellent display effects, but also has sensitive touch response. In addition to the touch screen, the device is also equipped with two physical buttons for quick operation and emergency stop functions, ensuring that users can easily control the device in any situation.

[0077] In summary, the wearable device provided by the present invention provides an ideal carrier for the neural signal coding generation method based on micro-vibration by combining advanced hardware technology with intelligent software algorithms. It can not only accurately execute various complex micro-vibration patterns, but also sense the user status and environmental changes in real time, and make intelligent adjustments and optimizations. This highly integrated and intelligent design makes the device have broad application prospects in many fields such as information transmission, emotion regulation, and cognitive enhancement.

[0078] In order to verify the superiority of the present invention, we designed a series of experiments to compare the performance of the embodiments of the present invention with other prior art solutions. The following is a detailed description of the embodiments and comparative examples, as well as the corresponding test results and analysis.

[0079] Example 1: A neural signal encoding generation method based on micro-vibration and a wearable device thereof according to the present invention.

[0080] Comparative Example 1: The traditional vibration feedback device can only provide vibration of a single frequency and intensity.

[0081] Comparative Example 2: Common smart bracelets on the market have a simple vibration reminder function, but lack personalized adjustment and neural feedback optimization.

[0082] Comparative Example 3: A tactile feedback system with a fixed mode can provide a variety of preset vibration modes, but cannot be adjusted in real time according to user reactions.

[0083] We selected information transmission accuracy, user comfort, neuroplasticity enhancement effect and emotion regulation effect as the main test indicators. These indicators directly reflect the innovation and superiority of the present invention in information encoding, user experience, cognitive enhancement and emotion management.

[0084] The test method is as follows:

[0085] 1. Information transmission accuracy: A series of coded information is transmitted to the subjects to test the proportion of their correct interpretation. It is expressed as a percentage.

[0086] 2. User comfort: Use the Likert seven-point scale to ask the subjects to rate their experience, with 1 point indicating extreme discomfort and 7 points indicating extreme comfort.

[0087] 3. Neuroplasticity enhancement effect: Through a standardized working memory test (n-back task), test the change in the score of the subjects before and after using the device. Record the percentage of score improvement.

[0088] 4. Emotional regulation effect: Use the Self-Assessment Affect Scale (SAM) to measure the change in the emotional state of the subjects before and after the use of the device. Record the degree of emotional regulation to the target state and express it as a percentage.

[0089] We recruited 100 volunteers, aged 20-50 years old, half male and half female. Each volunteer was tested for one week using four different protocols. The test results are shown in the following table:

[0090] Test indicators Example 1 Comparative Example 1 Comparative Example 2 Comparative Example 3 Information transmission accuracy 95％ 60％ 75％ 80％ User comfort 6.2 3.5 4.8 5.1 Neuroplasticity enhancement effect 18％ 2％ 5％ 8％ Emotion regulation effect 85％ 20％ 40％ 55％

[0091] It can be clearly seen from the test results that the embodiments of the present invention are significantly better than the comparative solutions in all indicators. Let us analyze these results one by one:

[0092] First, in terms of information transmission accuracy, the embodiment of the present invention achieves a high accuracy of 95%, far exceeding other solutions. This is due to the multi-dimensional vibration parameter space and deep learning-driven sequence generation technology we use. By precisely controlling multiple parameters of vibration, we are able to create richer and more distinguishable tactile signals, thereby greatly improving the capacity and accuracy of information encoding.

[0093] Secondly, the user comfort test results show that the embodiment of the present invention has obtained a high score of 6.2. This achievement is mainly due to our personalized neural response model and context-aware adaptation mechanism. The system can dynamically adjust the micro-vibration parameters according to each user's unique response characteristics and the current environment, ensuring comfort for long-term use. In contrast, other solutions lack this intelligent adjustment capability and have significantly lower comfort scores.

[0094] In terms of neuroplasticity enhancement, the embodiment of the present invention showed significant advantages, reaching an 18% improvement. This result fully demonstrates the effectiveness of our neuroplasticity enhancement training function. By designing targeted micro-vibration sequences and dynamically adjusting them according to the user's real-time EEG response, we successfully promoted plastic changes in the brain and improved cognitive abilities. Other solutions have relatively limited effects due to the lack of such a specialized training mechanism.

[0095] Finally, in terms of the effect of emotion regulation, the embodiment of the present invention also performed well, reaching an adjustment degree of 85%. This is due to our emotion state assessment system based on multi-dimensional physiological indicators and the carefully designed emotion regulation micro-vibration mode. By monitoring the user's emotional state in real time and applying targeted tactile stimulation, we can effectively guide the user's emotions to the target state. In contrast, other solutions either do not have this function at all or the effect is not ideal.

[0096] In summary, these test results strongly demonstrate the innovation and superiority of the present invention in multiple aspects such as information encoding, user experience, cognitive enhancement and emotion management. Our solution is not only more advanced in technology, but more importantly, it can bring tangible effect improvement and experience enhancement to users. This comprehensive performance advantage makes the present invention have broad application prospects in the fields of wearable devices, human-computer interaction, health management, etc.

[0097] It is worth mentioning that in all test scenarios, we found that the user group aged between 30 and 40 performed the best, and they showed the highest adaptability and learning ability when using the device of the present invention. This finding provides a valuable reference for our future product development and market positioning.

[0098] In general, these experimental results not only verify the technical superiority of the present invention, but also point out the direction for future improvement and optimization. We believe that with the further improvement of technology and the expansion of application scenarios, the neural signal coding generation method based on micro-vibration and its wearable device will bring about a revolutionary change in the fields of human-computer interaction and biomedical engineering.

[0099] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for generating neural signal encoding based on micro-vibration, characterized in that: The method comprises the following steps: receiving input information and extracting feature vectors; mapping the feature vectors to a multidimensional vibration parameter space; generating micro-vibration signals based on the vibration parameter space; adjusting the micro-vibration signals according to a user's personal neural response model; adaptively adjusting the micro-vibration signals based on current context information; collecting user neural response data; calculating the error between the neural response data and a target neural response pattern; optimizing the vibration parameters based on the error; updating the user's personal neural response model; and performing multi-modal fusion on the optimized vibration parameters to obtain a final micro-vibration output signal.

2. The method according to claim 1, characterized in that The multidimensional vibration parameter space includes five dimensions: amplitude, frequency, duration, spatial distribution and rhythmic pattern, wherein: the corresponding vibration parameter vector is obtained by inputting the feature vector into a pre-trained deep neural network; the amplitude range is 0-500μm, the frequency range is 1-500Hz, and the duration range is 10ms-1s; the spatial distribution is represented by a 4x4 matrix, and each element corresponds to the activation intensity of a vibration unit; the rhythmic pattern is represented by the spectral features obtained by Fourier transform.

3. The method according to claim 1, characterized in that The steps of constructing the personal neural response model include: collecting the user's response data to a standard micro-vibration sequence; using a support vector regression algorithm, with the standard micro-vibration sequence as input and the response data as output, to train an initial model; and continuously updating the model parameters according to the user's feedback data during daily use through an online learning algorithm.

4. The method according to claim 1, characterized in that: The adaptive adjustment step comprises: Environmental vibration and noise data are collected through acceleration sensors and microphones; the frequency spectrum characteristics of the environmental data are analyzed using a fast Fourier transform algorithm; when it is detected that the environmental interference in a specific frequency band exceeds a preset threshold, the frequency of the micro-vibration signal is adjusted to avoid the interference frequency band; according to the user's activity status, the intensity of the micro-vibration signal is dynamically adjusted, wherein the activity status is determined by the root mean square value of the acceleration sensor.

5. The method according to claim 1, characterized in that The neuroplasticity enhancement training steps include: setting training goals, such as improving attention or working memory; designing a micro-vibration sequence with progressive difficulty according to the training goals; monitoring the user's neural activity through electroencephalogram or functional near-infrared spectroscopy in each training stage; calculating a similarity score with the target state based on the neural activity data; and adaptively adjusting the micro-vibration sequence parameters of the next stage according to the similarity score.

6. The method according to claim 1, characterized in that The emotional state adjustment step includes: evaluating the user's emotional state in real time through heart rate variability, skin conductance and facial electromyography signals; mapping the emotional state to a two-dimensional emotional space, including two dimensions of pleasure and arousal; selecting a preset micro-vibration mode according to the difference between the target emotional state and the current state; guiding the user's emotional state to transition to the target state by gradually adjusting the micro-vibration parameters; continuously monitoring changes in emotional indicators and dynamically optimizing the adjustment strategy.

7. The method according to claim 1, characterized in that The safety control steps include: real-time monitoring of the user's physiological indicators, including heart rate, blood pressure and skin temperature; automatically reducing the micro-vibration intensity or pausing the output when the physiological indicators exceed the safety threshold; accumulating and recording the daily micro-vibration usage time, and when it exceeds the recommended usage time, issuing a reminder and limiting the usage intensity; setting an emergency stop mechanism to allow the user to immediately terminate all micro-vibration outputs through specific operations.

8. The wearable device according to the method of claim 1, characterized in that: include: A multi-modal micro-vibration generator array consisting of 16 piezoelectric ceramic actuators arranged in a 4x4 matrix; A high-precision sensor module, including a 3-axis accelerometer, a 3-axis gyroscope, a temperature sensor, a photoplethysmography sensor, and a skin conductance sensor; a microprocessor unit, including a main processor and a neural network coprocessor; a storage unit, including 256MB LPDDR4 RAM and 64GB eMMC flash memory; Wireless communication module, supporting Bluetooth 5.2, Wi-Fi 6 and NFC; Power management unit, including 500mAh lithium polymer battery and intelligent power management chip.

9. The wearable device according to claim 8, characterized in that: The specific parameters of the multimodal micro-vibration generator array are as follows: each driver can generate vibrations in the three directions of X, Y, and Z, and can rotate; the amplitude range is 0-500μm, with an accuracy of 0.1μm; the frequency range is 1-500Hz, with an accuracy of 0.1Hz; through PWM signal control, 256 levels of amplitude adjustment and 1024 levels of frequency adjustment can be achieved.

10. The wearable device according to claim 8, characterized in that: The specific parameters of the high-precision sensor module are: 3-axis accelerometer, range of ±16g, sampling rate of 1kHz, 16-bit resolution; 3-axis gyroscope, range of ±2000° / s, sampling rate of 1kHz, 16-bit resolution; temperature sensor, accuracy of ±0.1℃, sampling rate of 10Hz; photoelectric volumetric pulse wave sensor, sampling rate of 200Hz, 12-bit resolution; skin conductance sensor, range of 0-100μS, sampling rate of 100Hz, 14-bit resolution.

Images