Voice input system and method based on friction nanometer power generation

Abstract

The invention belongs to the technical field of voice recognition, and particularly discloses a voice input system and method based on friction nanometer power generation, and the method comprises the steps: collecting a friction electric signal sequence, extracting a signal intensity feature and a fluctuation mode analysis feature from the friction electric signal sequence, and obtaining a preliminary signal fluctuation description; according to the preliminary signal volatility description, processing a part of interference noise contained in the sequence by adopting a wavelet transform algorithm, and determining a pure signal sequence after noise removal; if the signal intensity characteristic in the pure signal sequence exceeds a preset threshold value, judging the signal segment as a potential effective signal segment, and obtaining a time domain characteristic related to the pronunciation action sense from the signal segment to obtain an action sense intensity index; classifying signals under dynamic environment adaptation conditions through matching of motion sense intensity indexes and fluctuation mode analysis to obtain an effective signal candidate set after classification; the invention aims to solve the problem that in the prior art, signal acquisition in a dynamic environment is easily interfered by diversity of noise, humidity and contact modes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of speech recognition, in particular to a speech input system and method based on friction nanogenerator. BACKGROUND

[0002] In the field of modern technology, speech input technology, as an important way of human-computer interaction, is widely used in intelligent devices, medical assistance and barrier-free communication scenarios, and its development has an undeniable value in improving the convenience of life and the inclusiveness of technology. Especially in some special environments, such as scenes where traditional microphones cannot be used, it is particularly urgent to explore new ways of speech input. Friction nanogenerator technology, as a new energy conversion and signal acquisition method, provides a new possibility for speech input. The tiny friction generated by body contact is converted into an electrical signal, which can capture the subtle movements of the user's pronunciation, opening up a new path for interaction in silent or low-voice environments. However, existing speech input methods still face many challenges in signal acquisition and processing, especially in complex environments. Many technologies are difficult to effectively distinguish between target signals and external interference, resulting in a decrease in input accuracy, especially when the user's pronunciation movements are weak or the environmental noise is strong, the system often cannot accurately identify effective speech features. This limitation makes it difficult for new technologies to achieve ideal results in practical applications, especially in scenarios that rely on body contact to generate signals, fluctuations in signal quality become an important bottleneck affecting system stability.

[0003] Further focusing on technical difficulties, the core challenge of friction nanogenerator in speech input lies in the evaluation and screening of signal quality. When the user pronounces, the contact between the oral cavity or tongue and the sensor will produce unstable frictional electric signals. The strength and fluctuation pattern of this signal is affected by the contact method, force, and even external humidity, making it difficult for the system to determine which signals are truly caused by speech actions and which are invalid interference or noise in a stationary state. The deeper problem is that this uncertainty in signal fluctuations directly affects the accuracy of subsequent signal analysis, for example, in some cases, the user's slight pronunciation may be misjudged as invalid input, while meaningless friction in the environment is treated as valid signal processing.

[0004] Therefore, how to accurately evaluate the fluctuation characteristics of frictional electric signals in a dynamically changing contact environment and screen out effective signal segments that truly reflect speech characteristics has become a key problem in improving the practicality and reliability of this technology. SUMMARY

[0005] The present application provides a speech input system and method based on friction nanogenerator, aiming to solve the problem of poor speech recognition accuracy caused by signal collection being easily disturbed by noise, humidity and contact method diversity in dynamic environment in the prior art.

[0006] To solve the above technical problems, the technical solution adopted by the present application is: A voice input method based on friction nanogenerator, the method comprising: collecting a triboelectric signal sequence, extracting signal intensity features and fluctuation pattern analysis features therefrom to obtain a preliminary signal fluctuation description; according to the preliminary signal fluctuation description, using a wavelet transform algorithm to process the part containing interference noise in the sequence to determine the pure signal sequence after removing noise; if the signal intensity feature in the pure signal sequence exceeds a preset threshold, it is judged as a potential effective signal segment, and the time domain feature related to the pronunciation action sense is obtained from it to obtain the action sense intensity index; through the matching of the action sense intensity index and the fluctuation pattern analysis, the signals under the condition of dynamic environment adaptation are classified to obtain the classified effective signal candidate set; according to the classified effective signal candidate set, using a support vector machine algorithm to verify the voice feature capture, judging which paragraphs in the candidate set truly reflect the voice action, and determining the final effective signal segment set; if there are abnormal fluctuations related to the influence of external humidity in the final effective signal segment set, the fluctuation pattern analysis of these paragraphs is adjusted through an adaptive filtering algorithm to obtain the corrected signal segment set; according to the corrected signal segment set, cluster analysis is performed for the diversity of contact methods to obtain the clustering grouping result; through the comparison of the clustering grouping result and the signal evaluation standard, using the k-means algorithm to refine the boundary of the signals within the group to determine the optimized effective signal segment boundary description.

[0007] In an aspect of the present disclosure, the preliminary signal fluctuation description is obtained by collecting a triboelectric signal sequence, extracting signal intensity features and fluctuation pattern analysis features therefrom, comprising: According to the above business content and the extracted related attributes, the following business solutions are generated, around the fluctuation description target of the triboelectric signal, combining the mutually related attributes of signal collection, signal intensity, fluctuation pattern, feature extraction, data processing, sequence analysis, intensity analysis and pattern recognition, the following technical process steps are generated: real-time collection of triboelectric signals by sensor equipment to obtain original signal sequence data, stored as an initial data set; According to the initial data set, using a segmentation processing method to divide the signal sequence into time windows to obtain a segmented signal segment set; For the segmented signal segment set, the signal intensity value in each segment is calculated, and the mean value calculation method is used to determine the intensity feature of each segment; Through fluctuation pattern analysis of the signal segment set, the fluctuation features in the frequency domain are extracted using the fast Fourier transform method to obtain the fluctuation pattern description; If the frequency components in the fluctuation pattern description exceed the preset threshold range, the corresponding signal segment is subjected to secondary filtering to obtain the adjusted fluctuation feature set. Based on the adjusted set of fluctuation characteristics and intensity characteristics, a comprehensive signal volatility description model is constructed to determine the volatility category of the signal; By classifying and storing the output results of the comprehensive signal volatility description model, the final volatility description data can be obtained.

[0008] In one aspect of this disclosure, the step of processing the interference noise portion of the sequence using a wavelet transform algorithm based on a preliminary signal fluctuation description to determine the clean signal sequence after noise removal includes: By initially acquiring the signal sequence, the raw data containing interference noise is obtained, resulting in the signal dataset to be processed. For the signal dataset to be processed, wavelet transform is used to separate the interference noise, and the separated noise components and the preliminary clean signal are obtained. Based on the initial clean signal, a signal reconstruction operation is performed to reassemble the separated signal data and obtain a reconstructed set of clean signals. For the reconstructed set of clean signals, wave characteristics are extracted to obtain the wave change patterns of the signals in the time domain and determine the wave characteristic description. If some segments in the fluctuation characteristic description exceed the preset threshold range, the corresponding segments are subjected to secondary smoothing to obtain the adjusted fluctuation characteristic data. Based on the adjusted volatility characteristic data, a signal volatility classification model is constructed to obtain the category classification of signal volatility and determine the final volatility attribution. By structuring and storing the final volatility attribution results, and using database management tools to save the categorized data, a long-term queryable signal volatility archive is established.

[0009] In one aspect of this disclosure, if the signal intensity characteristics in a pure signal sequence exceed a preset threshold, it is determined to be a potentially valid signal segment, and temporal features related to the sense of vocalization are obtained from it to obtain a sense of vocalization intensity index, including: By scanning the signal strength of the clean signal segment by segment, if a signal strength exceeds a preset threshold, it is marked as a valid signal segment, thus obtaining a preliminary set of filtered signal segments. For the initially screened set of signal segments, feature extraction methods are used to obtain time-domain feature data related to vocalization actions, and to determine the range of signal segments with strong sense of movement. Based on the range of signal segments with strong motion, calculate the corresponding intensity index value, obtain the quantized intensity data of each segment, and determine its salience ranking in the overall signal; By classifying the quantized intensity data after saliency ranking, a pre-established support vector machine model is used to divide the signal segments into categories with different motion intensity, resulting in classified signal groups. For the classified signal groups, the time domain distribution pattern of signal segments in each category is obtained to determine their correlation with the vocalization action. Based on the correlation data, each signal group is assigned a priority label. If the correlation data is higher than the preset standard, it is marked as a high-priority signal segment, thus obtaining the final set of priority-processed signals. By structuring and storing the final set of priority signals, and using database tools to save classification and priority information, long-term queryable signal archive data is determined.

[0010] In one aspect of this disclosure, the classification of signals under dynamic environmental adaptation conditions through matching motion intensity indices with fluctuation pattern analysis to obtain a classified set of effective signal candidates includes: By initially collecting signal data in a dynamic environment, a pre-established support vector machine model is used to classify the signals, resulting in a pre-classified set of signals. Based on the initial classification of the signal set, the intensity of the action is analyzed one by one for each classification group. If the intensity value of a certain classification group exceeds the preset threshold, it is marked as a high-intensity signal group, and a set of high-intensity signal groups is obtained. For a set of high-intensity signal groups, acquire the fluctuation pattern data within each group, compare the fluctuation patterns using a pattern matching method, and determine the range of signal groups that match the preset patterns; Based on the range of signal groups that match the preset mode, environmental adaptability testing is performed. If the adaptability of a certain group in a dynamic environment is found to be lower than the preset standard, it is removed, and a set of signal groups with qualified adaptability is obtained. For the set of signal groups that meet the adaptability requirements, a candidate screening operation is performed. By sorting the signal data within each group, the top-ranked signal data is obtained, and the final set of valid candidate signals is determined. By structuring and storing the final set of effective candidate signals, and using database tools to save classification information and intensity data, signal archive data that can be queried for a long time is obtained. Based on long-term queryable signal archive data, a periodic update mechanism is implemented. If a change in the dynamic environment is detected, a reclassification operation of the signal set is triggered to obtain an updated set of valid candidate signals.

[0011] In one aspect of this disclosure, the step of verifying the captured speech features using a support vector machine algorithm based on the classified set of valid signal candidates, determining which segments in the candidate set truly reflect speech actions, and identifying the final set of valid signal segments includes: By organizing the data of the classified effective signal candidate set, the original data records of each signal segment in the candidate set are obtained, and a preliminary list of signal segments is determined. Based on the preliminary list of signal segments, a pre-established support vector machine model is used to compare and analyze the speech features of each signal segment, resulting in signal segment grouping after feature analysis. For the signal segments grouped after feature analysis, a matching detection of action standards is performed. If the feature data of a certain signal segment does not match the action standard, it is removed, and a set of signal segments that meet the standard is obtained. From the set of signal segments that meet the standards, obtain the judgment criteria data for each signal segment, and determine the range of filtered signal segments by checking the criteria data one by one. For the filtered signal segment range, a signal verification operation is performed. If an anomaly is detected in a signal segment during the verification process, it is marked as pending processing, and a list of verified signal segments is obtained. Based on the list of verified signal segments, the classification results data for each signal segment are organized, and the data is saved using a structured storage tool to determine the final signal segment archive. By periodically checking the final signal segment archive, if the signal segment data in the archive is found to be inconsistent with the current environment, a data update process is triggered to obtain an updated set of signal segments.

[0012] In one aspect of this disclosure, if there are abnormal fluctuations related to external humidity in the final set of valid signal segments, then an adaptive filtering algorithm is used to adjust the fluctuation pattern analysis of these segments to obtain a corrected set of signal segments, including: For the final set of valid signal segments, a preliminary scan of humidity-related signal data is performed using a preset environmental monitoring tool. If the scan results show abnormal fluctuations, these signal segments are marked to obtain a list of marked signal segments. Based on the list of marked signal segments, an adaptive filtering algorithm is used to analyze and correct the fluctuation patterns of the marked signal segments one by one, resulting in corrected signal segment groups. For the corrected signal segment groups, the fluctuation pattern of each signal segment is matched with the preset standard using a data comparison tool. If the fluctuation pattern of a certain signal segment does not match the standard, it is classified as pending processing, and a set of matched signal segments is obtained. Based on the set of matched signal segments, obtain the environmental association data for each signal segment, and use data processing tools to store this data in a structured manner to determine the processed signal segment archive; For the organized signal segment files, the signal segments in the files are compared with the current external environmental data through a periodic scanning tool. If a signal segment is found to be inconsistent with the environmental data, it is marked as needing to be updated, and a list of signal segments to be updated is obtained. Based on the list of signal segments to be updated, the environmental data of the marked signal segments is reacquired and recorded through the data acquisition module to obtain the updated signal segment dataset. For the updated signal segment dataset, the updated data is merged with the original signal segment archives using a data integration tool to determine the final comprehensive signal segment archive.

[0013] In one aspect of this disclosure, the step of performing cluster analysis on the diversity of contact mode variations based on the corrected signal segment set to obtain cluster grouping results includes: To establish the relationship between the signal segment set and the calibration data, the calibration data is initially organized using a data filtering tool to obtain the basic file of the organized signal segments. Based on the compiled basic data of signal segments, and combined with the relationship between contact methods and diversity patterns, a classification tool is used to perform pattern recognition on the contact methods of signal segments to determine the list of signal segments after pattern recognition. Based on the list of signal segments after pattern recognition, and combining the correlation between clustering methods and grouping results, the signal segments are grouped using clustering analysis tools to obtain a set of grouped signal segments. Based on the grouped signal segment set, and combined with the correlation between feature analysis and mode differences, the contact mode feature data of each group is obtained, and the signal segment directory after feature extraction is determined. For the signal segment catalog after feature extraction, and combining the correlation between signal classification and result division, if the feature data of a certain group does not match the preset standard, it is marked to obtain the marked signal segment group. Based on the marked signal segment groups, and combined with the relationship between data combination and set processing, the marked groups are reorganized a second time through the data integration module to determine the final signal segment classification file. For the final signal segment classification file, combining the correlation between the differences in the method and the result classification, if the contact method characteristics of a certain category are detected to be inconsistent with the expectations, its classification label is updated through the recording tool to obtain an updated signal segment classification list.

[0014] In one aspect of this disclosure, the step of comparing the clustering results with signal evaluation criteria and using the k-means algorithm to refine the boundaries of signals within the groups to determine the optimized effective signal segment boundary description includes: By associating clustering with signal data, a data filtering module is used to perform preliminary classification of the signal data, resulting in a classified signal dataset. Based on the classified signal dataset, and combining the grouping results with the evaluation criteria, the signal dataset is matched against the criteria using a comparison tool to determine the matched signal data list. For the matched list of signal data, and combining the association between signal boundaries and group boundaries, a boundary delineation tool is used to mark the boundaries of the signal data to obtain the marked set of signal data. Based on the labeled signal data set, and combined with the correlation between optimization processing and boundary description, the signal boundaries are descriptively summarized through the data processing module to obtain the summarized boundary description record. For the summarized boundary description records, combined with the correlation between data processing and evaluation standards, if the boundary description of a certain segment of signal data is inconsistent with the preset standard, the boundary is corrected by adjustment tools to determine the corrected signal boundary file. By combining the revised signal boundary profile with the correlation between standard comparison and optimization processing, the information integration module is used to finally organize the signal boundary profile, resulting in the organized signal boundary dataset. For the organized signal boundary dataset, and combining the relationship between signal evaluation and data processing, the signal boundary dataset is classified and stored using a recording tool to obtain the stored signal boundary directory.

[0015] In another aspect, this disclosure also relates to a voice input system based on triboelectric nanogenerators, the system comprising: The signal acquisition and feature extraction module is used to acquire triboelectric signal sequences, extract signal intensity features and fluctuation pattern analysis features from them, and obtain a preliminary description of signal fluctuation. The wavelet denoising module is used to process the interference noise in the sequence based on the initial signal fluctuation description and the wavelet transform algorithm to determine the clean signal sequence after noise removal. The time-domain feature extraction module is used to determine whether the signal intensity features in a clean signal sequence exceed a preset threshold. If they exceed the threshold, they are judged as potentially effective signal segments, and the time-domain features related to the sense of pronunciation action are obtained from them to obtain the sense of action intensity index. The dynamic classification module is used to classify signals under dynamic environmental adaptation conditions by matching motion intensity index with fluctuation pattern analysis, and obtain a set of effective signal candidates after classification. The support vector machine verification module is used to verify the captured speech features using the support vector machine algorithm based on the classified effective signal candidate set, determine the segments in the candidate set that truly reflect speech actions, and determine the final set of effective signal segments. The adaptive filtering correction module is used to determine whether there are abnormal fluctuations related to the influence of external humidity in the final effective signal segment set. If there are abnormalities, the fluctuation pattern analysis of these segments is adjusted through the adaptive filtering algorithm to obtain the corrected signal segment set. The clustering analysis module is used to perform clustering analysis on the diversity of contact mode changes based on the corrected signal segment set, and obtain clustering results; The clustering optimization module is used to refine the boundaries of signals within a group by comparing the clustering results with the signal evaluation criteria and using the k-means algorithm to determine the optimized effective signal segment boundary description.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention achieves effective recognition and signal optimization of speech actions through multi-level signal processing and feature analysis. First, it extracts preliminary features through signal strength and fluctuation pattern analysis, then uses wavelet transform to remove noise interference, subsequently screening for potentially effective signal segments and extracting time-domain features. Support vector machines are then used to verify the authenticity of speech actions, determining the set of effective signal segments. For abnormal fluctuations caused by humidity, this invention utilizes adaptive filtering for correction and optimizes signal boundaries through cluster analysis and the k-means algorithm, ultimately obtaining accurate signal segment descriptions. The technical advantage of this invention lies in significantly improving the recognition accuracy and stability of speech action signals in complex environments, providing a reliable solution for signal processing in dynamic environments. Attached Figure Description

[0017] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained from these drawings without creative effort.

[0018] Fig. 1 This is one of the flowcharts for a voice input method based on triboelectric nanogenerator according to the present invention.

[0019] Fig. 2 This is the second flowchart of a voice input method based on triboelectric nanogenerator according to the present invention.

[0020] Fig. 3 This is the third flowchart of a voice input method based on triboelectric nanogenerator according to the present invention. Detailed Implementation

[0021] The present invention will be further described below with reference to embodiments. These embodiments are merely some, not all, of the embodiments described. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the protection scope of the present invention.

[0022] Please see Figs. 1-3 As shown in the figure, this embodiment discloses a voice input method based on triboelectric nanogenerators, which may specifically include: Step S101: By acquiring the triboelectric signal sequence, signal intensity features and fluctuation mode analysis features are extracted from it to obtain a preliminary description of signal fluctuation.

[0023] Based on the aforementioned business content and extracted relevant attributes, the following business solution is generated. Focusing on the description of the volatility of triboelectric signals, and combining interrelated attributes such as signal acquisition, signal strength, fluctuation patterns, feature extraction, data processing, sequence analysis, intensity analysis, and pattern recognition, the following technical process steps are generated: Real-time acquisition of triboelectric signals using sensor devices to obtain raw signal sequence data, which is stored as an initial dataset. Based on the initial dataset, the signal sequence is divided into time windows using a segmentation method, resulting in a set of segmented signal segments. For each segmented signal segment set, the signal strength value is calculated, and the intensity characteristics of each segment are determined using a mean calculation method. Fluctuation pattern analysis is performed on the signal segment set, and the frequency domain fluctuation features are extracted using the Fast Fourier Transform method to obtain a fluctuation pattern description. If the frequency components in the fluctuation pattern description exceed a preset threshold range, the corresponding signal segment undergoes secondary filtering to obtain an adjusted fluctuation feature set. Based on the adjusted fluctuation feature set and intensity characteristics, a comprehensive signal volatility description model is constructed to determine the signal volatility category. The output results of the comprehensive signal volatility description model are classified and stored to obtain the final volatility description data.

[0024] Specifically, in the process of acquiring triboelectric signal sequences and extracting signal strength and fluctuation mode features to obtain a preliminary description of signal fluctuation, triboelectric signal data is first acquired using a sensor device at a sampling rate of 1000 times per second. Assuming a acquisition time of 10 seconds, 10,000 data points are obtained, forming a time series dataset. Subsequently, signal processing algorithms are used to extract intensity features from this dataset, such as calculating the root mean square (RMS) value of the signal. Specifically, this is done by summing the squares of each data point, dividing by the total number of data points, and then taking the square root. Assuming the calculated result is an RMS value of 2.5 volts, it indicates that the overall signal strength is at a moderate level. Next, for fluctuation mode feature analysis, a fast Fourier transform algorithm can be used to convert the time-domain signal into a frequency-domain signal, extracting the main frequency components. Assuming the analysis results show a dominant frequency of 50 Hz and a secondary frequency of 100 Hz, it indicates that the signal fluctuation has periodic characteristics. Simultaneously, calculating the spectral energy distribution reveals that the energy proportion at 50 Hz is 60%, reflecting that this frequency is the dominant fluctuation mode. To further describe signal volatility, the fluctuation amplitude can be quantified by calculating the signal's standard deviation. Assuming a standard deviation of 0.8 volts, and combining the RMS value and spectral analysis results, the signal volatility can be initially described as "moderate intensity, significant periodic fluctuations, and relatively stable fluctuation amplitude." To establish a rigorous logical relationship, these characteristics can be correlated with the equipment's operating status. Assuming the signal originates from mechanical friction components, by comparing the standard deviation with historical data (assuming a historical average standard deviation of 0.5 volts), it can be inferred that the current fluctuation amplitude is relatively high, potentially indicating a slight risk of wear on the friction components. This allows the signal analysis results to be combined with the business scenario, forming a complete thought process chain from data acquisition to feature extraction and application inference.

[0025] Step S102: Based on the preliminary signal fluctuation description, the wavelet transform algorithm is used to process the interference noise in the sequence to determine the clean signal sequence after noise removal.

[0026] Initial acquisition of the signal sequence yields raw data containing interference noise, resulting in a signal dataset to be processed. Wavelet transform is then used to separate the interference noise from the dataset, obtaining the separated noise components and a preliminary clean signal. Based on the preliminary clean signal, signal reconstruction is performed, recombining the separated signal data to obtain a reconstructed clean signal set. For the reconstructed clean signal set, wave characteristics are extracted to obtain the signal's wave pattern in the time domain, defining the wave characteristic description. If certain segments in the wave characteristic description exceed a preset threshold, these segments undergo secondary smoothing to obtain adjusted wave characteristic data. Based on the adjusted wave characteristic data, a signal volatility classification model is constructed to determine the category of signal volatility and assign the final volatility classification. The final volatility classification results are then structured and stored using a database management tool to save the classification data, establishing a long-term queryable signal volatility archive.

[0027] Specifically, based on the preliminary description of signal fluctuations, wavelet transform algorithm is used to process the potential interference noise in the triboelectric signal sequence to obtain a clean signal sequence. The specific implementation method is as follows: First, assume that there is a triboelectric signal time series data with a sampling rate of 500 times per second and a collection time of 20 seconds, totaling 10,000 data points, which includes background noise and random interference. Next, the Daubechies wavelet basis (db8) in the wavelet transform algorithm is selected as the decomposition tool to decompose the signal into 5 layers of wavelet coefficients. By analyzing the energy distribution of the coefficients in each layer, it is determined that the high-frequency part (layers 1 and 2) mainly contains noise components, and its energy proportion is approximately 15% of the total energy. Subsequently, a threshold processing method is set, and a soft threshold function is used to filter the high-frequency coefficients. Assuming the threshold is set to 0.3, coefficients below this value are set to zero, and coefficients above this value are reduced proportionally. After processing, the energy distribution is recalculated, and it is found that the energy proportion of high-frequency noise is reduced to 3%. Finally, the signal was reconstructed using inverse wavelet transform to obtain a clean signal sequence. The signal-to-noise ratio (SNR) before and after processing was compared. Assuming the original SNR was 12.5 dB, the processed SNR increased to 18.7 dB, indicating a significant reduction in noise interference. To establish a rigorous logical relationship, the clean signal sequence was integrated with equipment health monitoring. Assuming the signal originates from bearing friction components, a comparison with historical clean signal waveforms (assuming a historical average SNR of 20.0 dB) revealed that the current SNR is slightly below the normal range, potentially indicating early abnormal wear on the bearing surface. This provides data support for subsequent maintenance decisions.

[0028] Step S103: If the signal intensity feature in the pure signal sequence exceeds a preset threshold, it is judged as a potential effective signal segment, and the time domain features related to the sense of pronunciation action are obtained from it to obtain the sense of movement intensity index.

[0029] By scanning the signal strength of the clean signal segment by segment, if a signal segment's strength exceeds a preset threshold, it is marked as a valid signal segment, resulting in a preliminary set of selected signal segments. For this preliminary set, feature extraction methods are used to obtain time-domain feature data related to the vocalization action, determining the range of signal segments with strong vocalization. Based on this range, corresponding intensity index values ​​are calculated, obtaining quantized intensity data for each segment and determining its salience ranking within the overall signal. The quantized intensity data, after salience ranking, is then classified using a pre-established support vector machine model, dividing the signal segments into categories with different vocalization intensities, resulting in classified signal groups. For each classified signal group, the time-domain distribution pattern of signal segments within that category is obtained, determining its correlation with the vocalization action. Based on this correlation data, each signal group is prioritized; if the correlation data exceeds a preset standard, it is marked as a high-priority signal segment, resulting in the final set of priority-processed signals. This final set of priority-processed signals is then structured and stored using a database tool to save the classification and priority information, ensuring long-term queryability of the signal archive data.

[0030] Specifically, assuming a processed, clean signal sequence is available, the data originates from vibration sensors related to human vocalization movements, with a sampling rate of 1000 times per second and a collection duration of 10 seconds, totaling 10,000 data points. First, the system automatically detects the signal strength, with a preset threshold of 0.5 volts. By traversing the entire sequence, it identifies consecutive time periods where the signal strength exceeds this threshold. Assuming three time periods meet this condition, with a total duration of 2.8 seconds, these are determined to be potentially valid signal segments. Next, for these valid signal segments, the system employs a time-domain feature extraction algorithm to calculate the root mean square (RMS) value of the signal as an indicator of the intensity of the movement. Assuming the calculated RMS value is 0.72 volts, peak features are also extracted, with a peak value reaching 1.2 volts. Further analysis of the signal fluctuation frequency within the time period reveals that it is mainly concentrated in the 5-10 Hz range, consistent with the typical vibration characteristics of vocalization movements. Subsequently, the system compares the extracted motion intensity index with the historical database. Assuming the RMS mean of normal vocalization movements in the historical data is 0.68 volts, and the current value is slightly higher than the average, combined with fluctuation frequency analysis, it infers a potential characteristic of excessive vocalization effort. To establish a rigorous logical relationship, this analysis result is integrated with speech rehabilitation training services. The system automatically generates an assessment report, indicating that the current vocalization movement may have an abnormal tendency towards excessive force, which can serve as a reference for subsequent training adjustments, thereby supporting the optimized design of personalized rehabilitation programs.

[0031] Step S104: By matching the motion intensity index with the fluctuation pattern analysis, the signals under dynamic environmental adaptation conditions are classified to obtain a set of effective signal candidates after classification.

[0032] Initial signal data is collected under dynamic conditions, and a pre-established support vector machine model is used to classify the signals, resulting in a preliminary signal set. Based on this preliminary set, the intensity of each category is analyzed. If the intensity of a category exceeds a preset threshold, it is marked as a high-intensity signal group, and a high-intensity signal group set is obtained. For each high-intensity signal group, fluctuation pattern data is acquired, and pattern matching is used to compare the fluctuation patterns and determine the range of signal groups that match preset patterns. Based on the range of signal groups matching preset patterns, environmental adaptability testing is performed. If a group's adaptability in the dynamic environment is found to be below a preset standard, it is removed, resulting in a qualified signal group set. For the qualified signal group set, a candidate selection process is performed. The signal data within each group is sorted, and the top-ranked signals are selected to determine the final valid candidate signal set. The final valid candidate signal set is then structured and stored using a database tool to save the classification information and intensity data, resulting in a long-term queryable signal archive. Based on long-term queryable signal archive data, a periodic update mechanism is implemented. If a change in the dynamic environment is detected, a reclassification operation of the signal set is triggered to obtain an updated set of valid candidate signals.

[0033] Specifically, under dynamic environment adaptation conditions, the system first acquires signal data from vibration sensors related to human vocalization. Assuming a sampling rate of 800 times per second and a collection time of 8 seconds, a total of 6400 data points are collected. The signal is affected by environmental noise and requires classification processing. The system preprocesses the signal using an adaptive filtering algorithm to filter out noise interference below 0.3 volts, retaining the main signal components. Then, it extracts the motion intensity index and uses short-time energy analysis to calculate the average energy value within each 0.5-second window, obtaining a set of energy distribution data, assuming a peak energy of 0.85 volts. Next, the system analyzes the signal fluctuation pattern, using a wavelet transform algorithm to decompose the signal and extract feature components in the frequency range of 3-8 Hz. It finds that the main fluctuation period is 0.2 seconds, which is related to the dynamic characteristics of vocalization. Based on this, the system classifies signals according to energy value and fluctuation period, setting the energy threshold at 0.6 volts and the period range at 0.15-0.25 seconds. Signal segments meeting these criteria are categorized into a valid candidate set. Assuming four candidate segments are ultimately selected, with a total duration of 1.9 seconds, the system establishes a robust logical relationship by integrating the classification results with pronunciation behavior analysis, automatically generating a feature matching report for subsequent pronunciation pattern evaluation in dynamic environments. This supports parameter adjustments in the adaptive training system, ensuring the application value of the classification results.

[0034] Step S105: Based on the classified effective signal candidate set, the speech feature capture is verified using the support vector machine algorithm to determine which segments in the candidate set truly reflect speech actions, and to determine the final set of effective signal segments.

[0035] By organizing the classified and valid candidate signal data, the original data records of each signal segment in the candidate set are obtained to determine a preliminary list of signal segments. Based on this preliminary list, a pre-established support vector machine model is used to compare and analyze the speech features of each signal segment, resulting in signal segment groups after feature analysis. For these groups, action standard matching detection is performed. If the feature data of a signal segment does not match the action standard, it is removed, resulting in a set of signal segments that meet the standard. From this set of standard-compliant signal segments, the judgment criteria data for each signal segment is obtained. By verifying each criterion data, the range of filtered signal segments is determined. For this range, signal verification is performed. If an anomaly is detected during verification, the segment is marked as pending processing, resulting in a list of verified signal segments. Based on this list, the classification results data for each signal segment are organized and stored using a structured storage tool to determine the final signal segment archive. The final signal segment archive is periodically checked. If the signal segment data in the archive does not match the current environment, a data update process is triggered, resulting in an updated set of signal segments.

[0036] Specifically, in speech signal processing under dynamic conditions, the system further verifies and filters the classified effective signal candidate set to determine the final speech action-related signal segments. First, the system extracts four signal segments from the candidate set, with a total duration of 2.2 seconds. It assumes that the average duration of each signal segment is 0.55 seconds, and the signal amplitude ranges from 0.4 to 0.9 volts. Next, the system uses a support vector machine algorithm to classify and verify the speech features of these signal segments. Specifically, each signal segment is divided into 0.1-second sub-segments, and the spectral characteristics of each sub-segment are calculated. Assuming a frequency resolution of 0.5 Hz for the spectral analysis, energy distribution data of the main frequency components within the range of 2 to 10 Hz is extracted. It is found that the energy concentration of two signal segments exceeds 75%. Subsequently, the system classifies the feature vectors using a support vector machine model, setting the classification boundary parameter to 0.7 and selecting the radial basis function as the kernel function. The calculated classification confidence scores for two signal segments were 0.82 and 0.78, exceeding the preset threshold of 0.75. The confidence scores for the other two signal segments were only 0.62 and 0.58, and were therefore classified as non-voice action-related signals. Finally, the system categorized the two signal segments with confidence scores above the threshold into the final set of valid signal segments, with a total duration of 1.1 seconds. To ensure logical rigor, the system correlates the verification results with the voice behavior analysis business, automatically generating a classification verification log that records the feature values ​​and confidence scores of each signal segment. This log is then used by the subsequent voice pattern optimization module to support adjustments to signal processing strategies in dynamic environments.

[0037] Step S106: If there are abnormal fluctuations related to the influence of external humidity in the final set of valid signal segments, the fluctuation pattern analysis of these segments is adjusted by an adaptive filtering algorithm to obtain the corrected set of signal segments.

[0038] For the final set of valid signal segments, a preliminary scan of humidity-related signal data is performed using a pre-set environmental monitoring tool. If the scan results show abnormal fluctuations, these signal segments are marked, resulting in a list of marked signal segments. Based on this list, an adaptive filtering algorithm is used to analyze and correct the fluctuation patterns of each marked signal segment, resulting in corrected signal segment groups. For these corrected groups, a data comparison tool is used to match the fluctuation patterns of each signal segment with a pre-set standard. If a signal segment's fluctuation pattern does not match the standard, it is categorized as needing further processing, resulting in a set of matched signal segments. Based on this set, environmental correlation data for each signal segment is acquired, and this data is structured and stored using a data processing tool, resulting in a processed signal segment archive. For this archive, a periodic scanning tool compares the signal segments in the archive with current external environmental data. If a signal segment is found to be inconsistent with the environmental data, it is marked as needing updating, resulting in a list of signal segments to be updated. Based on this list, the data acquisition module re-acquires and records environmental data for the marked signal segments, resulting in an updated signal segment dataset. For the updated signal segment dataset, the updated data is merged with the original signal segment archives using a data integration tool to determine the final comprehensive signal segment archive.

[0039] Specifically, in the field of speech signal processing in dynamic environments, the system detects and corrects abnormal fluctuations related to external humidity that may exist in the final set of valid signal segments to ensure signal quality. First, the system scans the environmental parameters of the final set of valid signal segments. Assuming the set contains three signal segments with a total duration of 1.5 seconds and an average duration of 0.5 seconds per segment, ambient humidity data is acquired through a built-in humidity sensor interface. Assuming the current humidity is 78%, and based on historical data analysis, it is found that signal fluctuation amplitude increases by approximately 15% when humidity exceeds 75%. Next, the system detects fluctuations in each signal segment. Specifically, the signal segment is divided into 0.2-second sub-intervals, and the peak and trough values ​​of each sub-interval are calculated. If the fluctuation amplitude of the first signal segment reaches 0.3 volts, exceeding the normal range threshold of 0.1 volts, it is determined to be an abnormal fluctuation. Subsequently, the system initiates an adaptive filtering algorithm to adjust the abnormal signal segments. Employing the minimum mean square error algorithm, with a filter order of 8 and an iteration step size of 0.01, the system smooths the time-domain data of the signal segments, reducing fluctuation amplitude to within 0.12 volts while preserving the main frequency components of the original signal within the 3-8 Hz range. Finally, the system generates a corrected set of signal segments, with a total duration of 1.5 seconds. The system then correlates the fluctuation data before and after correction with ambient humidity values ​​and automatically stores it in the signal quality monitoring database for subsequent use by the environmental adaptability optimization module, supporting adjustments to signal processing strategies under different humidity conditions.

[0040] Step S107: Based on the corrected signal segment set, perform cluster analysis on the diversity of contact mode changes to obtain cluster grouping results.

[0041] To establish the correlation between the signal segment set and the calibration data, a data filtering tool is used to initially organize the calibration data, resulting in a basic signal segment profile. Based on this basic profile and considering the correlation between contact methods and diversity patterns, a classification tool is used to perform pattern recognition on the contact methods of the signal segments, determining a list of signal segments after pattern recognition. For this list, a clustering analysis tool is used to group the signal segments, resulting in a set of grouped signal segments, considering the correlation between clustering methods and grouping results. Based on these grouped sets, and considering the correlation between feature analysis and method differences, contact method feature data for each group is obtained, determining a directory of signal segments after feature extraction. For this directory, and considering the correlation between signal classification and result segmentation, if the feature data of a group does not match a preset standard, it is marked, resulting in marked signal segment groups. Finally, based on these marked signal segment groups and considering the correlation between data combination and set processing, a data integration module performs a secondary processing on the marked groups, determining the final signal segment classification profile. For the final signal segment classification file, combining the correlation between the differences in the method and the result classification, if the contact method characteristics of a certain category are detected to be inconsistent with the expectations, its classification label is updated through the recording tool to obtain an updated signal segment classification list.

[0042] Specifically, in the field of speech signal processing in dynamic environments, the system performs cluster analysis based on the diversity of contact methods using a corrected set of signal segments to obtain grouping results. First, the system extracts features from the corrected set of signal segments. Assuming the set contains 5 signal segments with a total duration of 2.5 seconds and an average duration of 0.5 seconds per segment, the system analyzes the time-domain and frequency-domain characteristics of each signal segment to extract parameters related to the contact method, such as the average energy value of the signal being 2.3 milliwatts and the frequency distribution ranging from 5 to 10 Hz. Subsequently, the system employs the K-means clustering algorithm for grouping, setting the number of cluster centers to 3. The initial center points are randomly sampled based on the signal energy distribution, with an upper limit of 50 iterations. In each iteration, the Euclidean distance from each signal segment to the center point is calculated. Assuming that after the first iteration, the distance from signal segment 1 to center point 1 is 0.8, and the distance from signal segment 2 to center point 2 is 1.2, the system reassigns signal segments based on these distances and updates the center point positions. After 30 iterations, the system converges, resulting in three cluster groups: the first group contains two signal segments with an average energy of 2.1 milliwatts; the second group contains two signal segments with an average energy of 2.5 milliwatts; and the third group contains one signal segment with an average energy of 2.0 milliwatts. Next, the system performs feature validation on the clustering results. By calculating the within-group variance and between-group variance (assuming a within-group variance of 0.15 and a between-group variance of 0.85), the significance of the clustering results is verified, confirming the reasonableness of the grouping. Finally, the system performs correlation analysis between the clustering results and the physical characteristics of the contact methods, and generates classification labels by combining the contact pattern data stored in the historical database. For example, the first group corresponds to contact method A, the second group corresponds to contact method B, and the third group corresponds to contact method C. The results are automatically stored in the signal classification database to provide a reference for subsequent signal processing strategy optimization.

[0043] Step S108: By comparing the clustering results with the signal evaluation criteria, the k-means algorithm is used to refine the boundaries of the signals within the groups, and the optimized effective signal segment boundary description is determined.

[0044] By associating clustering with signal data, a data filtering module performs preliminary classification of the signal data, resulting in a classified signal dataset. Based on this classified dataset, and considering the association between the grouping results and evaluation criteria, a comparison tool is used to perform standard matching on the signal dataset, determining a list of matched signal data. For this list, a boundary delimitation tool is used to annotate the signal data boundaries, based on the association between signal boundaries and grouping boundaries, obtaining an annotated signal data set. Based on this annotated signal data set, and considering the association between optimization processing and boundary description, a data processing module performs descriptive summarization of the signal boundaries, obtaining summarized boundary description records. For these summarized boundary description records, and considering the association between data processing and evaluation criteria, if the boundary description of a segment of signal data is inconsistent with the preset standard, an adjustment tool is used to correct the boundary, determining a corrected signal boundary file. Using this corrected signal boundary file, and considering the association between standard comparison and optimization processing, an information integration module performs final processing of the signal boundary file, obtaining a processed signal boundary dataset. Finally, considering the association between signal evaluation and data processing, a recording tool is used to classify and store the signal boundary dataset, obtaining a stored signal boundary directory.

[0045] Specifically, in the field of speech signal processing in dynamic environments, the system compares the clustering results with signal evaluation criteria and uses the K-means algorithm to refine the signal boundaries within each group, ultimately determining the optimized effective signal segment boundary descriptions. The system first retrieves existing clustering results from the database. Assuming there are four signal segments in two groups, the first group has three segments with an average duration of 0.6 seconds, and the second group has one segment with a duration of 0.4 seconds. These segments are then compared with preset signal evaluation criteria, including a signal duration threshold of 0.5 seconds and an energy fluctuation range between 1.5 and 3.0 milliwatts. The comparison reveals that one segment in the first group has a duration of 0.45 seconds, which does not meet the threshold, and an energy fluctuation of 1.2 milliwatts, requiring further refinement. Next, the system initiates the K-means algorithm for boundary adjustment, setting the number of cluster centers to 2. The initial center points are based on the signal duration distribution, with an iteration limit of 20. In the first iteration, the distance from each signal segment to the center point is calculated. Assuming the distance from signal segment 1 to center point 1 in the first group is 0.3 and the distance from signal segment 2 to center point 2 is 0.5, the system reassigns signal segments based on these distances and updates the center points. After 15 iterations, the algorithm converges. The refined boundary of the first group of signal segments now contains two signal segments, the average duration is increased to 0.55 seconds, and the energy fluctuation is adjusted to 1.8 milliwatts, meeting the evaluation criteria. The system then verifies the adjusted boundary, calculating the duration consistency of the signals within the boundary. Assuming a consistency index of 0.9, reaching the preset threshold of 0.85, the boundary optimization is confirmed to be reasonable. Finally, the system automatically updates the optimized signal segment boundary description to the signal processing database and links it with the subsequent signal classification module to ensure the continuity of signal segments at different processing stages, forming a complete logical chain from boundary refinement to classification application.

[0046] This invention provides a voice input system based on triboelectric nanogenerators, mainly comprising: The signal acquisition and feature extraction module is used to acquire triboelectric signal sequences, extract signal intensity features and fluctuation pattern analysis features from them, and obtain a preliminary description of signal fluctuation. The wavelet denoising module is used to process the interference noise in the sequence based on the initial signal fluctuation description and the wavelet transform algorithm to determine the clean signal sequence after noise removal. The time-domain feature extraction module is used to determine whether the signal intensity features in a clean signal sequence exceed a preset threshold. If they exceed the threshold, they are judged as potentially effective signal segments, and the time-domain features related to the sense of pronunciation action are obtained from them to obtain the sense of action intensity index. The dynamic classification module is used to classify signals under dynamic environmental adaptation conditions by matching motion intensity index with fluctuation pattern analysis, and obtain a set of effective signal candidates after classification. The support vector machine verification module is used to verify the captured speech features using the support vector machine algorithm based on the classified effective signal candidate set, determine the segments in the candidate set that truly reflect speech actions, and determine the final set of effective signal segments. The adaptive filtering correction module is used to determine whether there are abnormal fluctuations related to the influence of external humidity in the final effective signal segment set. If there are abnormalities, the fluctuation pattern analysis of these segments is adjusted through the adaptive filtering algorithm to obtain the corrected signal segment set. The clustering analysis module is used to perform clustering analysis on the diversity of contact mode changes based on the corrected signal segment set, and obtain clustering results; The clustering optimization module is used to refine the boundaries of signals within a group by comparing the clustering results with the signal evaluation criteria and using the k-means algorithm to determine the optimized effective signal segment boundary description.

[0047] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1.A voice input method based on frictional nanogeneration, characterized by, The method comprises: acquiring a triboelectric signal sequence, extracting signal intensity features and fluctuation pattern analysis features therefrom, and obtaining a preliminary signal fluctuation description; processing the part containing interference noise in the sequence using a wavelet transform algorithm according to the preliminary signal fluctuation description, and determining a pure signal sequence after removing noise; if the signal intensity features in the pure signal sequence exceed a preset threshold, judging that it is a potential effective signal segment, and obtaining a time domain feature related to pronunciation action sensitivity from the potential effective signal segment, and obtaining an action sensitivity index; classifying the signal under the dynamic environment adaptation condition through matching of the action sensitivity index and the fluctuation pattern analysis, and obtaining a classified effective signal candidate set; verifying the speech feature capture using a support vector machine algorithm according to the classified effective signal candidate set, judging the segment in the candidate set that truly reflects the speech action, and determining a final effective signal segment set; if there is an abnormal fluctuation related to the influence of external humidity in the final effective signal segment set, adjusting the fluctuation pattern analysis of the segment through an adaptive filtering algorithm, and obtaining a corrected signal segment set; performing clustering analysis on the corrected signal segment set for the diversity of contact modes, and obtaining a clustering grouping result; comparing the clustering grouping result with a signal evaluation standard, refining the boundaries of the signals in the grouping using a k-means algorithm, and determining an optimized effective signal segment boundary description. 2.The voice input method based on friction nanogenerator according to claim 1, wherein, The method comprises: acquiring a triboelectric signal sequence, extracting signal intensity features and fluctuation pattern analysis features therefrom, and obtaining a preliminary signal fluctuation description; acquiring the original signal sequence data through real-time acquisition of the triboelectric signal by a sensor device, and storing it as an initial data set; dividing the signal sequence into time windows using a segmentation processing method according to the initial data set, and obtaining a segmented signal segment set; calculating the signal intensity value in each segment using a mean value calculation method, and determining the intensity feature of each segment; extracting the fluctuation features in the frequency domain using a fast Fourier transform method through fluctuation pattern analysis of the signal segment set, and obtaining a fluctuation pattern description; if the frequency components in the fluctuation pattern description exceed a preset threshold range, performing secondary filtering processing on the corresponding signal segment, and obtaining an adjusted fluctuation feature set; constructing a comprehensive signal fluctuation description model according to the adjusted fluctuation feature set and the intensity feature, and judging the fluctuation category of the signal; 3.The voice input method based on friction nanogenerator according to claim 1, wherein, storing the output results of the comprehensive signal fluctuation description model, and obtaining the final fluctuation description data. The method comprises: acquiring a triboelectric signal sequence, extracting signal intensity features and fluctuation pattern analysis features therefrom, and obtaining a preliminary signal fluctuation description; acquiring the original signal sequence data through real-time acquisition of the triboelectric signal by a sensor device, and storing it as an initial data set; dividing the signal sequence into time windows using a segmentation processing method according to the initial data set, and obtaining a segmented signal segment set; calculating the signal intensity value in each segment using a mean value calculation method, and determining the intensity feature of each segment; extracting the fluctuation features in the frequency domain using a fast Fourier transform method through fluctuation pattern analysis of the signal segment set, and obtaining a fluctuation pattern description; if the frequency components in the fluctuation pattern description exceed a preset threshold range, performing secondary filtering processing on the corresponding signal segment, and obtaining an adjusted fluctuation feature set; constructing a comprehensive signal fluctuation description model according to the adjusted fluctuation feature set and the intensity feature, and judging the fluctuation category of the signal; storing the output results of the comprehensive signal fluctuation description model, and obtaining the final fluctuation description data. The fluctuation characteristic extraction is implemented for the reconstructed pure signal set, the fluctuation change mode of the signal in the time domain is obtained, and the fluctuation characteristic description is determined; If some segments in the fluctuation characteristic description exceed the preset threshold range, secondary smoothing processing is performed on the corresponding segments to obtain adjusted fluctuation characteristic data; According to the adjusted fluctuation characteristic data, a signal fluctuation classification model is constructed, the category division of signal fluctuation is obtained, and the final fluctuation attribution is determined; Through the structured storage of the final fluctuation attribution result, the database management tool is used to save the classification data, and the long-term queryable signal fluctuation archive is determined. 4.The voice input method based on friction nanogenerator according to claim 1, wherein, If the signal intensity feature in the pure signal sequence exceeds the preset threshold, it is judged as a potential effective signal segment, and the time domain feature related to the pronunciation action is obtained, and the action intensity index is obtained, including: By scanning the signal intensity of the pure signal segment by segment, if it is detected that the signal intensity of a certain segment exceeds the preset threshold, it is marked as an effective signal paragraph, and a preliminary screened signal segment set is obtained; For the preliminary screened signal segment set, a feature extraction method is used to obtain time domain feature data related to pronunciation action, and the signal segment range with strong action feeling is determined; According to the signal segment range with strong action feeling, the corresponding intensity index value is calculated, the quantized intensity data of each segment is obtained, and the significance ranking thereof in the overall signal is determined; Through classification processing of the quantized intensity data after significance ranking, the signal segment is divided into different action intensity categories by using a pre-established support vector machine model, and a classified signal group is obtained; For the classified signal group, the time domain distribution mode of the signal paragraph in each category is obtained, and the correlation degree data thereof with pronunciation action is determined; According to the correlation degree data, each signal group is marked with priority, and if the correlation degree data is higher than the preset standard, it is marked as a high-priority signal segment, and a final priority processing signal set is obtained; Through the structured storage of the final priority processing signal set, the database tool is used to save the classification and priority information, and the long-term queryable signal archive data is determined. 5.The voice input method based on triboelectric nanogenerator according to claim 1, wherein, The matching of the action intensity index and the fluctuation mode analysis is used to classify the signals under the condition of dynamic environment, and an effective signal candidate set after classification is obtained, including: Through preliminary collection of signal data under the dynamic environment, a pre-established support vector machine model is used for classification processing of the signal, and a preliminary classified signal set is obtained; According to the preliminary classified signal set, the action intensity of each classification group is analyzed one by one, and if the intensity value of a certain classification group exceeds the preset threshold, it is marked as a high-intensity signal group, and a high-intensity signal group set is obtained; For the high-intensity signal group set, the fluctuation mode data in each group is obtained, the fluctuation mode is compared by a mode matching method, and the signal group range consistent with the preset mode is determined; According to the signal group range consistent with the preset mode, environmental adaptability detection is implemented, and if it is detected that the adaptability of a certain group under the dynamic environment is lower than the preset standard, it is excluded, and an adaptability qualified signal group set is obtained; The candidate screening operation is implemented for the adaptive qualified signal group set, the signal data in each group is sorted, the signal data with high ranking is obtained, and the final effective candidate signal set is determined; The final effective candidate signal set is stored in a structured manner, the classification information and intensity data are saved by using a database tool, and signal archive data that can be queried for a long time is obtained; According to the signal archive data that can be queried for a long time, a regular update mechanism is implemented, if it is detected that the dynamic environment changes, a reclassification operation of the signal set is triggered, and an updated effective candidate signal set is obtained. 6.The voice input method based on triboelectric nanogenerator according to claim 1, wherein, According to the classified effective signal candidate set, the support vector machine algorithm is used for verification of voice feature capture, the paragraph in the candidate set that truly reflects the voice action is judged, and a final effective signal segment set is determined, including: The original data record of each signal paragraph in the candidate set is obtained by data arrangement of the classified effective signal candidate set, and a preliminary signal paragraph list is determined; According to the preliminary signal paragraph list, the voice features of each signal paragraph are compared and analyzed by using a pre-established support vector machine model, and a signal paragraph grouping after feature analysis is obtained; For the signal paragraph grouping after feature analysis, the matching detection of action standard is implemented, if the feature data of a certain signal paragraph does not match the action standard, it is removed, and a signal paragraph set that meets the standard is obtained; From the signal paragraph set that meets the standard, the judgment basis data of each signal paragraph is obtained, the basis data is checked one by one, and the signal paragraph range after screening is determined; For the signal paragraph range after screening, a signal verification operation is implemented, if it is detected that a certain signal paragraph appears abnormally in the verification process, it is marked as to be processed, and a signal paragraph list that passes the verification is obtained; According to the signal paragraph list that passes the verification, the classification result data of each signal paragraph is arranged, the data is saved by using a structured storage tool, and a final signal paragraph archive is determined; Through regular detection of the final signal paragraph archive, if it is found that the signal paragraph data in the archive does not match the current environment, a data update process is triggered, and an updated signal paragraph set is obtained. 7.The voice input method based on triboelectric nanogenerator according to claim 1, wherein, If there is abnormal fluctuation related to external humidity influence in the final effective signal segment set, the fluctuation mode analysis of the paragraph is adjusted by using an adaptive filtering algorithm, and a corrected signal segment set is obtained, including: For the final effective signal segment set, the signal data related to humidity is preliminarily scanned by using a pre-set environment monitoring tool, if the scanning result shows that there is abnormal fluctuation, the signal segments are marked, and a marked signal segment list is obtained; According to the marked signal segment list, the fluctuation mode of the marked signal segment is analyzed and corrected one by one by using an adaptive filtering algorithm, and a corrected signal segment grouping is obtained; For the corrected signal segment grouping, the fluctuation mode of each signal segment is matched and detected with the pre-set standard by using a data comparison tool, if the fluctuation mode of a certain signal segment does not match the standard, it is classified as to be processed, and a signal segment set that passes the matching is obtained; According to the matching passing signal segment set, the environmental correlation data of each signal segment is obtained, the data is stored by a data arrangement tool, and a signal segment archive after arrangement is determined; According to the signal segment archive after arrangement, the signal segments in the archive are compared with current external environment data by a periodic scanning tool, if it is detected that a certain signal segment is inconsistent with the environment data, it is marked as needing to be updated, and a list of signal segments to be updated is obtained; According to the list of signal segments to be updated, the environmental data of the marked signal segments is re-acquired and recorded by a data acquisition module, and an updated signal segment data set is obtained; According to the updated signal segment data set, the update data is merged with the original signal segment archive by a data integration tool, and a final signal segment comprehensive archive is determined. 8.The voice input method based on triboelectric nanogenerator of claim 1, wherein, According to the corrected signal segment set, cluster analysis is performed on the diversity of contact modes, and a clustering grouping result is obtained, including: According to the association between the signal segment set and the correction data, the correction data is preliminarily arranged by a data screening tool, and a signal segment basic archive after arrangement is obtained; According to the signal segment basic archive after arrangement, in combination with the association between the contact mode and the diversity state, a classification tool is used to identify the mode of the signal segment, and a signal segment list after mode identification is determined; According to the signal segment list after mode identification, in combination with the association between the clustering method and the grouping result, a clustering analysis tool is used to group the signal segments, and a signal segment set after grouping is obtained; According to the signal segment set after grouping, in combination with the association between the feature analysis and the mode difference, the contact mode feature data of each group is obtained, and a signal segment directory after feature extraction is determined; According to the signal segment directory after feature extraction, in combination with the association between the signal classification and the result division, if the feature data of a certain group does not match the preset standard, it is marked, and a signal segment group after marking is obtained; According to the signal segment group after marking, in combination with the association between the data combination and the set processing, a data integration module is used to secondarily arrange the marked group, and a final signal segment classification archive is determined; According to the final signal segment classification archive, in combination with the association between the mode difference and the result division, if it is detected that the contact mode feature of a certain classification does not match the expectation, the classification label is updated by a recording tool, and an updated signal segment classification list is obtained. 9.The voice input method based on triboelectric nanogenerator of claim 1, wherein, The association between the clustering grouping result and the signal evaluation standard is compared, a k-means algorithm is used to refine the boundary of the signal in the group, an optimized effective signal segment boundary description is determined, including: According to the association between the clustering grouping and the signal data, a data screening module is used to preliminarily classify the signal data, and a classified signal data set is obtained; According to the classified signal data set, in combination with the association between the grouping result and the evaluation standard, a comparison tool is used to match the signal data set with the standard, and a matched signal data list is determined; According to the matched signal data list, in combination with the association between the signal boundary and the grouping boundary, a boundary division tool is used to mark the boundary of the signal data, and a marked signal data set is obtained; According to the labeled signal data set, in combination with the association of optimization processing and boundary description, the signal boundary is described and induced by a data arrangement module, and an induced boundary description record is obtained; According to the induced boundary description record, in combination with the association of data processing and evaluation standard, if the boundary description of a certain segment of signal data is inconsistent with the preset standard, the boundary is corrected by an adjustment tool, and a corrected signal boundary file is determined; According to the corrected signal boundary file, in combination with the association of standard comparison and optimization processing, the signal boundary file is finally arranged by an information integration module, and an arranged signal boundary data set is obtained; According to the arranged signal boundary data set, in combination with the association of signal evaluation and data processing, the signal boundary data set is stored by a recording tool, and a stored signal boundary directory is obtained. 10.A voice input system based on frictional nanogenerator, characterized in that, The system comprises: A signal acquisition and feature extraction module is used to acquire a triboelectric signal sequence, extract signal intensity features and fluctuation pattern analysis features, and obtain a preliminary signal fluctuation description; A wavelet denoising module is used to process the part of the sequence containing interference noise according to the preliminary signal fluctuation description, and determine a pure signal sequence after removing noise by using a wavelet transform algorithm; A time domain feature extraction module is used to determine whether the signal intensity feature in the pure signal sequence exceeds a preset threshold, if it exceeds the threshold, it is determined as a potential effective signal segment, and the time domain feature related to the pronunciation action is obtained, and the action intensity index is obtained; A dynamic classification module is used to classify the signals under dynamic environmental adaptation conditions by matching the action intensity index and the fluctuation pattern analysis, and obtain a classified effective signal candidate set; A support vector machine verification module is used to verify the speech feature capture by using a support vector machine algorithm according to the classified effective signal candidate set, determine the paragraph in the candidate set that truly reflects the speech action, and determine a final effective signal segment set; An adaptive filter correction module is used to determine whether there is an abnormal fluctuation related to the influence of external humidity in the final effective signal segment set, if there is an abnormality, the fluctuation pattern analysis of these paragraphs is adjusted by using an adaptive filter algorithm, and a corrected signal segment set is obtained; A clustering analysis module is used to perform clustering analysis on the diversity of contact methods according to the corrected signal segment set, and obtain a clustering grouping result; A clustering optimization module is used to compare the clustering grouping result with the signal evaluation standard, and use a k-means algorithm to refine the boundary of the signal in the group, and determine an optimized effective signal segment boundary description.

Images