Handlebar type health data real-time analysis and prompt system

Abstract

The invention relates to the technical field of health data processing, in particular to a handlebar-based health data real-time analysis and prompt system, which comprises a physiological signal sensing module, a trend feature extraction module, a state level identification module, a prompt scheme selection module and a multimode channel prompt module. According to the method, a multi-source signal sequence is constructed for trend characteristic analysis through synchronous acquisition and time window alignment of four types of signals including pulse beat cycle, grip rate, blood oxygen decreasing time period and contact area frequency, the health state level is judged according to a trend cross-coincidence relation, and a corresponding prompt scheme is matched; through cooperation of the signal change trend, real-time perception and grade classification of the physiological abnormal state of the rider are realized, the pertinence and effectiveness of prompt response are improved, the precision and stability of a continuous prompt intervention mechanism are enhanced, and the problems of potential health risk identification delay and lack of personalized regulation and control of prompt are effectively solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of health data processing, and in particular to a handle-based health data real-time analysis and prompting system. BACKGROUND

[0002] The technical field of health data processing involves collecting, storing, analyzing and feeding back physiological information related to individual health, and covers core matters such as real-time monitoring of physiological parameters, structured management of data, assessment of health status, and establishment of early warning mechanisms. This technical field systematically combines sensor information, artificial intelligence, and other means, and is widely used in remote medical health management, vehicle health monitoring, and smart wearable devices, aiming to achieve continuous sensing of individual health status, data-driven risk identification, and personalized recommendations. Among them, the traditional handle-based health data real-time analysis and prompting system refers to a system that collects health data of drivers or riders through sensor devices installed on the handlebars of vehicles and performs data analysis and prompting on terminal devices. The technical matter it addresses is the real-time monitoring and prompting of physiological parameter status of vehicle operators during riding or driving. The traditional handle-based health data real-time analysis and prompting system uses pressure sensors, photoelectric volume sensors, temperature sensors, and other methods to collect physiological indicators such as heart rate, body temperature, and blood oxygen, and performs parameter calculation through embedded microprocessors, and presents the processing results to the operator through liquid crystal displays, voice broadcasts, and other means.

[0003] In the prior art, real-time monitoring and prompting are dependent on a single physiological indicator such as heart rate, blood oxygen, or body temperature. In actual application, individual differences or transient fluctuations can easily lead to misjudgment, and there is no collaborative analysis mechanism across indicators, which cannot effectively identify abnormal correlation relationships between multi-dimensional physiological signals. In high-intensity riding or long-term operation, it is difficult to accurately perceive mild abnormalities or early trend changes, the prompting mechanism lacks flexible control ability, the prompt content is difficult to adapt to the actual risk level, resulting in unstable prompt interference frequency, response lag, and affecting the health management experience and safety response efficiency of the operator. SUMMARY

[0004] The purpose of the present application is to solve the shortcomings in the prior art, and to provide a handle-based health data real-time analysis and prompting system.

[0005] To achieve the above purpose, the present application adopts the following technical solution, the handle-based health data real-time analysis and prompting system comprises: The physiological signal sensing module obtains the pulse beat cycle signal, palm grip change rate, blood oxygen content decline time period, and real-time contact area change frequency of the rider's palm and handlebar during handlebar operation, and sets a data synchronization period to bind and align the four groups of signals within the time window to generate a multi-source aligned time window signal sequence. The trend feature extraction module labels trend change trajectories according to the change trend direction within the signal sequence based on the multi-source aligned time window signal sequence, to obtain a trend trajectory labeling group; The state level identification module selects the intersection of the grip rate mutation point and the pulse period jump interval according to the trend trajectory labeling group, and performs logical judgment on whether the blood oxygen drop delay start point and the palm area frequency highlight segment are synchronized in the intersection, to generate a trend coordination state level; The prompt scheme selection module calls the trend coordination state level, matches the audio reminder frequency, vibration frequency control value and light signal output intensity level in the preset prompt scheme configuration, selects the corresponding scheme number according to the state level, and generates a state corresponding prompt configuration group.

[0006] As a further scheme of the present application, the multi-source aligned time window signal sequence includes a main signal trend change mode, a cross-modal time alignment feature, a signal coupling relationship in a continuous segment, and a signal synchronization index. The trend trajectory labeling group includes a jumping discontinuity feature, a rate change interval distribution, a drop segment boundary identifier, and a frequency change node trajectory. The trend coordination state level includes a trend coincidence number identifier, a trend coincidence time length record, a trend coincidence time sequence structure, and a trend synchronization stability index. The state corresponding prompt configuration group includes an audio output parameter, a vibration execution parameter, a light signal control parameter, and a prompt unit enabling flag.

[0007] As a further scheme of the present application, the physiological signal sensing module includes: The multi-source data synchronization submodule acquires the pulse jumping period signal, the palm grip change rate, the blood oxygen content drop time period, and the real-time contact area change frequency of the palm and the handlebar of the rider during the handlebar operation, sets a unified data synchronization period, calls four groups of signal original sampling sequences, adjusts the signal data frame to a unified time window structure according to the signal timestamp information, and interpolates and corrects the time axis difference of the differentiated sampling rate, to obtain a synchronized multi-source signal sequence; The signal change extraction submodule normalizes the pulse period difference, the grip change rate, the contact area change frequency, and the blood oxygen drop rate according to the original signal change trajectory within the time window based on the synchronized multi-source signal sequence, calculates the joint fluctuation amplitude of the signal change within the time window, and generates a signal change joint fluctuation value sequence; The time window sequence construction submodule calls the signal change joint fluctuation value sequence, segments the sequence according to the signal change amplitude of the time window according to the sequence continuity and mutation amplitude, labels the segmented sequence with a labeling structure, and generates a multi-source aligned time window signal sequence.

[0008] As a further scheme of the present application, the trend feature extraction module includes: The pulse interruption extraction submodule locates the direction of pulse signal change based on the multi-source signal sequence within the time window, identifies the peak spacing offset segment in the pulse rhythm within a unit time window, counts the number of pulse interruption segments with continuous intervals greater than the reference period, and obtains the pulse interruption count value. The rate distribution labeling submodule calls the pulse interruption count value to locate the local inflection point in the grip force signal change curve, obtains the duration sequence of the contraction segment and the rebound segment within the time window, calculates the amplitude of the rate difference between the two segments, and divides the labeling distribution time period according to the threshold to obtain the grip force rate difference distribution label group. The trend trajectory recognition submodule extracts the start and end points of the descending segment in the blood oxygen signal within the time window based on the grip force rate difference distribution label group, records the duration period, extracts the position of the frequency rising segment in the contact area change curve, calculates the trend trajectory difference measure, and generates a trend trajectory label group by matching the signal segment identifier structure after arranging them in order.

[0009] As a further aspect of the present invention, the trend trajectory difference measurement adopts the formula: ; in, A measure of the difference in trend trajectories. Indicates the first The duration of the normalized blood oxygen saturation decrease within a time window. This represents the average frequency of the contact area after normalization. This represents the normalized grip strength contraction rate. This represents the grip rebound rate after normalization. Indicates the number of time windows.

[0010] As a further aspect of the present invention, the status level identification module includes: The trend cross-extraction submodule extracts grip force rate mutation points and pulse cycle jump intervals based on the trend trajectory annotation group. It obtains the continuous fluctuation time series of grip force rate and the pulse cycle change time segment, respectively, detects the corresponding time nodes on the time axis, and filters the cross-over segments in the time intervals of the two to obtain the trend cross-over segment index group. The delay synchronization discrimination submodule calls the trend crossover segment index group to obtain the blood oxygen fluctuation trend time series and palm area frequency signal change sequence in the crossover segment. According to the time synchronization judgment mechanism, it filters the blood oxygen decline delay start point and the frequency change peak time point in the palm area frequency band. It performs synchronization judgment based on whether the time difference is lower than the set delay synchronization threshold, records the synchronization mark position information in the time window, and generates the trend item synchronization mark result. The coordination level analysis submodule generates the trend coordination state level according to the trend item synchronous marking result, respectively counts the synchronization of the four trends of the grip rate, the pulse period, the blood oxygen delay and the palm area frequency in three continuous time windows, screens the corresponding trend combination state level label by accumulating the trend coincidence times, and generates the trend coordination state level.

[0011] As a further scheme of the present application, the prompting scheme selection module comprises: The state level matching submodule calls the trend coordination state level label group, obtains the state level label corresponding to the time window, matches the corresponding scheme number according to the mapping relationship between the state level and the level scheme number in the preset prompting scheme, generates the state level matching number sequence through the prompt scheme number corresponding to the state level label index, and generates the state level matching number sequence. The prompt configuration submodule retrieves the audio reminder frequency value, the vibration frequency control value and the light signal output intensity level value corresponding to the number in the preset prompting scheme according to the state level matching number sequence, reads the prompt unit enabling flag, and generates the state corresponding prompt configuration group.

[0012] As a further scheme of the present application, the system further comprises a multi-mode channel prompting module: The multi-mode channel prompting module activates the sound prompting module and the front-end visual prompting device of the handle by using the state corresponding prompt configuration group, synchronously triggers the three prompting channels according to the working time length and the frequency control signal marked in the configuration group, constructs the prompt signal persistence control chain, and generates the prompt mode linkage state. The prompt mode linkage state comprises a channel prompting linkage structure, a prompt output synchronization, a persistence control parameter and a module working state label.

[0013] As a further scheme of the present application, the multi-mode channel prompting module comprises: The prompt channel activation submodule calls the state corresponding prompt configuration group, obtains the audio reminder frequency value, the vibration frequency control value and the light signal output intensity level value, analyzes the enabling flag of the corresponding prompt channel, determines whether to activate the sound prompting module and the front-end visual prompting device of the handle through the flag, and generates the multi-channel prompt activation state label group. The control signal synchronization submodule extracts the control parameters of the activated channel according to the multi-channel prompt activation state label group, obtains the audio working time length, the vibration duration and the light signal maintenance time marked in the prompt configuration group, correspondingly generates the frequency control signals of the three types of channels, aligns the control signals according to the time reference, completes the synchronous signal output control, and obtains the channel prompting control synchronization sequence. The linkage state construction submodule calls the channel prompt control synchronization sequence, extracts the control signal duration and frequency parameters of the channel, constructs the prompt channel output state change record according to the time slice, combines the three types of channel states according to the synchronization time axis, and generates the prompt mode linkage state.

[0014] Compared with the prior art, the advantages and positive effects of the present application are that: In the present application, through the synchronous collection and time window alignment of four types of signals including pulse beat period, grip rate, blood oxygen drop period and contact area frequency, a multi-source signal sequence is constructed for trend feature analysis, the health state level is judged according to the trend cross-over relationship, and the corresponding prompt scheme is matched, the multi-channel linkage output of prompt frequency, vibration intensity and light signal control is realized, the trend linkage logic and time sequence coincidence recognition strategy between cross-modal data are established, the accuracy of risk identification is improved by combining the trend consistency and synchronism, the real-time perception and level classification of physiological abnormal state of the rider are realized through the cooperative signal change trend, the pertinence and effectiveness of prompt response are improved, the precision and stability of the continuous prompt intervention mechanism are enhanced, and the problems of delay in identifying potential health risks and lack of personalized control in prompting are effectively solved. BRIEF DESCRIPTION OF DRAWINGS

[0015] Figure 1 The system flowchart of the present application is shown in the figure; Figure 2 The physiological signal perception module flowchart in the present application is shown in the figure; Figure 3 The trend feature extraction module flowchart in the present application is shown in the figure; Figure 4 The state level identification module flowchart in the present application is shown in the figure; Figure 5 The prompt scheme selection module flowchart in the present application is shown in the figure; Figure 6 The multi-modal channel prompt module flowchart in the present application is shown in the figure. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical scheme and advantages of the present application clearer and more understandable, the present application will be further described in detail below in combination with the figures and examples. It should be understood that the specific examples described herein are only used to explain the present application, and are not used to limit the present application.

[0017] In the description of the present application, it should be understood that the terms "length", "width", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, in the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise explicitly and specifically limited.

[0018] Please refer to Figure 1 The real-time analysis and prompt system based on handlebar health data comprises: The physiological signal sensing module acquires the pulse beat cycle signal, palm grip change rate, blood oxygen content drop time period and real-time contact area change frequency of the rider's palm and handlebar during the operation of the handlebar, binds the four groups of signals in the time window respectively by setting the data synchronization period, aligns them, identifies the main signal variation sequence in the time window according to the continuous segmentation mode, and generates a multi-source aligned time window signal sequence; The trend feature extraction module extracts the number of intermittent pulse periods, the time period distribution of contraction and rebound rate in grip change, the start and end points of blood oxygen drop duration, and the position of contact area frequency rising section according to the trend direction in the signal sequence, respectively marks the trend change trajectory, and obtains a trend trajectory marking group based on the multi-source aligned time window signal sequence; The state level identification module selects the intersection of the grip rate mutation point and the pulse period jump interval, and performs logical judgment on whether the blood oxygen drop delay start point and the palm area frequency prominent section are synchronized in the intersection, if the four trends form a coincidence structure in three consecutive time windows, it is marked as level three, if three trends coincide, it is marked as level two, if any two trends coincide, it is marked as level one, and a trend coordination state level is generated; The prompt scheme selection module calls the trend coordination state level, matches the audio reminder frequency, vibration frequency control value and light signal output intensity level in the preset prompt scheme configuration, selects the corresponding scheme number according to the state level, and retrieves the enable flag of the prompt unit in the output combination scheme according to the corresponding number, and generates a state corresponding prompt configuration group; The multi-mode channel prompt module uses the state corresponding prompt configuration group to activate the sound prompt module and the visual prompt device at the front end of the handlebar, synchronously triggers the three prompt channels according to the working time length and frequency control signal marked in the configuration group, constructs a prompt signal persistence control chain, and generates a prompt mode linkage state; The multi-source alignment time window signal sequence includes a main signal trend change mode, a cross-modal time alignment feature, a signal coupling relationship in a continuous segment, and a signal synchronicity index. The trend trajectory label group includes a jump discontinuity feature, a rate change interval distribution, a falling segment boundary identification, a frequency change node trajectory, and a trend coordination state level. The trend coordination state level includes a trend overlap number identification, a trend overlap time length record, a trend overlap time sequence structure, and a trend synchronization stability index. The state corresponding prompt configuration group includes an audio output parameter, a vibration execution parameter, a light signal control parameter, and a prompt unit enable flag. The prompt mode linkage state includes a channel prompt linkage structure, a prompt output synchronicity, a persistence control parameter, and a module working state marker.

[0019] Please refer to Figure 2 The physiological signal sensing module includes: The multi-source data synchronization submodule acquires a pulse jump cycle signal, a palm grip force change rate, a blood oxygen content falling time period, and a real-time contact area change frequency of the palm and the handlebar of the rider during the operation of the handlebar. A uniform data synchronization period is set, four groups of signal original sampling sequences are called, signal data frames are adjusted to a uniform time window structure according to signal timestamp information, and the time axis difference of the differentiated sampling rate is interpolated and corrected, to obtain a synchronized multi-source signal sequence. The original data sequence of the pulse jump cycle signal is collected. The pulse signal is continuously recorded by a sampling device at a sampling rate of 500 Hz, 500 sampling points per second in the time domain are acquired, a total of 5000 sample points are formed within 10 seconds, a force value change sequence generated by the palm grip force through the pressure sensing unit is recorded, the sampling period is 0.01 s, the difference value between adjacent points of the recording interval constitutes the grip force change rate sequence, the blood oxygen value is collected by a skin conductance sensor, the time interval segment with a detection value lower than 95% is selected at the falling segment of the oxygenated hemoglobin concentration, the sampling interval is uniformly set as 1 s, the blood oxygen value is recorded and the first-order difference value is calculated every 1 s within the time, the falling time period is determined, the contact area between the palm and the handlebar is recorded by the thin film capacitive array attached to the surface of the handlebar, the area change corresponding to the change of the sensing contact capacitance is taken as the differential frequency to construct the real-time contact area change frequency sequence, the time stamps of the above four types of signals are matched, after the synchronous retrieval of the original sequence is completed, the time stamp recorded in the pulse signal is taken as the main reference time axis to construct a uniform time window structure, a time window with a length of 1 second is established according to the equal-interval window division strategy every second, the sample alignment of the time sequence of each type of signal is performed, the time axis offset amount of each type of signal due to the difference in sampling frequency is corrected by linear interpolation, the value of the missing time point in the blood oxygen signal is linearly interpolated by using the values of the adjacent two points, and the value is uniformly set as one frame of sample per second, to obtain the synchronized multi-source signal sequence.

[0020] The signal variation extraction submodule extracts the signal variation from the synchronized multi-source signal sequence. According to the original signal change trajectory in the time window, the pulse period difference, the grip change rate, the contact area change frequency and the blood oxygen drop rate are normalized respectively. The formula is as follows: ; The joint fluctuation amplitude of the signal change in the time window is calculated to generate a signal variation joint fluctuation value sequence. Wherein, represents the joint fluctuation amplitude of the signal change in the time window, represents the normalized pulse period difference in the first time window, represents the normalized grip rate variation value, represents the normalized contact area change frequency, represents the normalized blood oxygen drop rate, represents the number of time windows; The formula calculation logic is used to calculate the joint fluctuation amplitude of the multi-source signal change in each time window. The core logic is to standardize and extract the characteristic quantity of the four types of signals, and then aggregate the numerical values. The pulse period difference in the current time window reflects the heart rate rhythm change degree. The normalized grip change rate and the contact area change frequency are taken. After squaring and summing, the square root is taken to evaluate the action intensity. The normalized drop rate extracted from the blood oxygen signal represents the physiological load state. By integrating the time domain differences of the multi-source signals, a single amplitude value is output. The comprehensive state fluctuation degree of the measured object can be reflected in different time windows. The joint fluctuation amplitude of the signal change in the time window is a result of unified measurement of the change degree of multiple physiological and operation signals in the same time window. The amplitude value is calculated by the pulse period difference, grip fluctuation, contact area change and blood oxygen drop rate. It reflects the overall synchronicity and fluctuation intensity of the physiological rhythm and operation action of the measured individual in the time window. The larger the value, the more intense the signal variation in the time window, and the more unstable the signal state. Parameter significance and calculation process: The signal change trajectory in the corresponding time window is extracted with 1 second as the time window period. Each type of signal is normalized. In each time window, the pulse period signal is calculated. The distance between adjacent peaks is set to 0.1s and 0.85s respectively. Then =0.75s; Extract the grip sensor output value within the time window, set the total number of data points collected in the time window to 100, and process the difference value of each point by root mean square to obtain the grip change amplitude , set the grip change value sequence in a certain time window as [12, 15, 18, …, 22], then the corresponding variable value sequence is [3, 3, …, 4], square the mean value and take the square root, the result is 3.5N / s; For the contact area change rate in the time window , take the change value in the differential sequence of the capacitance to obtain the value is 0.025m 2 / s; Take the difference between the average value of each point in the current window and the average value of the previous window in the blood oxygen value as , if the current window average is 94% and the previous window average is 96%, then =2%; The following example parameters (see Table 1) are used, and are substituted into the formula to derive the actual example: Table 1: Joint fluctuation amplitude parameter table (unit unified) Time window number Pulse period difference (s) Grip force change rate (N / s) Contact area change rate (m 2 / s) Blood oxygen difference (%) 1 0.75 3.5 0.025 2 2 0.68 4.1 0.030 1.5 3 0.80 2.9 0.020 2.2 4 0.70 3.2 0.027 1.8 As shown in Table 1, based on the multi-source signals collected in each time window, the fluctuation amplitude in each window is calculated, and the specific operation is as follows: For the first time window: ; For the second time window: ; For the third time window: ; For the fourth time window: ; Jointly average the above four time windows: ; The results show that in the four time windows of the sample test, the average fluctuation amplitude of the observed multi-source signal joint change is 2.2828, which will be an important input basis for dividing continuous change segments and mutation segments.

[0021] The time window sequence construction submodule calls the signal change joint fluctuation value sequence, according to the signal change amplitude of the time window, the segmented processing is carried out according to the sequence continuity and mutation amplitude, the segmented sequence is marked with a labeling structure, and a multi-source aligned time window signal sequence is generated; With the same time window structure as the basis for division, the corresponding amplitude value in each time window in the combined fluctuation amplitude sequence is extracted, and the continuous fluctuation amplitude sequence is sequentially stored. Taking 4 time windows as an example, the corresponding fluctuation amplitude sequence is [2.251, 3.2801, 1.5001, 2.1001], the sub-sequence is segmented, and the condition for identifying continuous change segments is set as the amplitude difference of adjacent two time windows not exceeding 0.8, that is, when |SV(i)-SV(i-1)|≤0.8, the current window is merged into the last segment, otherwise it is the starting window of a new segment. The first pair (2.251, 3.2801) in [2.251, 3.2801, 1.5001, 2.1001] has a difference of 1.0291, which is greater than 0.8, and is set as a break point. The second pair (3.2801, 1.5001) has a difference of 1.780, which is greater than 0.8, and again forms a segment boundary. The third pair (1.5001, 2.1001) has a difference of 0.6, which meets the condition, so it is merged into the same segment. Therefore, the segmented structure is: segment 1: [2.251], segment 2: [3.2801], segment 3: [1.5001, 2.1001]; the dispersion degree of the fluctuation amplitude value in each segment is screened for internal stability, and the standard deviation is not more than 0.5. If it exceeds, the segment is further divided into multiple sub-segments. The standard deviation of [1.5001, 2.1001] in segment 3 is set as: ; The standard deviation is lower than the threshold value 0.5, so segment 3 remains merged; after the division of each segment, a flag structure is assigned to each segment, and the starting window number, ending window number, sample number and average amplitude value are recorded in the form of labels. Segment 1: starting window number 1, ending window number 1, sample number 1, average amplitude 2.251; segment 2: starting window number 2, ending window number 2, sample number 1, average amplitude 3.2801; segment 3: starting window number 3, ending window number 4, sample number 2, average amplitude (1.5001+2.1001) / 2=1.8001, output multi-source alignment time window signal sequence.

[0022] Please refer to Figure 3 , the trend feature extraction module comprises: The beat discontinuity extraction submodule is based on the multi-source signal sequence in the time window, locates the pulse signal change direction, identifies the peak interval offset segment in the beat rhythm in the unit time window, counts the number of beat discontinuity segments with a continuous interval greater than the reference period, and obtains the pulse beat discontinuity count value; The direction of the pulse signal change is defined, the interval deviation section between the wave peaks in the rhythm of the pulse is identified in a unit time window, the number of pulse discontinuous sections with a continuous interval greater than a reference period is counted, a pulse discontinuous count value is obtained, a local peak point sequence in a time window is extracted from a pulse original signal sequence, the reference period is set to 0.75 seconds, the value is obtained by sampling the resting pulse signals of 50 testers for 10 minutes from the baseline test data, the mean value is 0.75 s, the standard deviation is 0.08 s, and ±2 times the standard deviation, that is, [0.59, 0.91], is selected as the normal interval for determining the period, the time interval between two consecutive wave peaks in the target time window is detected, if there is a wave peak interval greater than 0.91 seconds or less than 0.59 seconds, it is judged as a pulse discontinuous section, if multiple intervals exceeding the above interval are detected in a time window, the cumulative number is taken as the number of pulse discontinuous sections in the current time window, the time for detecting the wave peaks in the time window is set to 0.12 s, 0.95 s, 1.9 s and 2.8 s, and the adjacent intervals are 0.83 s, 0.95 s, 0.9 s and 0.95 s, which is greater than the upper limit 0.91 s, and is judged as a discontinuous section, the count value is 1, the wave peak interval deviation in the time window is processed, and the pulse discontinuous count value is obtained.

[0023] The pulse discontinuous count value is called by the rate distribution labeling sub-module, the local inflection point position in the grip signal change curve is located, the duration sequence of the contraction section and the rebound section in the time window is obtained, the difference amplitude of the two sections is calculated, and the distribution time section is labeled according to the threshold value, and a grip rate difference distribution label group is obtained; The local inflection point position in the grip signal change curve is located, the duration sequence of the contraction section and the rebound section in the time window is obtained, the change sequence after the grip signal is normalized is taken as the input, the time length between the local minimum point and the maximum point is extracted as the contraction time, and the time length from the maximum point to the next minimum point is taken as the rebound time, the extreme value pairs are identified in each time window, and the duration of each section is recorded, the contraction and rebound time sequences are constructed, the contraction section is detected for 0.4 seconds and the rebound section is detected for 0.3 seconds in a certain time window, and then the [0.4, 0.3] is recorded as the rate change section structure, the grip contraction time difference and the rebound time difference of adjacent two time windows are calculated, the window 1 contraction is 0.4 seconds, the window 2 is 0.6 seconds, the difference is 0.2 seconds, the rate difference amplitude array is constructed in this way, the difference array constructed for the time window is set, the distribution time section interval is 5 windows, and the standard deviation is calculated in each 5-window section to represent the grip change rate stability level, if the standard deviation is greater than the set threshold value 0.15, it is labeled as a “fluctuation section”, otherwise it is labeled as a “stable section”, the window sequence [0.4, 0.6, 0.5, 0.65, 0.55] is set, the standard deviation is calculated as 0.089, which is less than 0.15, and is labeled as a stable section, the segmented labeling structure is given a unique number, and a grip rate difference distribution label group is obtained.

[0024] The trend trajectory recognition submodule distributes labels based on grip force rate differences, extracts the start and end points of the descending segment in the blood oxygen signal within a time window, records the duration, and extracts the position of the frequency rising segment in the contact area change curve using the following formula: ; Calculate the trend trajectory difference measure, arrange them in order, match the signal segment identifier structure, and generate trend trajectory annotation groups; in, A measure of the difference in trend trajectories. Indicates the first The duration of the normalized blood oxygen saturation decrease within a time window. This represents the average frequency of the contact area after normalization. This represents the normalized grip strength contraction rate. This represents the grip rebound rate after normalization. Indicates the number of time windows; Formula calculation logic: The formula aims to assess the synergistic trend of various key signals across different time windows, fusing features from three dimensions, including the duration of the blood oxygen saturation decline. Average frequency of contact area change and grip contraction rate With rebound rate First calculate and The absolute difference between them quantifies the time shift between blood oxygenation and changes in contact area. and The summation and square root of the squares are used to quantify the intensity of dynamic changes in actions within a time window. Essentially, by superimposing the differences and rate intensities between signal trends in a structured manner, the signal intensity measures whether the signals of each channel change synchronously in the time trend, thereby assessing the coordination and degree of difference between signal changes within a time window. The trend trajectory difference measure indicates whether there is coordination or deviation between the changing trends of key signals within multiple time windows. It combines the time difference between the decrease in blood oxygen and the change in contact area, as well as the dynamic intensity of grip force, to quantify the degree of asynchrony between signals. The larger the value, the more asynchronous the changes in trends of multiple signal channels are, and the more inconsistent the overall fluctuation trend is. Meaning of parameters and calculation process: The window numbers within the fluctuation segment are mapped and matched with the corresponding blood oxygen change sequences to identify the time range within which the blood oxygen value continuously decreases within the fluctuation segment. For example, in windows 3 to 5, the blood oxygen value drops from 97% to 93%, and the duration of the decrease is recorded as 3 seconds. Second; The contact area frequency sequence in the same time window sequence is processed, the differential value sequence is taken to obtain the change rate of each window, and the average contact area change rate in each fluctuation section is taken as a parameter The differential mean of windows 3 to 5 is set to 0.023 m 2 / s, and the normalized value is taken ; The contraction rate and the rebound rate are extracted from the grip strength change curve, the contraction rate is represented as the grip strength growth amplitude per unit time, and the rebound rate is represented as the grip strength decline amplitude per unit time, which are taken as parameters 、 , and if the corresponding values are normalized, they are 0.45 and 0.38 respectively; Now, actual data is substituted for calculation, and each parameter is as follows: Table 2: Trend trajectory difference calculation parameter table Window number Blood oxygen drop duration (s) Contact area change average Grip force contraction rate Grip force rebound rate 1 3 0.6 0.45 0.38 2 2.5 0.52 0.40 0.35 3 2.8 0.58 0.43 0.36 As shown in Table 2, the formula is expanded for calculation: The first term: ; The second term: ; The third term: ; The formula is substituted for calculation: ; The result shows that the trend trajectory difference measure of multi-source signals in the current sampling window under the corresponding fluctuation section is 2.7604.

[0025] Referring to Figure 4 , the state level recognition module comprises: The trend intersection extraction submodule extracts the grip rate mutation point and the pulse period jump interval based on the trend trajectory label group, obtains the continuous fluctuation time sequence of the grip rate and the pulse period change time segment, detects the corresponding time nodes on the time axis, screens the intersection overlap section on the time interval, and obtains the trend intersection overlap section index group; A continuous grip force data stream is acquired through a sensor, which can be collected in real time by a force-sensitive sensor integrated in a smart glove or a fitness handle, forming a complete time series record. The sequence is extracted by a sliding window method to identify the mutation point where the rate rapidly rises or falls. The pulse period jump interval is derived from the time pulse waveform signal recorded by a finger clip pulse detector or a wristband device. The length of the period is calculated cycle by cycle, and the jump position is labeled to generate a pulse jump interval set. The grip mutation point and the pulse jump interval are mapped to a unified time axis for comparison. If there is an overlapping time period, it is considered as a cross-over segment between the two trends. In actual use scenarios, if the user is holding the handle tightly while experiencing physiological fluctuations such as a sudden increase in pulse, the time period will be identified as a trend cross-over state, and the trend cross-over segment index set is obtained.

[0026] The time delay synchronization discrimination submodule calls the trend cross-over segment index set to obtain the corresponding blood oxygen fluctuation trend time series and palm area frequency signal change sequence in the cross-over segment. According to the time synchronization judgment mechanism, the blood oxygen drop delay starting point and the frequency change peak time point in the palm area frequency band are selected. The synchronization is determined according to whether the time difference is less than the set delay synchronization threshold. The synchronization marker position information in the time window is recorded to generate the trend item synchronization marker result. The blood oxygen and palm area frequency signal sequences in the segment need to be extracted separately. The extraction of the blood oxygen fluctuation trend depends on the continuous recording values of the wearable blood oxygen instrument, and the drop trend is identified as a point where the relatively stable state starts to rapidly decrease. The point is marked as the delay starting point. After the user performs intense exercise, the blood oxygen value starts to decrease due to temporary oxygen deficiency. At this time, the system can record the position. The palm area frequency change signal is captured by the pressure-sensitive area or strain gauge. The system needs to extract the time point corresponding to the frequency peak value. The two event time points are compared to determine whether they are within the preset time tolerance range, i.e., whether the time difference between the two is close enough to be considered as a synchronous event. In application scenarios, such as during cycling, if the user's blood oxygen value suddenly decreases and the palm frequency abnormally fluctuates at almost the same time, the system determines that it is a set of valid synchronous events, and adds the event marker in the recorded time window. The synchronization marker results formed in multiple cross-over segments are collected and sorted into a structured state identifier to generate the trend item synchronization marker result.

[0027] The coordination level analysis submodule marks the trend item synchronization result, respectively counts the synchronization of the four trends of grip rate, pulse period, blood oxygen delay and palm area frequency in the continuous three time windows, filters the corresponding trend combination state level mark through the cumulative trend coincidence number, and generates the trend coordination state level; In the state level identification link, the process needs to equally divide the entire monitoring time axis, adopts the fixed time window sliding forward mode, so as to make a stability judgment on the synchronization, respectively counts the synchronization state of the four trends of grip rate, pulse period, blood oxygen delay and palm area frequency in each time window. In actual application, when the system traverses the time window, it judges whether there are two or more trends in the same window. If the four trends are synchronized, the time window is determined to be the highest level of synchronization, if only three trends are synchronized, it is the second level, and if any two trends are synchronized, it is the first level. Through the level mapping mechanism, each time window can be assigned a specific level label. In actual scenarios, if the user appears four trends synchronization in the continuous three time windows, it indicates that the current physiological fluctuation is highly consistent, which is marked as level three and the corresponding time index is recorded. The level results are sorted and stored in the data structure with time sequence as the main index, and the trend coordination state level is generated.

[0028] Please refer to Figure 5 The prompt scheme selection module includes: The state level matching submodule calls the trend coordination state level label group, obtains the state level label corresponding to the time window, matches the corresponding scheme number according to the mapping relationship between the state level and the level scheme number in the preset prompt scheme, generates the state level matching number sequence through the prompt scheme number corresponding to the state level mark index. When calling the trend coordination state level label group, the system will traverse the time window label record contained therein in turn, read the corresponding state level value at each time index point. The level value is a discrete numerical type, and the first, second and third levels are set to correspond to different degrees of physiological trend coordination strength. The system enters the preset prompt scheme configuration table according to the level value. The configuration table is a static definition structure, which contains the binding relationship between each state level and the scheme number. The state level 1 corresponds to the prompt scheme number A1, the level 2 corresponds to A2, and the level 3 corresponds to A3. In actual deployment, the mapping relationship can be preset in the embedded controller, and the table or hash mapping form is used for quick query. Through the state level label as the index item, the system quickly retrieves the corresponding prompt scheme number from the table, and converts the level mark in each time window into the corresponding prompt number to form a sequence structure. In application scenarios, if the state level labels of the continuous three time windows in a certain period of time are 3, 2 and 2, the system will retrieve the prompt scheme numbers A3, A2 and A2, and generate the state level matching number sequence.

[0029] The prompt configuration submodule retrieves the audio reminder frequency value, vibration frequency control value, and light signal output intensity level value corresponding to the number in the preset prompt scheme according to the state level matching number sequence, reads the prompt unit enable flag, and generates a state corresponding prompt configuration group; The configuration content corresponding to each number is read in turn. The configuration content has been set and loaded into the internal database or control chip of the device before the system is deployed. The structure records the prompt parameters associated with each scheme number, including the audio reminder frequency, vibration prompt frequency control value, and light signal output intensity level. Each parameter is a numerical variable. Each configuration also contains a flag indicating whether the prompt unit is enabled, which is used to determine whether the sound, vibration, or light signal channel is enabled. In actual application, the configuration content set for number A2 is an audio frequency of 1000 Hz, a vibration frequency of 3 times per second, a light signal intensity level of 2, and a prompt unit enable flag of "110", which means that the sound and vibration are enabled but the light prompt is not. The system reads the above parameters in table form during the traversal of the number sequence and arranges the results in chronological order in the data structure to generate a state corresponding prompt configuration group.

[0030] Please refer to Figure 6 , the multi-mode channel prompt module includes: The prompt channel activation submodule calls the state corresponding prompt configuration group to obtain the audio reminder frequency value, vibration frequency control value, and light signal output intensity level value, parses the enable flag of the corresponding prompt channel, determines whether to activate the sound prompt module and the front-end visual prompt device of the handle through the flag, and generates a multi-channel prompt activation state identification group. The three core prompt parameters in each configuration record, namely the audio reminder frequency, vibration frequency control value, and light signal output intensity level, are interpreted. At the same time, the prompt channel enable flag is read. The flag is a three-bit binary value, and each bit corresponds to the enable state of the audio, vibration, and light signal three types of prompt channels. After reading the flag, the system parses the value of each bit to determine whether the prompt function of a specific channel needs to be enabled at the current time window. In actual scenarios, if the enable flag in a configuration record is "101", it means that the audio and light signal prompts need to be enabled at the current state, but the vibration channel is not enabled. The system generates corresponding activation commands based on the results and sends channel start signals to the sound prompt module and the front-end integrated visual prompt device of the handle. At the same time, the system constructs an activation state identification structure for the three types of channels to record whether each prompt module is in an activated state at the current time, and generates a multi-channel prompt activation state identification group.

[0031] The control signal synchronization submodule extracts the control parameters of the activated channel according to the multi-channel prompt activation state identifier group, obtains the audio working time, vibration duration and light signal maintenance time marked in the prompt configuration group, generates frequency control signals of the three types of channels, aligns the control signals according to the time reference, completes the synchronization signal output control, and obtains the channel prompt control synchronization sequence. The system generates the corresponding control pulse signal according to the frequency parameter of each activated channel, and the signal will be constructed into a control sequence with fixed frequency and periodic characteristics according to the specific control requirements of the channel. For example, if the activation state is set to "1" and the corresponding frequency value is 1200Hz and the working time is 3 seconds, the system will generate a 1200Hz square wave control signal with a duration of 3 seconds. If the vibration channel is activated and the frequency is set to 4 times per second and the maintenance time is 2 seconds, the system will generate an equally spaced excitation signal pulse sequence. After the channel control signal is generated, all signals are aligned according to the time reference line to ensure that the multi-channel prompts can start and maintain synchronization output state within the same time window, and the channel prompt control synchronization sequence is obtained.

[0032] The linkage state construction submodule calls the channel prompt control synchronization sequence, extracts the control signal duration and frequency parameters of the channel, constructs the prompt channel output state change record according to the time slice, combines the three types of channel states according to the synchronization time axis, and generates the prompt mode linkage state. The three types of prompt signals synchronized and output in the system are structured according to the time slice. The control signal of each type of channel includes specific start and end time, control frequency and duration, etc. The system will record the output state change of each channel in the slice as the basic unit of time. In actual application scenarios, the audio channel is set to output a 1200Hz prompt sound from t=5.0s to t=8.0s, the light channel is set to output a three-level intensity flicker signal from t=5.5s to t=9.0s, and the vibration channel is set to output vibration pulses at intervals from t=6.0s to t=8.0s. The system converts the above three types of states into state vectors corresponding to the time period, splices and combines the state vectors of different channels on the same time reference, constructs a cross-channel linkage state matrix, and reflects the interaction between the multi-channel prompts through the time alignment structure of the matrix. The system uses the linkage state structure for real-time prompt control and data recording, and outputs the prompt mode linkage state.

[0033] The above merely describes the preferred embodiments of the present application, and is not intended to limit the present application in other forms. Any skilled person in the art can modify or change the disclosed technical content into equivalent embodiments with equivalent changes, and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments according to the technical essence of the present application, without departing from the technical solution content of the present application, still falls within the protection scope of the present application.

Claims

1. A real-time analysis and prompting system based on handlebar health data, characterized in that, The system comprises: The physiological signal sensing module acquires the pulse beat period signal, palm grip rate of change, blood oxygen content falling time period and real-time contact area change frequency of the palm and the handlebar of the rider during the handlebar operation process, binds the four groups of signals in the time window respectively through setting a data synchronization period, aligns them, generates a multi-source aligned time window signal sequence, and outputs the signal sequence. The trend feature extraction module labels the trend change trajectory according to the change trend direction in the signal sequence based on the multi-source aligned time window signal sequence, obtains a trend trajectory labeling group, and outputs the trend trajectory labeling group. The state level identification module selects the intersection of the grip rate mutation point and the pulse period jump interval according to the trend trajectory labeling group, performs logical judgment on whether the blood oxygen falling delay starting point and the palm area frequency prominent segment are synchronized in the intersection, generates a trend coordination state level, and outputs the trend coordination state level. The prompt scheme selection module calls the trend coordination state level, matches the audio reminder frequency, vibration frequency control value and light signal output intensity level in the preset prompt scheme configuration, selects a corresponding scheme number according to the state level, generates a state corresponding prompt configuration group, and outputs the state corresponding prompt configuration group.

2. The handlebar-based real-time health data analysis and prompting system of claim 1, wherein, The multi-source aligned time window signal sequence comprises a main signal trend change mode, a cross-modal time alignment feature, a signal coupling relationship in a continuous segment, and a signal synchronization index; the trend trajectory labeling group comprises a beat discontinuity feature, a rate change interval distribution, a falling segment boundary identifier and a frequency change node trajectory; the trend coordination state level comprises a trend coincidence number identifier, a trend coincidence time length record, a trend coincidence time sequence structure and a trend synchronization stability index; and the state corresponding prompt configuration group comprises an audio output parameter, a vibration execution parameter, a light signal control parameter and a prompt unit enabling flag.

3. The handlebar-based real-time health data analysis and prompting system of claim 1, wherein, The physiological signal sensing module comprises: The multi-source data synchronization submodule acquires the pulse beat period signal, palm grip rate of change, blood oxygen content falling time period and real-time contact area change frequency of the palm and the handlebar of the rider during the handlebar operation process, sets a unified data synchronization period, calls four groups of signal original sampling sequences, adjusts signal data frames to a unified time window structure according to signal timestamp information, and inserts and corrects the time axis difference of differentiated sampling rates to obtain a synchronized multi-source signal sequence; The signal change extraction submodule normalizes the pulse period difference, grip change rate, contact area change frequency and blood oxygen falling rate according to the original signal change trajectory in the time window based on the synchronized multi-source signal sequence, calculates the joint fluctuation amplitude of the signal change in the time window, generates a signal change joint fluctuation value sequence, and outputs the signal change joint fluctuation value sequence. The time window sequence construction submodule calls the signal change joint fluctuation value sequence, segments the sequence according to the signal change amplitude of the time window in sequence continuity and mutation amplitude, labels the segmented sequence with a labeling structure, generates a multi-source aligned time window signal sequence, and outputs the multi-source aligned time window signal sequence.

4. The handlebar-based real-time health data analysis and prompting system of claim 3, wherein, The trend feature extraction module comprises: The beat discontinuity extraction submodule is based on the multi-source signal sequence in the time window, locates the pulse signal change direction, identifies the peak interval offset segment in the beat rhythm in a unit time window, counts the number of beat discontinuity segments with a continuous interval greater than a reference period, and obtains a pulse beat discontinuity count value; The rate distribution labeling submodule calls the pulse beat discontinuity count value, locates the local inflection point position in the grip signal change curve, obtains the duration sequence of the contraction segment and the rebound segment in the time window, calculates the rate difference amplitude of the two segments, and labels the distribution time segment according to the threshold value to obtain a grip rate difference distribution label group; The trend trajectory identification submodule extracts the start and end points of the falling segment in the blood oxygen signal in the time window according to the grip rate difference distribution label group, records the duration, extracts the position of the frequency rising segment in the contact area change curve, calculates the trend trajectory difference measure, and matches the signal segment identification structure after arranging in sequence to generate a trend trajectory label group.

5. The handlebar-based real-time health data analysis and prompting system of claim 4, wherein, The trend trajectory difference measure uses the formula: ； wherein, denotes a trend trajectory difference measure, denotes a normalized processed blood oxygen drop duration in the denotes a normalized processed contact area frequency average, denotes a normalized processed grip contraction rate, denotes a normalized processed grip rebound rate, denotes a normalized processed grip rebound rate, denotes a number of time windows.

6. The handlebar-based real-time health data analysis and prompting system of claim 4, wherein, The state level identification module includes: The trend intersection extraction submodule extracts the grip rate mutation point and the pulse period jump interval based on the trend trajectory label group, respectively obtains the continuous fluctuation time sequence of the grip rate and the pulse period change time segment, detects the corresponding time nodes on the time axis, screens the intersection overlapping segment on the time interval of the two, and obtains a trend intersection overlapping segment index group; The delay synchronization discrimination submodule calls the trend intersection overlapping segment index group, obtains the corresponding blood oxygen fluctuation trend time sequence and palm area frequency signal change sequence in the intersection overlapping segment, screens the blood oxygen falling delay start point and the frequency change peak time point in the palm area frequency band according to the time synchronization judgment mechanism, performs synchronization judgment according to whether the time difference value is lower than the set delay synchronization threshold, records the synchronization marker position information in the time window, and generates a trend item synchronization marker result; The collaborative level analysis submodule generates a trend collaborative state level according to the trend item synchronization marker result, respectively counts the synchronization of the grip rate, the pulse period, the blood oxygen delay, and the palm area frequency in the trend in the continuous three time windows, screens the corresponding trend combination state level label by accumulating the trend coincidence times, and generates the trend collaborative state level.

7. The handlebar-based real-time health data analysis and prompting system of claim 6, wherein, The prompt scheme selection module includes: The state level matching submodule calls the trend collaborative state level label group, obtains the state level label corresponding to the time window, matches the corresponding scheme number according to the mapping relationship between the state level and the level scheme number in the preset prompt scheme, generates a state level matching number sequence through the prompt scheme number corresponding to the state level label index, and generates a state level matching number sequence. The prompt configuration submodule retrieves the audio reminder frequency value, the vibration frequency control value, and the light signal output intensity level value corresponding to the number in the preset prompt scheme according to the state level matching number sequence, reads the prompt unit enable flag, and generates a state corresponding prompt configuration group.

8. The handlebar-based real-time health data analysis and prompting system of claim 1, wherein, The system further includes a multi-mode channel prompt module: The multi-mode channel prompting module activates the sound prompting module and the front-end visual prompting device of the handlebar by using the state corresponding prompting configuration group, synchronously triggers the three prompting channels according to the working time length and frequency control signal marked in the configuration group, constructs a prompting signal persistent control chain, and generates a prompting mode linkage state; The prompting mode linkage state includes a channel prompting linkage structure, prompting output synchronicity, persistent control parameters, and module working state markers.

9. The handlebar-based real-time health data analysis and prompting system of claim 8, wherein, The multi-mode channel prompting module includes: The prompting channel activation submodule calls the state corresponding prompting configuration group, obtains an audio prompting frequency value, a vibration frequency control value, and a light signal output intensity level value, analyzes an activation flag bit of the corresponding prompting channel, determines whether to activate the sound prompting module and the front-end visual prompting device of the handlebar through the flag bit, generates a multi-channel prompting activation state marker group, and outputs a synchronous signal. The control signal synchronization submodule extracts the control parameters of the activated channel according to the multi-channel prompting activation state marker group, obtains the audio working time length, vibration duration, and light signal maintenance time marked in the prompting configuration group, correspondingly generates frequency control signals of the three types of channels, aligns the control signals according to a time reference, completes synchronous signal output control, and obtains a channel prompting control synchronization sequence. The linkage state construction submodule calls the channel prompting control synchronization sequence, extracts the control signal duration and frequency parameters of the channel, constructs a prompting channel output state change record according to a time segment, combines the states of the three types of channels according to a synchronous time axis, and generates a prompting mode linkage state.

Images