Keyboard design method and system integrated with health monitoring

Abstract

The invention relates to the technical field of computer input equipment, and discloses a keyboard design method and system fused with health monitoring, and the method comprises the steps: analyzing a keystroke behavior type corresponding to keyboard operation data, and generating a physiological signal feature of a target user; setting a baseline parameter threshold of health monitoring of the target user, calculating rhythm entropy change corresponding to the keyboard rhythm data, and analyzing a potential physiological risk type corresponding to the target user; analyzing a hand fatigue accumulation track and an instantaneous operation steady state curve in the historical operation data of the user, and evaluating a hand and brain coordination health score of the target user; dynamically adjusting the baseline parameter threshold to obtain a sensitivity threshold, analyzing the current scene demand corresponding to the keystroke behavior type in real time, and generating a multi-dimensional hierarchical early warning strategy of the target user; and executing health monitoring processing on the target user to obtain a health monitoring result. According to the invention, the accuracy of health detection when a user uses the keyboard can be improved.

Description

Technical Field

[0001] The present invention relates to a keyboard design method and system integrated with health monitoring, belonging to the technical field of computer input devices. Background Art

[0002] With the gradual improvement of people's health awareness and the rapid development of smart technology, the integrated application of health monitoring technology in various devices has become an important trend. Keyboards, as the core device for computer input, are widely used in multiple scenarios such as office, study, and entertainment. In order to understand the user's health status in a timely manner, it is necessary to monitor the user's health during the use of the keyboard.

[0003] At present, health monitoring is mainly achieved with the help of wearable devices and professional medical testing equipment. Although wearable devices are easy to carry and can collect some physiological data in real time, such as heart rate and step count, these devices have relatively single functions and cannot comprehensively monitor hand behavior and physiological health status. Moreover, the monitoring data is difficult to be closely associated with specific behavioral scenarios. Although professional medical testing equipment has high detection accuracy, it is usually bulky and expensive, and its usage scenarios are limited, making it impossible to achieve daily real-time health monitoring. Some smart keyboards only implement simple function expansions, such as adding shortcut keys and adjusting backlight, and cannot realize health monitoring capabilities. Therefore, there is an urgent need for a keyboard design method that integrates health monitoring. Summary of the Invention

[0004] The present invention provides a keyboard design method and system integrating health monitoring, the main purpose of which is to improve the accuracy of health detection when a user uses the keyboard.

[0005] To achieve the above objectives, the present invention provides a keyboard design method integrating health monitoring, comprising:

[0006] Acquiring keyboard operation data and keyboard rhythm data of a target user, the keyboard operation data including key pressure distribution, acceleration timing signals, and input interval duration, analyzing the keystroke behavior type corresponding to the keyboard operation data, and collecting hand surface electromyographic signals and skin impedance change data of the target user, and combining the hand surface electromyographic signals and the skin impedance change data to generate a physiological signal signature of the target user;

[0007] Combining the keystroke behavior type and the physiological signal characteristics, setting a baseline parameter threshold for the target user's health monitoring, calculating the rhythm entropy change corresponding to the keyboard rhythm data, and constructing a hand control degradation model for the target user based on the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change. Based on the hand control degradation model, analyzing the potential physiological risk type corresponding to the target user;

[0008] Dispatching the user's historical operation data of the target user, analyzing the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user's historical operation data, and evaluating the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve;

[0009] Combined with the potential physiological risk type and the hand-brain coordination score, the baseline parameter threshold is dynamically adjusted to obtain a sensitivity threshold. The current scenario requirements corresponding to the keystroke behavior type are analyzed in real time. Combined with the preset pathological risk mapping rules, the current scenario requirements and the sensitivity threshold, a multi-dimensional hierarchical early warning strategy for the target user is generated;

[0010] Based on the multi-dimensional hierarchical early warning strategy, health monitoring processing is performed on the target user to obtain a health monitoring result.

[0011] Optionally, combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user includes:

[0012] Performing denoising processing on the hand surface electromyographic signal to obtain a denoised electromyographic signal;

[0013] Segment-processing the denoised electromyographic signal to obtain electromyographic signal segments;

[0014] Extracting time-frequency features of the electromyographic signal segments to obtain time-frequency features of the electromyographic signal;

[0015] performing baseline correction processing on the skin impedance change data to obtain corrected skin impedance change data;

[0016] extracting skin impedance features from the corrected skin impedance change data;

[0017] Performing feature alignment processing on the electromyographic signal time-frequency feature and the skin impedance feature to obtain an aligned joint feature;

[0018] Performing feature splicing on the aligned joint features to obtain a spliced joint feature;

[0019] The health status of the spliced joint features is encoded to generate physiological signal features of the target user.

[0020] Optionally, the setting of a baseline parameter threshold for health monitoring of the target user in combination with the keystroke behavior type and the physiological signal feature includes:

[0021] Quantizing and encoding the keystroke behavior type to obtain a keystroke behavior feature vector;

[0022] performing standardization processing on the physiological signal characteristics to obtain standard physiological signal characteristics;

[0023] Calculating feature similarity between the keystroke behavior feature vector and the standard physiological signal feature;

[0024] When the feature similarity is greater than a preset similarity, performing feature fusion processing on the keystroke behavior feature vector and the standard physiological signal feature to obtain a fused feature vector;

[0025] A baseline parameter threshold for health monitoring of the target user is set based on the fused feature vector.

[0026] Optionally, the step of constructing a hand control degradation model for the target user by combining the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change includes:

[0027] Performing sliding window processing on the key position pressure distribution to obtain key position pressure sliding characteristics;

[0028] Calculating the dynamic balance of the key position pressure distribution based on the key position pressure sliding characteristics;

[0029] performing signal decomposition processing on the acceleration time series signal to obtain a frequency sub-band signal;

[0030] extracting a tremor frequency subband signal from the frequency subband signals, and calculating the tremor spectrum energy of the acceleration time series signal based on the tremor frequency subband signal;

[0031] The timing fluctuation characteristics of the input interval duration are analyzed, and a hand control degradation model of the target user is constructed by combining the timing fluctuation characteristics, the dynamic balance, the tremor spectrum energy, and the rhythm entropy change.

[0032] Optionally, the calculating the dynamic balance of the key pressure distribution based on the key pressure sliding characteristics includes:

[0033] The characteristic standard deviation corresponding to the key position pressure sliding characteristic is calculated, and based on the characteristic standard deviation, the dynamic balance degree of the key position pressure distribution is calculated by the following formula:

[0034]

[0035] Among them, A represents the dynamic balance of key pressure distribution, σ a represents the characteristic standard deviation corresponding to the ath feature in the key pressure sliding feature, a represents the serial number of the key pressure sliding feature, q represents the number of key pressure sliding features, max(σ a ,σ a+1 ,…,σ q ) represents the function of selecting the maximum value of the characteristic standard deviation.

[0036] Optionally, calculating the tremor spectrum energy of the acceleration time series signal based on the tremor frequency subband signal includes:

[0037] Performing Fourier transform processing on the tremor frequency subband signal to obtain a tremor subband signal spectrum;

[0038] performing a spectrum modulo operation on the spectrum of the tremor subband signal to obtain a tremor spectrum amplitude;

[0039] Calculating the spectrum energy density corresponding to the tremor frequency subband signal based on the tremor spectrum amplitude;

[0040] A tremor frequency interval corresponding to the tremor frequency subband signal is determined, and an integration operation is performed on the spectrum energy density within the tremor frequency interval to obtain tremor spectrum energy of the acceleration time series signal.

[0041] Optionally, analyzing the timing fluctuation characteristics of the input interval duration includes:

[0042] Calculating timestamp differences between adjacent intervals in the input interval duration, and determining an input interval sequence corresponding to the input interval duration based on the timestamp differences;

[0043] Performing window segmentation processing on the input interval sequence to obtain a window interval sequence;

[0044] Calculate the sequence mean corresponding to the window interval sequence, and calculate the sequence variation coefficient corresponding to the window interval sequence according to the sequence mean using the following formula:

[0045]

[0046] Among them, B represents the sequence variation coefficient corresponding to the window interval sequence, F b represents the sequence mean of the bth sequence in the window interval sequence, Δt be represents the e-th time series within the b-th sequence in the window interval sequence, b represents the sequence number of the window interval sequence, e represents the sequence number of the time series in the window interval sequence, and R represents the number of window interval sequences;

[0047] Based on the sequence variation coefficient, the time series fluctuation characteristics of the input interval duration are analyzed.

[0048] Optionally, the step of evaluating the target user's hand-brain coordination score by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve includes:

[0049] Performing trajectory smoothing on the fatigue accumulation trajectory to obtain a smoothed fatigue accumulation trajectory;

[0050] extracting morphological parameters from the smooth fatigue accumulation trajectory, and calculating the muscle load aging coefficient of the target user based on the morphological parameters;

[0051] Calculating a control response attenuation rate of the target user according to the stability curve;

[0052] The hand-brain coordination score of the target user is evaluated by combining the muscle load aging coefficient and the control response attenuation rate.

[0053] Optionally, calculating the target user's muscle load aging coefficient based on the morphological parameters includes:

[0054] Mining the morphological dimension features corresponding to the morphological parameters, performing feature screening processing on the morphological dimension features, and obtaining target morphological dimension features;

[0055] Analyzing the independent contribution of the morphological dimension features to muscle load aging to obtain a contribution set;

[0056] Based on the contribution degree set, allocating the dimension importance corresponding to the morphological dimension feature;

[0057] The muscle load aging coefficient of the target user is calculated by combining the dimension importance and the morphological dimension characteristics.

[0058] In order to solve the above problems, the present invention also provides a keyboard design system integrating health monitoring, the system comprising:

[0059] a physiological signal feature generation module, configured to obtain keyboard operation data and keyboard rhythm data of a target user, the keyboard operation data including key pressure distribution, acceleration timing signals, and input interval duration, analyze the keystroke behavior type corresponding to the keyboard operation data, and collect hand surface electromyography signals and skin impedance change data of the target user, and combine the hand surface electromyography signals and the skin impedance change data to generate a physiological signal feature of the target user;

[0060] a potential physiological risk type analysis module, configured to set a baseline parameter threshold for the target user's health monitoring based on the keystroke behavior type and the physiological signal characteristics, calculate the rhythm entropy change corresponding to the keyboard rhythm data, construct a hand control degradation model for the target user based on the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change, and analyze the potential physiological risk type corresponding to the target user based on the hand control degradation model;

[0061] a hand-brain coordination scoring and evaluation module, configured to schedule the user's historical operation data of the target user, analyze the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user's historical operation data, and evaluate the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve;

[0062] A multi-dimensional hierarchical warning strategy generation module is used to dynamically adjust the baseline parameter threshold based on the potential physiological risk type and the hand-brain coordination score to obtain a sensitivity threshold, analyze the current scenario requirements corresponding to the keystroke behavior type in real time, and generate a multi-dimensional hierarchical warning strategy for the target user based on preset pathological risk mapping rules, the current scenario requirements, and the sensitivity threshold;

[0063] The health monitoring module is used to perform health monitoring processing on the target user based on the multi-dimensional layered early warning strategy to obtain health monitoring results.

[0064] Compared with the problems described in the background technology, the present invention can accurately grasp the user's operating habits by analyzing the keystroke behavior type corresponding to the keyboard operation data, providing a basis for subsequent comprehensive analysis, and combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user, which can further enrich the understanding of the user status from a physiological level and provide more comprehensive data support for applications in multiple fields. Furthermore, the present invention sets the baseline parameter threshold for the health monitoring of the target user by combining the keystroke behavior type and the physiological signal characteristics, and can tailor personalized health assessment benchmarks for the unique operating behaviors and physiological characteristics of different users, greatly improving the accuracy of abnormal state identification during health monitoring and reducing the occurrence of misjudgment and missed judgment. The present invention can understand the fatigue evolution and real-time state of the target user during hand operation by analyzing the hand fatigue accumulation trajectory and instantaneous operation steady-state curve in the user's historical operation data. State stability, combined with the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve, evaluates the hand-brain coordination score of the target user, which can understand the health level of the user's hand-brain coordination, and provide a scientific and quantitative basis for preventing hand strain, optimizing the human-computer interaction experience, and diagnosing potential nervous system or motor system abnormalities. Furthermore, the present invention dynamically adjusts the baseline parameter threshold by combining the potential physiological risk type and the hand-brain coordination score to obtain a sensitivity threshold, so that the threshold setting is more in line with the user's individual physiological health status, which can improve the perception accuracy of changes in the user's physiological state and ensure the sensitivity and accuracy of subsequent early warning mechanisms. Furthermore, the present invention performs health monitoring processing on the target user based on the multi-dimensional hierarchical early warning strategy, thereby realizing accurate analysis of the health risks of the target user, ensuring that an effective early warning is issued at the initial stage of the risk, assisting the user to take timely intervention measures, reducing the chance of illness, and comprehensively protecting the user's physical and mental health. Therefore, the keyboard design method and system for integrating health monitoring provided in the embodiment of the present invention can improve the accuracy of health detection when the user uses the keyboard. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 A flowchart of a keyboard design method integrating health monitoring provided by one embodiment of the present invention;

[0066] Figure 2 A schematic diagram of a module for implementing the keyboard design method for integrating health monitoring provided by an embodiment of the present invention.

[0067] The purpose, features and advantages of the present invention will be further described with reference to the accompanying drawings and in conjunction with the embodiments. DETAILED DESCRIPTION

[0068] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0069] The embodiment of the present application provides a keyboard design method that integrates health monitoring. The execution subject of the keyboard design method that integrates health monitoring includes but is not limited to at least one of the electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiment of the present application. In other words, the keyboard design method that integrates health monitoring can be executed by software or hardware installed on a terminal device or a server device. The server includes but is not limited to: a single server, a server cluster, a cloud server or a cloud server cluster, etc.

[0070] Example 1:

[0071] Reference Figure 1 FIG. 1 is a flow chart of a keyboard design method integrating health monitoring according to an embodiment of the present invention. In this embodiment, the keyboard design method integrating health monitoring includes:

[0072] S1. Obtain keyboard operation data and keyboard rhythm data of the target user, wherein the keyboard operation data includes key pressure distribution, acceleration timing signal and input interval duration, analyze the keystroke behavior type corresponding to the keyboard operation data, and collect the surface electromyography signal and skin impedance change data of the target user's hand, and combine the surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user.

[0073] By analyzing the keystroke behavior type corresponding to the keyboard operation data, the present invention can accurately grasp the user's operating habits, providing a basis for subsequent comprehensive analysis, and combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user, which can further enrich the understanding of the user status from a physiological level and provide more comprehensive data support for applications in multiple fields.

[0074] Among them, the target users include people of different ages and occupations, who will generate personal unique operation data and physiological signals when using the keyboard. For example, due to the high frequency of keyboard use, the keystroke behavior and hand physiological state of writers will show different characteristics from those of ordinary users; the keyboard operation data is the key information reflecting the user's keyboard input behavior, the key pressure distribution reflects the difference in the force applied by the user when hitting different keys, the acceleration timing signal records the change in speed during each keystroke, and the input interval duration shows the time interval characteristics of the user when inputting characters; the keystroke behavior type refers to the user's keystroke habit category summarized based on the keyboard operation data, such as fast and continuous input, slow and cautious input, etc., and the hand The surface electromyographic signal is a signal generated by the electrical activity of the hand muscles of the target user when performing keyboard operations. The skin impedance change data is data on the change of the electrical impedance of the hand skin of the target user over time during the keyboard operation. Furthermore, the analysis of the keystroke behavior type corresponding to the keyboard operation data can be achieved through machine learning algorithms, such as using support vector machines (SVM), decision trees and other algorithms to train and classify data such as key pressure distribution, acceleration timing signals and input interval duration to determine different keystroke behavior types; the surface electromyographic signal and skin impedance change data of the target user's hand can be achieved through physiological signal acquisition equipment, such as using electromyographic sensors to collect surface electromyographic signals of the hand, and using skin impedance meters to obtain skin impedance change data.

[0075] As an embodiment of the present invention, combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user includes:

[0076] Performing denoising processing on the hand surface electromyographic signal to obtain a denoised electromyographic signal;

[0077] Segment-processing the denoised electromyographic signal to obtain electromyographic signal segments;

[0078] Extracting time-frequency features of the electromyographic signal segments to obtain time-frequency features of the electromyographic signal;

[0079] performing baseline correction processing on the skin impedance change data to obtain corrected skin impedance change data;

[0080] extracting skin impedance features from the corrected skin impedance change data;

[0081] Performing feature alignment processing on the electromyographic signal time-frequency feature and the skin impedance feature to obtain an aligned joint feature;

[0082] Performing feature splicing on the aligned joint features to obtain a spliced joint feature;

[0083] The health status of the spliced joint features is encoded to generate physiological signal features of the target user.

[0084] Among them, the denoised electromyographic signal is the output of the hand surface electromyographic signal after filtering and other denoising processes, the electromyographic signal segment is the part of the denoised electromyographic signal divided according to time or amplitude standards, the electromyographic signal time-frequency feature is the product of time-frequency analysis of the electromyographic signal segment, the corrected skin impedance change data is the result of baseline correction of the skin impedance change data, the skin impedance feature is the key data obtained from the corrected skin impedance change data through feature extraction technology, the aligned joint feature is a feature group formed after matching the electromyographic signal time-frequency feature and the skin impedance feature in time or dimension, and the spliced joint feature is a new feature set obtained by serial or merging the aligned joint features.

[0085] Furthermore, the hand surface electromyographic signal can be denoised by filtering tools such as bandpass filters to obtain denoised electromyographic signals; the denoised electromyographic signal can be segmented according to rules such as fixed time intervals or signal amplitude thresholds to obtain electromyographic signal segments; the electromyographic signal segments can be subjected to time-frequency feature extraction by means of time-frequency analysis algorithms such as short-time Fourier transform and wavelet transform to obtain electromyographic signal time-frequency features; the skin impedance change data can be baseline corrected by using a reference signal comparison method or a baseline estimation method based on statistics to obtain corrected skin impedance change data; the skin impedance features in the corrected skin impedance change data can be extracted by feature extraction methods such as mean, standard deviation calculation and principal component analysis; the skin impedance features in the corrected skin impedance change data can be extracted by time scale matching, feature dimension The time-frequency features of the electromyographic signal and the skin impedance features are aligned with each other by alignment and other methods to obtain aligned joint features; the aligned joint features can be spliced by horizontal splicing, vertical splicing and other methods to obtain spliced joint features; the health status of the spliced joint features can be encoded by constructing a health status assessment model to generate the physiological signal features of the target user, such as constructing a classification model based on a neural network, taking the spliced joint features as input, and outputting coding labels for different health states, such as "healthy", "mild fatigue", "severe fatigue", etc. after model training and learning; or constructing a regression model based on a support vector machine to map the spliced joint features to a continuous health status score interval, so as to quantitatively encode the physiological state of the target user.

[0086] S2. In combination with the keystroke behavior type and the physiological signal characteristics, set the baseline parameter threshold for the target user's health monitoring, calculate the rhythm entropy change corresponding to the keyboard rhythm data, and construct a hand control degradation model for the target user in combination with the key pressure distribution, the acceleration timing signal, the input interval duration and the rhythm entropy change. Based on the hand control degradation model, analyze the potential physiological risk type corresponding to the target user.

[0087] The present invention sets the baseline parameter threshold for the target user's health monitoring by combining the keystroke behavior type and the physiological signal characteristics. This can tailor personalized health assessment benchmarks based on the unique operating behaviors and physiological characteristics of different users, greatly improving the accuracy of abnormal state identification during health monitoring and reducing the occurrence of misjudgments and missed judgments. Among them, the baseline parameter threshold is a personalized reference standard for the target user's health monitoring, which is set based on the user's personal keystroke behavior and physiological signal characteristics, and is used to determine whether his or her health status deviates from the normal range.

[0088] As an optional embodiment of the present invention, the step of setting a baseline parameter threshold for health monitoring of the target user in combination with the keystroke behavior type and the physiological signal feature includes:

[0089] Quantizing and encoding the keystroke behavior type to obtain a keystroke behavior feature vector;

[0090] performing standardization processing on the physiological signal characteristics to obtain standard physiological signal characteristics;

[0091] Calculating feature similarity between the keystroke behavior feature vector and the standard physiological signal feature;

[0092] When the feature similarity is greater than a preset similarity, performing feature fusion processing on the keystroke behavior feature vector and the standard physiological signal feature to obtain a fused feature vector;

[0093] A baseline parameter threshold for health monitoring of the target user is set based on the fused feature vector.

[0094] Among them, the keystroke behavior feature vector is a digital vector representation obtained after the keystroke behavior type is quantized and encoded; the standard physiological signal feature is a feature obtained after the physiological signal feature is standardized to eliminate the dimension and numerical range differences; the feature similarity represents the degree of similarity between the keystroke behavior feature vector and the standard physiological signal feature in numerical value and feature space; the fused feature vector is a new feature vector formed by merging the keystroke behavior feature vector and the standard physiological signal feature through a specific method (such as weighted summation, etc.).

[0095] Furthermore, the keystroke behavior type can be quantized and encoded by the one-hot encoding method to obtain a keystroke behavior feature vector; the physiological signal feature can be standardized by the Z-score normalization method to obtain a standard physiological signal feature; the feature similarity between the keystroke behavior feature vector and the standard physiological signal feature can be calculated by the cosine similarity algorithm; when the feature similarity is greater than the preset similarity, the keystroke behavior feature vector and the standard physiological signal feature are feature fused by weighted fusion and other methods to obtain a fused feature vector; based on the fused feature vector, the baseline parameter threshold of the target user health monitoring can be set by means of statistical analysis, machine learning modeling and other means. For example, when using statistical analysis means, the mean and standard deviation of the fused feature vector can be calculated, and a reasonable interval can be determined as the baseline parameter threshold based on the standard deviation with the mean as the center. For example, for the fusion features of keystroke speed and hand electromyography, if the mean is 50 and the standard deviation is 5, 45-55 can be set as the preliminary baseline parameter threshold.

[0096] The present invention can understand the complexity and dynamic changes of the keyboard operation rhythm by calculating the rhythm entropy change corresponding to the keyboard rhythm data, and quantify the stability of the operation behavior with the help of the entropy value, thereby laying an important basis for subsequent processing. The rhythm entropy change reflects the complexity and degree of change of the keyboard rhythm data. The higher the entropy value, the more unstable the rhythm of the keyboard operation, indicating an abnormal change in the user's physical state. Furthermore, the calculation of the rhythm entropy change corresponding to the keyboard rhythm data can be achieved through the Shannon entropy algorithm. The rhythm frequency corresponding to the keyboard rhythm data in each time window is first counted, and the rhythm frequency is substituted into the Shannon entropy algorithm to calculate the rhythm entropy change corresponding to the keyboard rhythm data.

[0097] The present invention constructs a hand control degradation model for the target user by combining the key pressure distribution, the acceleration timing signal, the input interval duration and the rhythm entropy change. This can comprehensively simulate the changes in the target user's hand control ability and provide a reliable model support for the analysis of potential physiological risks. The hand control degradation model is a mathematical representation of the target user's hand operation control ability changing over time, which is a model that describes the degradation trend of the user's hand motion control function.

[0098] As an embodiment of the present invention, the step of constructing the target user's hand control degradation model by combining the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change includes:

[0099] Performing sliding window processing on the key position pressure distribution to obtain key position pressure sliding characteristics;

[0100] Calculating the dynamic balance of the key position pressure distribution based on the key position pressure sliding characteristics;

[0101] performing signal decomposition processing on the acceleration time series signal to obtain a frequency sub-band signal;

[0102] extracting a tremor frequency subband signal from the frequency subband signals, and calculating the tremor spectrum energy of the acceleration time series signal based on the tremor frequency subband signal;

[0103] The timing fluctuation characteristics of the input interval duration are analyzed, and a hand control degradation model of the target user is constructed by combining the timing fluctuation characteristics, the dynamic balance, the tremor spectrum energy, and the rhythm entropy change.

[0104] Among them, the key pressure sliding feature is the characteristic data obtained after the key pressure distribution is processed by the sliding window, the dynamic balance is an indicator used to measure the degree of balance of the key pressure distribution in the dynamic process, the frequency sub-band signal is the signal of different frequency ranges obtained after the acceleration time series signal is frequency decomposed, the vibration frequency sub-band signal is the signal corresponding to the vibration-related frequency in the frequency sub-band signal, the vibration spectrum energy is the energy distribution of the acceleration time series signal at the vibration-related frequency, and the time series fluctuation feature is the fluctuation characteristic of the input interval duration in the time series.

[0105] Furthermore, the key pressure distribution can be subjected to sliding window processing by setting a specific sliding window size and step size, and calculating the key pressure mean, variance and other statistical quantities within the window to obtain key pressure sliding characteristics; the acceleration timing signal can be subjected to signal decomposition processing by wavelet transform to obtain frequency sub-band signals; the tremor frequency sub-band signal in the frequency sub-band signal can be extracted by band-pass filtering based on a predetermined tremor frequency range; the hand control degradation model of the target user can be constructed by combining the timing fluctuation characteristics, the dynamic balance, the tremor spectrum energy and the rhythm entropy change, and the construction steps are: standardizing the above-mentioned multivariate feature data, using principal component analysis to screen key features, training the model with the help of machine learning algorithms such as support vector machines and random forests, and optimizing the model parameters through cross-validation to determine the final hand control degradation model.

[0106] Furthermore, as an optional embodiment of the present invention, the calculating the dynamic balance of the key pressure distribution based on the key pressure sliding characteristics includes:

[0107] The characteristic standard deviation corresponding to the key position pressure sliding characteristic is calculated, and based on the characteristic standard deviation, the dynamic balance degree of the key position pressure distribution is calculated by the following formula:

[0108]

[0109] Among them, A represents the dynamic balance of key pressure distribution, σ a represents the characteristic standard deviation corresponding to the ath feature in the key pressure sliding feature, a represents the serial number of the key pressure sliding feature, q represents the number of key pressure sliding features, max(σ a ,σ a+1 ,…,σ q ) represents the function of selecting the maximum value of the characteristic standard deviation.

[0110] The characteristic standard deviation is a measure of the degree of dispersion corresponding to the key position pressure sliding feature. Furthermore, the characteristic standard deviation corresponding to the key position pressure sliding feature can be calculated using a standard deviation formula.

[0111] Furthermore, as an optional embodiment of the present invention, calculating the tremor spectrum energy of the acceleration time series signal based on the tremor frequency subband signal includes:

[0112] Performing Fourier transform processing on the tremor frequency subband signal to obtain a tremor subband signal spectrum;

[0113] performing a spectrum modulo operation on the spectrum of the tremor subband signal to obtain a tremor spectrum amplitude;

[0114] Calculating the spectrum energy density corresponding to the tremor frequency subband signal based on the tremor spectrum amplitude;

[0115] A tremor frequency interval corresponding to the tremor frequency subband signal is determined, and an integration operation is performed on the spectrum energy density within the tremor frequency interval to obtain tremor spectrum energy of the acceleration time series signal.

[0116] Among them, the tremor subband signal spectrum is the frequency domain representation obtained after Fourier transform processing of the tremor frequency subband signal; the tremor spectrum amplitude is the signal strength value at each frequency point obtained after spectral modulo operation of the tremor subband signal spectrum; the spectral energy density is the value of the squaring of the spectral amplitude corresponding to the tremor frequency subband signal, which reflects the energy distribution of the signal at different frequencies; the tremor frequency interval is a predetermined interval corresponding to the tremor frequency subband signal that includes the target tremor frequency range.

[0117] Furthermore, the tremor frequency subband signal can be subjected to Fourier transform processing using a fast Fourier transform algorithm to obtain a tremor subband signal spectrum; a spectral modulus operation can be performed on the tremor subband signal spectrum using a complex modulus method to obtain a tremor spectrum amplitude; based on the tremor spectrum amplitude, the spectral energy density corresponding to the tremor frequency subband signal can be calculated by squaring it; the tremor frequency interval corresponding to the tremor frequency subband signal can be determined by analyzing historical data or based on relevant field knowledge; and the spectral energy density can be integrated within the tremor frequency interval using a numerical integration algorithm to obtain the tremor spectrum energy of the acceleration time series signal.

[0118] Furthermore, as an optional embodiment of the present invention, analyzing the timing fluctuation characteristics of the input interval duration includes:

[0119] Calculating timestamp differences between adjacent intervals in the input interval duration, and determining an input interval sequence corresponding to the input interval duration based on the timestamp differences;

[0120] Performing window segmentation processing on the input interval sequence to obtain a window interval sequence;

[0121] Calculate the sequence mean corresponding to the window interval sequence, and calculate the sequence variation coefficient corresponding to the window interval sequence according to the sequence mean using the following formula:

[0122]

[0123] Among them, B represents the sequence variation coefficient corresponding to the window interval sequence, F b represents the sequence mean of the bth sequence in the window interval sequence, Δt be represents the e-th time series within the b-th sequence in the window interval sequence, b represents the sequence number of the window interval sequence, e represents the sequence number of the time series in the window interval sequence, and R represents the number of window interval sequences;

[0124] Based on the sequence variation coefficient, the time series fluctuation characteristics of the input interval duration are analyzed.

[0125] Among them, the timestamp difference is the time difference between adjacent intervals in the input interval duration; the input interval sequence is a set of time intervals arranged in sequence corresponding to the input interval duration; and the window interval sequence is a set of interval sequences with a fixed window size and sliding step size obtained after window segmentation processing of the input interval sequence.

[0126] Furthermore, based on the timestamp difference, the input interval sequence corresponding to the input interval duration is determined by arranging the calculated timestamp difference in sequence; the input interval sequence can be window-segmented by setting the window size and sliding step size, and the input interval sequence fragments can be intercepted one by one according to the set parameters to obtain a window interval sequence; based on the sequence variation coefficient, the time series fluctuation characteristics of the input interval duration are analyzed, and based on the sequence variation coefficient, the permutation entropy is first calculated through time delay reconstruction, symbolic coding, and statistical pattern probability to quantify the degree of disorder of the input interval duration fluctuation; then the permutation entropy is compared with the baseline value to calculate the fluctuation intensity quantification value, and based on this, whether the time series fluctuation of the input interval duration is abnormal is analyzed, thereby obtaining the time series fluctuation characteristics of the input interval duration.

[0127] The present invention analyzes the potential physiological risk types corresponding to the target user based on the hand control degradation model, and then determines the possible health problems of the target user, providing an important basis for subsequent analysis and processing. The potential physiological risk types include hand muscle fatigue, early symptoms of nervous system diseases, excessive psychological stress, etc. The analysis results of the hand control degradation model can preliminarily screen out the risk types that the user may face. Furthermore, based on the hand control degradation model, the step of analyzing the potential physiological risk types corresponding to the target user is as follows: inputting the real-time keyboard operation data of the target user into the hand control degradation model, simulating the predicted value of the hand control degradation degree of the target user, comparing the predicted value with the preset risk threshold, and determining whether the hand control degradation degree exceeds the normal range; determining the potential physiological risk type according to the degree and direction of the excess range, combined with medical knowledge and clinical experience.

[0128] S3. Dispatching the user historical operation data of the target user, and analyzing the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user historical operation data, and evaluating the hand-brain coordination score of the target user in combination with the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve.

[0129] The present invention can understand the fatigue evolution and real-time state stability of the target user's hand operation process by analyzing the hand fatigue accumulation trajectory and instantaneous operation steady-state curve in the user's historical operation data. Combined with the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve, the hand-brain coordination score of the target user is evaluated to understand the health level of the user's hand-brain coordination, and provide a scientific and quantitative basis for preventing hand strain, optimizing human-computer interaction experience, and diagnosing potential abnormalities of the nervous system or motor system. Among them, the user historical operation data is a series of recorded information generated by the target user in the past when using related equipment or performing specific operations. The hand fatigue accumulation trajectory and the instantaneous operation steady-state curve are combined to evaluate the hand-brain coordination score of the target user, thereby understanding the health level of the user's hand-brain coordination. The operation steady-state curves are specific data curves in the user's historical operation data that can reflect the hand fatigue accumulation process and the operation stable state at a certain moment. The hand-brain coordination score is a numerical value obtained by the target user after analyzing and evaluating his hand operation data to measure the health of hand-brain coordination. Furthermore, the user historical operation data of the target user can be scheduled by connecting to a user data storage database and using SQL query statements or specific API interfaces; outliers can be removed through data cleaning, and the hand fatigue accumulation trajectory and instantaneous operation steady-state curve in the user's historical operation data can be analyzed from the cleaned data using time series analysis algorithms and statistical calculation methods.

[0130] As an optional embodiment of the present invention, the evaluating the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve includes:

[0131] Performing trajectory smoothing on the fatigue accumulation trajectory to obtain a smoothed fatigue accumulation trajectory;

[0132] extracting morphological parameters from the smooth fatigue accumulation trajectory, and calculating the muscle load aging coefficient of the target user based on the morphological parameters;

[0133] Calculating a control response attenuation rate of the target user according to the stability curve;

[0134] The hand-brain coordination score of the target user is evaluated by combining the muscle load aging coefficient and the control response attenuation rate.

[0135] Among them, the smooth fatigue accumulation trajectory is the trajectory obtained after the fatigue accumulation trajectory is smoothed; the morphological parameters are quantitative parameters in the smooth fatigue accumulation trajectory used to describe its shape, size and other characteristics; the muscle load aging coefficient is a numerical value determined based on the morphological parameters, reflecting the degree of aging caused by the target user's muscle load changing over time; the control response attenuation rate represents the degree to which the target user's response gradually weakens over time under neural control.

[0136] Furthermore, the fatigue accumulation trajectory can be smoothed by a Gaussian filtering algorithm to obtain a smooth fatigue accumulation trajectory; the morphological parameters in the smooth fatigue accumulation trajectory can be extracted by a geometric calculation method; the control response attenuation rate of the target user can be calculated according to the stability curve with the help of a slope analysis method; the muscle load aging coefficient and the control response attenuation rate are normalized respectively, and the normalized values are added together, and the hand-brain coordination score of the target user is evaluated according to the result of the addition, such as: when the addition result is in the interval [0, 0.2], it is determined that the hand-brain coordination score of the target user is in an excellent state. When the sum is in the range of (0.2, 0.5], the target user's hand-brain coordination score is at a good level, which means that the hand-brain coordination state is acceptable and there is room for improvement; when the sum is in the range of (0.5, 0.8], the target user's hand-brain coordination score is at a medium level, which means that there are some problems with hand-brain coordination and muscle load or neural control response may need attention; when the sum is in the range of (0.8, 1], the target user's hand-brain coordination score is at a poor level, which indicates that the hand-brain coordination state is not good.

[0137] Furthermore, as an optional embodiment of the present invention, the calculating the muscle load aging coefficient of the target user based on the morphological parameters includes:

[0138] Mining the morphological dimension features corresponding to the morphological parameters, performing feature screening processing on the morphological dimension features, and obtaining target morphological dimension features;

[0139] Analyzing the independent contribution of the morphological dimension features to muscle load aging to obtain a contribution set;

[0140] Based on the contribution degree set, allocating the dimension importance corresponding to the morphological dimension feature;

[0141] The muscle load aging coefficient of the target user is calculated by combining the dimension importance and the morphological dimension characteristics.

[0142] Among them, the morphological dimension feature is a feature subset corresponding to the morphological parameters. The morphological parameters contain information in multiple aspects. The morphological dimension feature is a feature related to the spatial dimension mined from these parameters. It is a specific subset of morphological parameters, focusing on describing the spatial characteristics of the morphology; the target morphological dimension feature is the key part obtained after the morphological dimension feature is screened. When analyzing the morphological dimension feature, not all features are equally important for calculating the muscle load aging coefficient. Through feature screening, those features that have little impact on the results or are redundant are removed. The remaining target morphological dimension features are the key parts for calculating the muscle load aging coefficient; the contribution set is the contribution of the morphological dimension feature to the muscle load aging. Quantitative representation of the degree of independent effect. Each morphological dimension feature has a certain influence on muscle load aging. The contribution set is a set of numerical values used to represent the degree to which each morphological dimension feature individually affects muscle load aging, reflecting the relative importance of each feature in influencing muscle load aging. The dimension importance is the weight corresponding to the morphological dimension feature after comprehensive consideration. Based on the contribution set, the relationship between the various morphological dimension features and their importance in the overall calculation are further considered, and a dimension importance is assigned to each morphological dimension feature. This dimension importance is a weight value after comprehensive consideration, which is used to reflect the relative importance of each feature when calculating the muscle load aging coefficient, so as to more accurately reflect the influence of morphological dimension features on muscle load aging.

[0143] Furthermore, the morphological dimension features corresponding to the morphological parameters can be mined through spatial geometry algorithms and machine learning feature extraction models; the morphological dimension features can be subjected to feature screening processing through a correlation analysis screening method to obtain target morphological dimension features, and the correlation coefficient between each morphological dimension feature and the muscle load aging index, such as the Pearson correlation coefficient, is calculated. A correlation threshold is set, and features with a correlation with the muscle load aging index higher than the threshold are retained, and features with a low correlation are removed, thereby obtaining target morphological dimension features; the independent contribution of the morphological dimension features to muscle load aging can be analyzed by constructing a regression model to obtain a contribution set, and the muscle load aging index is used as the dependent variable. The morphological dimension features are introduced one by one to construct a regression model, and each time a feature is introduced, the changes in indicators such as the model's goodness of fit and coefficient significance are observed to evaluate the independent contribution of the feature to muscle load aging; based on the contribution set, the dimensional importance corresponding to the morphological dimension features is allocated through the hierarchical analysis method, and a hierarchical structure model is constructed, with the morphological dimension features as the criterion layer. The relative importance of each feature is determined based on the contribution set, and a judgment matrix is constructed. The eigenvector and the maximum eigenvalue of the judgment matrix are calculated to perform a consistency test. If the test passes, the eigenvector is the dimension importance corresponding to each morphological dimension feature; the muscle load aging coefficient of the target user is calculated by weighted summation based on the dimension importance and the morphological dimension feature.

[0144] S4. Combine the potential physiological risk type and the hand-brain coordination score, dynamically adjust the baseline parameter threshold to obtain the sensitivity threshold, analyze the current scenario requirements corresponding to the keystroke behavior type in real time, and combine the preset pathological risk mapping rules, the current scenario requirements and the sensitivity threshold to generate a multi-dimensional hierarchical early warning strategy for the target user.

[0145] The present invention dynamically adjusts the baseline parameter threshold by combining the potential physiological risk type and the hand-brain cooperation score to obtain a sensitivity threshold, so that the threshold setting is more in line with the individual physiological health status of the user, can improve the perception accuracy of changes in the user's physiological state, and ensure the sensitivity and accuracy of the subsequent early warning mechanism. Among them, the sensitivity threshold is the baseline parameter threshold combined with the potential physiological risk type and the hand-brain cooperation score, and is targetedly adjusted to obtain a key value to adapt to the user's real-time health status. Furthermore, the steps of dynamically adjusting the baseline parameter threshold are: first, different risk level intervals are divided according to the potential physiological risk type, and the hand-brain cooperation score is also divided into different health level intervals. Then, a risk-health level matrix is constructed, and the rows of the matrix represent different risk levels and the columns represent different health levels. Next, a corresponding adjustment coefficient is pre-set for each cell in the matrix. After that, the target user's risk level range is determined according to their potential physiological risk type, and their health level range is determined according to their hand-brain cooperation score, so as to locate the corresponding cell in the matrix and obtain the adjustment coefficient of the cell. Finally, the adjustment coefficient is multiplied by the baseline parameter threshold to obtain the dynamically adjusted sensitivity threshold.

[0146] The present invention can obtain the actual context and needs of the user when using the device by real-time analysis of the current scenario needs corresponding to the keystroke behavior type, and generate a multi-dimensional hierarchical warning strategy for the target user by combining the preset pathological risk mapping rules, the current scenario needs and the sensitivity threshold. It can provide comprehensive and orderly warnings for the physiological risks that the user may face from multiple dimensions and different levels, helping the user to understand their own health status in a timely manner and take corresponding preventive and intervention measures. Among them, the current scenario needs are the actual needs of the user at the current moment inferred by combining the keystroke behavior type and the device usage scenario information. For example, in a fast typing scenario, the user may prefer the system to reduce Unnecessary interference warnings, the multi-dimensional layered warning strategy is a multi-level, multi-dimensional warning plan formulated according to different risk levels, scenario requirements and individual user characteristics. Furthermore, the real-time analysis of the current scenario requirements corresponding to the keystroke behavior type can adopt behavioral pattern recognition algorithms, such as hidden Markov models, support vector machines, etc., through the learning of historical keystroke behavior data and scenario information, to achieve accurate inference of current scenario requirements; the generation of multi-dimensional layered warning strategies can be assisted by rule engine tools, such as Drools, Jess, etc., according to the preset pathological risk mapping rules, combined with the current scenario requirements and sensitivity thresholds, to automatically generate corresponding warning strategies.

[0147] S5. Based on the multi-dimensional layered early warning strategy, perform health monitoring processing on the target user to obtain a health monitoring result.

[0148] The present invention performs health monitoring processing on the target user based on the multi-dimensional hierarchical early warning strategy, thereby realizing accurate analysis of the health risks of the target user, ensuring that effective early warnings are issued at the initial stage of the risk, assisting users to take timely intervention measures, reducing the chance of illness, and comprehensively protecting the user's physical and mental health. The health monitoring results are obtained by real-time analysis of relevant data such as the target user's keystroke behavior, and based on the multi-dimensional hierarchical early warning strategy, evaluation information on the user's current health status is obtained, including whether there are potential health risks and the level of the risk.

[0149] Compared with the problems described in the background technology, the present invention can accurately grasp the user's operating habits by analyzing the keystroke behavior type corresponding to the keyboard operation data, providing a basis for subsequent comprehensive analysis, and combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user, which can further enrich the understanding of the user status from a physiological level and provide more comprehensive data support for applications in multiple fields. Furthermore, the present invention sets the baseline parameter threshold for the health monitoring of the target user by combining the keystroke behavior type and the physiological signal characteristics, and can tailor personalized health assessment benchmarks for the unique operating behaviors and physiological characteristics of different users, greatly improving the accuracy of abnormal state identification during health monitoring and reducing the occurrence of misjudgment and missed judgment. The present invention can understand the fatigue evolution and real-time state of the target user during hand operation by analyzing the hand fatigue accumulation trajectory and instantaneous operation steady-state curve in the user's historical operation data. State stability, combined with the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve, evaluates the hand-brain coordination score of the target user, which can understand the health level of the user's hand-brain coordination, and provide a scientific and quantitative basis for preventing hand strain, optimizing the human-computer interaction experience, and diagnosing potential nervous system or motor system abnormalities. Furthermore, the present invention dynamically adjusts the baseline parameter threshold by combining the potential physiological risk type and the hand-brain coordination score to obtain a sensitivity threshold, so that the threshold setting is more in line with the user's individual physiological health status, which can improve the perception accuracy of changes in the user's physiological state and ensure the sensitivity and accuracy of subsequent early warning mechanisms. Furthermore, the present invention performs health monitoring processing on the target user based on the multi-dimensional hierarchical early warning strategy, thereby realizing accurate analysis of the health risks of the target user, ensuring that an effective early warning is issued at the initial stage of the risk, assisting the user to take timely intervention measures, reducing the chance of illness, and comprehensively protecting the user's physical and mental health. Therefore, the keyboard design method and system for integrating health monitoring provided in the embodiment of the present invention can improve the accuracy of health detection when the user uses the keyboard.

[0150] Example 2:

[0151] like Figure 2 The figure shows a functional module diagram of a keyboard design system integrating health monitoring according to the present invention.

[0152] The keyboard design system 200 integrated with health monitoring described in the present invention can be installed in an electronic device. Depending on the functions implemented, the keyboard design system integrated with health monitoring can include a physiological signal feature generation module 201, a potential physiological risk type analysis module 202, a hand-brain coordination scoring and evaluation module 203, a multi-dimensional hierarchical early warning strategy generation module 204, and a health monitoring module 205. The module described in the present invention can also be referred to as a unit, which refers to a series of computer program segments that can be executed by an electronic device processor and can perform fixed functions, and is stored in the memory of the electronic device.

[0153] In the embodiment of the present invention, the functions of each module / unit are as follows:

[0154] The physiological signal feature generation module 201 is used to obtain keyboard operation data and keyboard rhythm data of the target user, wherein the keyboard operation data includes key pressure distribution, acceleration timing signal, and input interval duration, analyze the keystroke behavior type corresponding to the keyboard operation data, and collect the hand surface electromyography signal and skin impedance change data of the target user, and combine the hand surface electromyography signal and the skin impedance change data to generate the physiological signal feature of the target user;

[0155] The potential physiological risk type analysis module 202 is configured to set a baseline parameter threshold for the target user's health monitoring based on the keystroke behavior type and the physiological signal characteristics, calculate the rhythm entropy change corresponding to the keyboard rhythm data, and construct a hand control degradation model for the target user based on the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change. Based on the hand control degradation model, the target user's corresponding potential physiological risk type is analyzed.

[0156] The hand-brain coordination scoring and evaluation module 203 is used to schedule the user's historical operation data of the target user, analyze the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user's historical operation data, and evaluate the hand-brain coordination score of the target user based on the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve;

[0157] The multi-dimensional hierarchical warning strategy generation module 204 is used to dynamically adjust the baseline parameter threshold based on the potential physiological risk type and the hand-brain coordination score to obtain a sensitivity threshold, analyze the current scenario requirements corresponding to the keystroke behavior type in real time, and generate a multi-dimensional hierarchical warning strategy for the target user based on the preset pathological risk mapping rules, the current scenario requirements, and the sensitivity threshold;

[0158] The health monitoring module 205 is configured to perform health monitoring processing on the target user based on the multi-dimensional layered early warning strategy to obtain a health monitoring result.

[0159] In detail, the modules in the keyboard design system 200 integrating health monitoring in the embodiment of the present invention are used in the same manner as above. Figure 1 The keyboard design method integrating health monitoring is the same technical means as described in , and can produce the same technical effects, so I will not go into details here.

[0160] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0161] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A keyboard design method integrating health monitoring, characterized in that: The method comprises: Acquiring keyboard operation data and keyboard rhythm data of a target user, the keyboard operation data including key pressure distribution, acceleration timing signals, and input interval duration, analyzing the keystroke behavior type corresponding to the keyboard operation data, and collecting hand surface electromyographic signals and skin impedance change data of the target user, and combining the hand surface electromyographic signals and the skin impedance change data to generate a physiological signal signature of the target user; Combining the keystroke behavior type and the physiological signal characteristics, setting a baseline parameter threshold for the target user's health monitoring, calculating the rhythm entropy change corresponding to the keyboard rhythm data, and constructing a hand control degradation model for the target user based on the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change. Based on the hand control degradation model, analyzing the potential physiological risk type corresponding to the target user; Dispatching the user's historical operation data of the target user, analyzing the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user's historical operation data, and evaluating the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve; Combined with the potential physiological risk type and the hand-brain coordination score, the baseline parameter threshold is dynamically adjusted to obtain a sensitivity threshold. The current scenario requirements corresponding to the keystroke behavior type are analyzed in real time. Combined with the preset pathological risk mapping rules, the current scenario requirements and the sensitivity threshold, a multi-dimensional hierarchical early warning strategy for the target user is generated; Based on the multi-dimensional hierarchical early warning strategy, health monitoring processing is performed on the target user to obtain a health monitoring result.

2. The keyboard design method integrating health monitoring according to claim 1, characterized in that: The step of combining the hand surface electromyography signal and the skin impedance change data to generate the physiological signal characteristics of the target user includes: Performing denoising processing on the hand surface electromyographic signal to obtain a denoised electromyographic signal; Segment-processing the denoised electromyographic signal to obtain electromyographic signal segments; Extracting time-frequency features of the electromyographic signal segments to obtain time-frequency features of the electromyographic signal; performing baseline correction processing on the skin impedance change data to obtain corrected skin impedance change data; extracting skin impedance features from the corrected skin impedance change data; Performing feature alignment processing on the electromyographic signal time-frequency feature and the skin impedance feature to obtain an aligned joint feature; Performing feature splicing on the aligned joint features to obtain a spliced joint feature; The health status of the spliced joint features is encoded to generate physiological signal features of the target user.

3. The keyboard design method integrating health monitoring according to claim 1, wherein: The step of setting a baseline parameter threshold for health monitoring of the target user in combination with the keystroke behavior type and the physiological signal feature includes: Quantizing and encoding the keystroke behavior type to obtain a keystroke behavior feature vector; performing standardization processing on the physiological signal characteristics to obtain standard physiological signal characteristics; Calculating feature similarity between the keystroke behavior feature vector and the standard physiological signal feature; When the feature similarity is greater than a preset similarity, performing feature fusion processing on the keystroke behavior feature vector and the standard physiological signal feature to obtain a fused feature vector; A baseline parameter threshold for health monitoring of the target user is set based on the fused feature vector.

4. The keyboard design method integrating health monitoring according to claim 1, wherein: The step of constructing a hand control degradation model of the target user by combining the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change includes: Performing sliding window processing on the key position pressure distribution to obtain key position pressure sliding characteristics; Calculating the dynamic balance of the key position pressure distribution based on the key position pressure sliding characteristics; performing signal decomposition processing on the acceleration time series signal to obtain a frequency sub-band signal; extracting a tremor frequency subband signal from the frequency subband signals, and calculating the tremor spectrum energy of the acceleration time series signal based on the tremor frequency subband signal; The timing fluctuation characteristics of the input interval duration are analyzed, and a hand control degradation model of the target user is constructed by combining the timing fluctuation characteristics, the dynamic balance, the tremor spectrum energy, and the rhythm entropy change.

5. The keyboard design method integrating health monitoring according to claim 4, characterized in that: The calculating the dynamic balance degree of the key position pressure distribution based on the key position pressure sliding characteristics includes: The characteristic standard deviation corresponding to the key position pressure sliding characteristic is calculated, and based on the characteristic standard deviation, the dynamic balance degree of the key position pressure distribution is calculated by the following formula: Among them, A represents the dynamic balance of key pressure distribution, σ a represents the characteristic standard deviation corresponding to the ath feature in the key pressure sliding feature, a represents the serial number of the key pressure sliding feature, q represents the number of key pressure sliding features, max(σ a ,σ a+1 ,…,σ q ) represents the function of selecting the maximum value of the characteristic standard deviation.

6. The keyboard design method integrating health monitoring as claimed in claim 4, characterized in that: The calculating, based on the tremor frequency sub-band signal, the tremor spectrum energy of the acceleration time series signal includes: Performing Fourier transform processing on the tremor frequency subband signal to obtain a tremor subband signal spectrum; performing a spectrum modulo operation on the spectrum of the tremor subband signal to obtain a tremor spectrum amplitude; Calculating the spectrum energy density corresponding to the tremor frequency subband signal based on the tremor spectrum amplitude; A tremor frequency interval corresponding to the tremor frequency subband signal is determined, and an integration operation is performed on the spectrum energy density within the tremor frequency interval to obtain tremor spectrum energy of the acceleration time series signal.

7. The keyboard design method integrating health monitoring according to claim 4, wherein: The analyzing the timing fluctuation characteristics of the input interval duration includes: Calculating timestamp differences between adjacent intervals in the input interval duration, and determining an input interval sequence corresponding to the input interval duration based on the timestamp differences; Performing window segmentation processing on the input interval sequence to obtain a window interval sequence; Calculate the sequence mean corresponding to the window interval sequence, and calculate the sequence variation coefficient corresponding to the window interval sequence according to the sequence mean using the following formula: Among them, B represents the sequence variation coefficient corresponding to the window interval sequence, F b represents the sequence mean of the bth sequence in the window interval sequence, Δt be represents the e-th time series within the b-th sequence in the window interval sequence, b represents the sequence number of the window interval sequence, e represents the sequence number of the time series in the window interval sequence, and R represents the number of window interval sequences; Based on the sequence variation coefficient, the time series fluctuation characteristics of the input interval duration are analyzed.

8. The keyboard design method integrating health monitoring as claimed in claim 1, characterized in that: The step of evaluating the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve includes: Performing trajectory smoothing on the fatigue accumulation trajectory to obtain a smoothed fatigue accumulation trajectory; extracting morphological parameters from the smooth fatigue accumulation trajectory, and calculating the muscle load aging coefficient of the target user based on the morphological parameters; Calculating a control response attenuation rate of the target user according to the stability curve; The hand-brain coordination score of the target user is evaluated by combining the muscle load aging coefficient and the control response attenuation rate.

9. The keyboard design method integrating health monitoring according to claim 8, characterized in that: The calculating the muscle load aging coefficient of the target user based on the morphological parameters includes: Mining the morphological dimension features corresponding to the morphological parameters, performing feature screening processing on the morphological dimension features, and obtaining target morphological dimension features; Analyzing the independent contribution of the morphological dimension features to muscle load aging to obtain a contribution set; Based on the contribution degree set, allocating the dimension importance corresponding to the morphological dimension feature; The muscle load aging coefficient of the target user is calculated by combining the dimension importance and the morphological dimension characteristics.

10. A keyboard design system integrating health monitoring, characterized in that: The system comprises: a physiological signal feature generation module, configured to obtain keyboard operation data and keyboard rhythm data of a target user, the keyboard operation data including key pressure distribution, acceleration timing signals, and input interval duration, analyze the keystroke behavior type corresponding to the keyboard operation data, and collect hand surface electromyography signals and skin impedance change data of the target user, and combine the hand surface electromyography signals and the skin impedance change data to generate a physiological signal feature of the target user; a potential physiological risk type analysis module, configured to set a baseline parameter threshold for the target user's health monitoring based on the keystroke behavior type and the physiological signal characteristics, calculate the rhythm entropy change corresponding to the keyboard rhythm data, construct a hand control degradation model for the target user based on the key pressure distribution, the acceleration timing signal, the input interval duration, and the rhythm entropy change, and analyze the potential physiological risk type corresponding to the target user based on the hand control degradation model; a hand-brain coordination scoring and evaluation module, configured to schedule the user's historical operation data of the target user, analyze the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve in the user's historical operation data, and evaluate the hand-brain coordination score of the target user by combining the hand fatigue accumulation trajectory and the instantaneous operation steady-state curve; A multi-dimensional hierarchical warning strategy generation module is used to dynamically adjust the baseline parameter threshold based on the potential physiological risk type and the hand-brain coordination score to obtain a sensitivity threshold, analyze the current scenario requirements corresponding to the keystroke behavior type in real time, and generate a multi-dimensional hierarchical warning strategy for the target user based on preset pathological risk mapping rules, the current scenario requirements, and the sensitivity threshold; The health monitoring module is used to perform health monitoring processing on the target user based on the multi-dimensional layered early warning strategy to obtain health monitoring results.

Images