Sports injury state early warning system based on intelligent flexible sensing device

Abstract

The application discloses a motion injury state early warning system based on an intelligent flexible sensing device and relates to the technical field of motion injury early warning. In order to solve the problem that motion injury cannot be judged and handled in the process of motion, the application is verified by sensitivity and effectiveness, quantitatively combines medical standards and individual baselines, assesses time according to dimensions, assigns weights and defines grade thresholds, improves risk identification accuracy and assessment standardization, matches self-checking calibration and time synchronization to guarantee data accuracy, can extract multi-dimensional injury correlation parameters and quantitatively load, motion and fatigue state, avoids single monitoring limitations, and lays a high-quality data foundation for subsequent risk analysis.

Description

Technical Field

[0001] This invention relates to the field of sports injury early warning technology, specifically a sports injury status early warning system based on intelligent flexible sensing devices. Background Technology

[0002] Existing sports injury early warning technologies have significant shortcomings: poor adaptability of sensing devices, difficulty in covering multiple target parts of the human body, and lack of self-calibration and standardized preprocessing of collected data, resulting in insufficient data accuracy; lack of sensitivity and effectiveness verification in risk feature extraction, failure to assign weights according to the importance of indicators during assessment, and vague criteria for level determination; single early warning channels without priority, lack of personalized suggestions based on user status and historical data, leading to delayed early warnings, low accuracy, and difficulty in effectively preventing and controlling sports injuries. Summary of the Invention

[0003] The purpose of this invention is to provide a sports injury status early warning system based on intelligent flexible sensing devices. Through dual verification of sensitivity and effectiveness, it quantifies and combines medical standards and individual baselines, assigns weights to dimensions and clarifies level thresholds during assessment, thereby improving the accuracy of risk identification and the standardization of assessment. With self-calibration and time synchronization, it ensures data accuracy and can also extract multi-dimensional injury-related parameters and quantify load, movement, and fatigue status, avoiding the limitations of single monitoring and laying a high-quality data foundation for subsequent risk analysis. This invention can solve the problems in existing technologies.

[0004] To achieve the above objectives, the present invention provides the following technical solution:

[0005] A sports injury status early warning system based on intelligent flexible sensing devices includes:

[0006] The flexible sensor data acquisition and processing unit is used to: acquire raw motion data of the target part of the human body in real time during the movement, and preprocess the acquired raw motion data.

[0007] The motion parameter analysis unit is used to extract key motion parameters related to sports injuries from the preprocessed data and quantify the human motion state.

[0008] The risk feature extraction unit is used to: filter out injury-related sensitive features from motion parameters and convert them into quantitative indicators;

[0009] First, based on the mechanism of motion pathology, highly sensitive features in key motion parameters are extracted;

[0010] High sensitivity features are typical injuries to different target parts of the human body, including ankle sprains, anterior cruciate ligament injuries of the knee, and lumbar muscle strain;

[0011] The injury risk assessment unit is used to: determine the level of injury risk under the current motion state based on quantitative indicators;

[0012] Among them, risk threshold standards related to the current type of movement and target parts of the human body are retrieved from the database, and the quantitative indicators of sensitive features are matched with the risk thresholds. After the matching is completed, a single indicator risk threshold is obtained.

[0013] The early warning signal generation unit is used to generate a corresponding early warning signal based on the assessed damage risk level.

[0014] The early warning instruction transmission unit is used to: parse the generated early warning signal and transmit the parsed instruction data to the user terminal;

[0015] The generated early warning signal is broken down into basic instruction information, content data, and transmission control parameters;

[0016] The disassembled early warning signals are then reassembled into structured data;

[0017] The user terminal personalization management unit is used to generate personalized sports and health suggestions based on the user's basic status, exercise history, and warning commands, and to display the generated sports and health suggestions.

[0018] Preferably, the flexible sensor data acquisition and processing unit is further used for:

[0019] First, wear the intelligent flexible sensing device on the target part of the human body. The intelligent flexible sensing device includes mechanical parameter sensing device, motion state sensing device, physiological parameter sensing device and auxiliary device; the target part of the human body includes the lower limb joints, trunk area, upper limb joints and muscle and tendon groups.

[0020] The intelligent flexible sensing device is turned on after being worn, and then performs self-testing and calibration.

[0021] After the self-test and calibration are completed and qualified, the raw motion data is collected in real time. During the collection process, the intelligent flexible sensing device synchronizes time through an internal clock or an external synchronization signal.

[0022] After real-time data acquisition is completed, it is transmitted to the local processing unit in real time through low-power wireless communication technology. At the same time, the data packet format is used during the transmission process, which includes sensor ID, timestamp and measurement value.

[0023] The local processing unit performs data preprocessing on the received raw motion data, including data cleaning and outlier handling, signal filtering and noise reduction, and data normalization and calibration.

[0024] The final step is to complete the process of acquiring and preprocessing raw motion data.

[0025] Preferably, the motion parameter analysis unit is further used for:

[0026] Based on biomechanical mechanisms, damage-related parameters were screened from the raw motion data, where the biomechanical mechanisms were extracted from the database;

[0027] Damage-related parameters include mechanical load parameters, kinematic parameters, dynamic parameters, and symmetry parameters;

[0028] The selected damage-related parameters are subjected to feature extraction, which includes time-domain feature extraction and periodic feature extraction.

[0029] Temporal feature extraction involves: extracting the key extreme values ​​of the damage-related parameters within the motion cycle, calculating the statistics of the damage-related parameters within the motion cycle, and finally capturing the changing trend of the damage-related parameters over time.

[0030] Periodic feature extraction involves: segmenting continuous motion into action cycles, extracting parameter features within each cycle after segmentation, calculating the mean and coefficient of variation of multiple cycles, and finally analyzing the parameter change trend of adjacent cycles.

[0031] The injury-related parameters after feature extraction are integrated into quantitative indicators under motion conditions, including load level quantification, motion quality quantification, and fatigue degree quantification.

[0032] Among them, load level quantification is as follows: the absolute load index is constructed by using the characteristic values ​​of mechanical parameters, and then normalized by combining human physiological parameters, and relative load index is constructed; movement quality quantification is as follows: the actual movement parameters are compared with the standard movement parameters, and then the variation of the parameters is quantified, and finally the difference of the parameters of the two limbs is calculated; fatigue level quantification is as follows: the change trend of parameters with exercise duration is quantified, and then the synchronous change of multiple parameters is quantified.

[0033] Ultimately, the key motion parameters were extracted, and the human motion state was quantified.

[0034] Preferably, the risk feature extraction unit is further configured to:

[0035] The extracted highly sensitive features undergo sensitivity verification. The sensitivity verification process is as follows:

[0036] S1: Clinical data comparison: Compare highly sensitive features with a database of historical sports injury cases, and confirm the correlation between the degree of feature abnormality and the probability of injury based on the comparison results;

[0037] S2: Dynamic threshold test: Resimulate the gradual changes in motion parameters and confirm the synchronicity between the characteristic values ​​and the pre-injury signals;

[0038] S3: Redundant Feature Removal: Finally, perform correlation analysis between features, and confirm the sensitivity verification process for highly sensitive features based on the correlation analysis results.

[0039] The highly sensitive feature data that has completed sensitivity verification will then undergo validity verification, which includes cross-group verification, cross-movement scenario verification, and timeliness verification.

[0040] Preferably, confirming the synchronicity between the characteristic value and the pre-injury signal includes:

[0041] The time series of eigenvalues ​​and precursor signals of damage are divided to generate multiple continuous time periods;

[0042] For each time period, based on the characteristic values ​​and precursor signals of damage, obtain the characteristic value change trend parameters and the precursor signal change trend parameters for every two adjacent time periods;

[0043] The characteristic value change trend parameter includes the slope of the characteristic value and the absolute difference between the maximum and minimum values ​​of the characteristic value within the time period; the damage precursor signal change trend parameter includes the slope of the damage precursor signal and the absolute difference between the maximum and minimum values ​​of the damage precursor signal within the time period.

[0044] The trend of change of characteristic value parameters and the trend of change of damage precursor signal parameters are compared between every two adjacent time periods to obtain the trend comparison results.

[0045] When the trend comparison results show that the slope trend in the characteristic value change trend parameter is not exactly the same as the slope trend of the damage precursor signal, it is determined that the characteristic value and the damage precursor signal are not synchronized, and an asynchronous warning is issued.

[0046] When the trend comparison results show that the slope trend of the characteristic value change trend parameter is exactly the same as the slope trend of the damage precursor signal, the synchronization quality between the characteristic value and the damage precursor signal is determined.

[0047] Preferably, determining the synchronization quality between the feature value and the precursor signal includes:

[0048] Retrieve the slope value of each feature value within each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data within those time periods;

[0049] Normalize the slope value of each feature value in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in the time period to obtain the normalized slope value of each feature value and the absolute difference between the maximum and minimum data values ​​in the time period.

[0050] Retrieve the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data in that time period;

[0051] Normalize the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in that time period to obtain the normalized slope value of each pre-injury signal and the absolute difference between the maximum and minimum data values ​​in that time period.

[0052] Synchronization quality assessment parameters are obtained by combining the normalized slope value corresponding to each feature value in each two adjacent time periods with the absolute difference between the maximum and minimum data values ​​in that time period, and the normalized slope value corresponding to each pre-injury signal with the absolute difference between the maximum and minimum data values ​​in that time period.

[0053] The synchronization quality assessment parameters are compared with preset assessment thresholds;

[0054] If the synchronization quality assessment parameter does not exceed the preset assessment threshold, the synchronization quality is determined to be non-compliant, and a synchronization anomaly alarm is triggered.

[0055] Preferably, the risk feature extraction unit is further configured to:

[0056] Transform highly sensitive features that have undergone validity verification into quantitative indicators;

[0057] The process for converting quantitative indicators is as follows:

[0058] S4: Hierarchical quantification of feature values: First, set the safety value, warning value and danger value of the feature according to the medical safety standard. The medical safety standard is extracted from the database. Then, the feature change rate is calculated in combination with the individual baseline. The duration or frequency of persistent abnormal features is quantified.

[0059] S5: Multi-feature integration and quantification: Based on the influence weight of different features on the injury, a comprehensive risk value is calculated. At the same time, based on the linkage relationship between features, the mechanism of multi-factor synergy leading to injury in real sports is simulated. Finally, the quantification standard is adjusted in real time according to the duration of sports and the degree of fatigue.

[0060] S6: Standardization of quantitative indicators: Map the comprehensive quantitative value to high risk, medium risk and low risk, then normalize the characteristic indicators of different types, and finally organize and output the quantitative indicators in the structure of characteristic name, current value, risk level and benchmark value.

[0061] Finally, the quantitative indicators of sensitive features are transformed.

[0062] Preferably, the damage risk assessment unit is further used for:

[0063] After retrieving risk threshold standards related to the current type of movement and target body part and matching single-index risk thresholds, a virtual movement scene is constructed through a quantum simulator to simulate the tissue stress distribution under different combinations of mechanical parameters.

[0064] Risk levels are determined based on the range of single-indicator risk thresholds, including low risk, medium risk, and high risk. Low risk means that the single-indicator risk threshold is within the safe range; medium risk means that the single-indicator risk threshold reaches the warning threshold but does not reach the danger threshold; and high risk means that the single-indicator risk threshold reaches or exceeds the danger threshold.

[0065] Weights are assigned to different dimensions based on the risk level determination results of single-indicator risk thresholds, including core risk dimensions, auxiliary risk dimensions, and temporary weight adjustments in special scenarios;

[0066] The risk level determination result of the single indicator risk threshold is combined with the corresponding assigned weight to calculate the comprehensive risk value. After the comprehensive risk value is calculated, the overall risk level is obtained.

[0067] If the overall risk value is less than 1.5, it is judged as low risk overall; if 1.5 ≤ overall risk value < 2.5, it is judged as medium risk overall; if the overall risk value is greater than or equal to 2.5, it is judged as high risk overall.

[0068] Finally, the level of injury risk under the current motion state is determined.

[0069] Preferably, the warning signal generation unit is further configured to:

[0070] The overall risk level is mapped to the early warning strategy templates in the database, and different early warning strategies for different risk levels are obtained after the mapping is completed;

[0071] Early warning content is generated based on different early warning strategies. The early warning content includes core information, action instructions, and feedback instructions.

[0072] The core information includes risk characterization, risk positioning, and risk quantification; action instructions provide immediate corrective suggestions for instructions with medium risk overall, and send stop instructions for instructions with high overall risk; feedback instructions generate encouraging information and reinforce correct behavior when the risk level decreases or user actions improve.

[0073] The generated warning content is transmitted to the user through physical signals via different sensory channels, including tactile, auditory, and visual signals;

[0074] Among them, the tactile signal is: adjusting the vibration intensity of the intelligent flexible sensing device according to the risk level, and using different vibration modes to distinguish information; the auditory signal is: using different frequencies of sound, for medium and high risks, directly playing the generated text content through headphones or device speakers; the visual signal is: displaying the warning content on the user terminal with icons and text simultaneously, and using color to distinguish the risk level.

[0075] When multiple risks occur simultaneously or signal channels are occupied, high-risk warnings have the highest interruption priority.

[0076] Finally, based on the target terminal, signal mode, specific content, and duration, a warning signal corresponding to the damage risk level is obtained.

[0077] Preferably, the early warning instruction transmission unit is further configured to:

[0078] Structured data reorganized into:

[0079] S7: Adopts a hierarchical structure: the top layer contains basic control fields, the middle layer contains core risk information, and the bottom layer contains content data;

[0080] S8: Unified Data Identification: Add standardized labels to each field;

[0081] S9: Compress redundant information: Simplify repetitive or minor content and retain differentiated data;

[0082] The structured data is adjusted according to the type, status, and communication environment of the user terminal;

[0083] The adjustment of structured data is as follows: adapt terminal characteristics according to the type of user terminal, select the corresponding transmission path according to the communication method, configure communication parameters, and finally encrypt the transmitted structured data.

[0084] After the structured data is adjusted, it is transmitted to the corresponding user terminal.

[0085] Preferably, the early warning instruction transmission unit is further configured to:

[0086] At the same time, when structured data is transmitted to the user terminal, the fastest transmission channel is automatically selected.

[0087] First, confirm the amount of structured data transmitted. The amount of data transmitted can be found in the data information database.

[0088] Real-time monitoring of the remaining capacity and signal strength of each transmission channel;

[0089] Select a transmission channel with a data transmission volume less than the remaining signal capacity as the first transmission channel for structured data.

[0090] When there is more than one first transmission channel, the transmission channel with the strongest channel signal strength is selected as the second transmission channel for structured data.

[0091] Choose either the first transmission channel or the second transmission channel as the communication channel for transmitting structured data to the user terminal.

[0092] Preferably, the user terminal personalization management unit is further configured to:

[0093] After receiving the warning signal, the user terminal first integrates the data to obtain a real-time user profile, which includes real-time warning instructions, basic user status, and historical motion data.

[0094] Personalized suggestions are generated based on real-time user profiles. These personalized suggestions include the generation of immediate response strategies, planning of short-term adjustment plans, and recommendations for long-term improvement plans.

[0095] The system includes real-time response strategy generation, which consists of high-risk and medium-risk warnings. High-risk warnings are mandatory recommendations, while medium-risk warnings are corrective recommendations. Short-term adjustment plan planning combines real-time warnings with user goals and plans subsequent actions. Long-term improvement plan recommendations are based on recurring error patterns, automatically generating long-term suggestions to address the root causes, while also considering the user's medical history.

[0096] The generated personalized suggestions undergo a feasibility check. Once the feasibility check is completed and the suggestions pass, they are displayed on the user's terminal.

[0097] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0098] 1. The sports injury status early warning system based on intelligent flexible sensing device provided by the present invention, combined with self-test calibration and time synchronization to ensure data accuracy, can also extract multi-dimensional injury-related parameters and quantify load, movement and fatigue status, avoiding the limitations of single monitoring, and laying a high-quality data foundation for subsequent risk analysis.

[0099] 2. The sports injury status early warning system based on intelligent flexible sensing device provided by the present invention improves the accuracy of risk identification and the standardization of assessment by verifying both sensitivity and effectiveness, quantifying and combining medical standards and individual baselines, assigning weights by dimension and specifying level thresholds during assessment.

[0100] 3. The sports injury status early warning system based on intelligent flexible sensing devices provided by this invention features structured transmission adapted to terminals and encrypted, generates full-dimensional suggestions based on user profiles and performs feasibility checks, ensuring timely early warning, stable transmission, and practical and personalized suggestions. Attached Figure Description

[0101] Figure 1 This is a schematic diagram of the sports injury state early warning unit of the present invention;

[0102] Figure 2 This is a schematic diagram of the sports injury state early warning steps of the present invention. Detailed Implementation

[0103] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0104] To address the shortcomings of existing technologies in flexible sensing data acquisition, such as lack of device classification and adaptation, self-calibration, time synchronization, and standardized preprocessing; and issues with motion parameter analysis including incomplete biomechanical parameter selection, incomplete feature extraction, and limited quantification dimensions, which prevent accurate acquisition of key motion data, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0105] A sports injury status early warning system based on intelligent flexible sensing devices includes:

[0106] The flexible sensor data acquisition and processing unit is used to: acquire raw motion data of the target part of the human body in real time during the movement, and preprocess the acquired raw motion data.

[0107] The motion parameter analysis unit is used to extract key motion parameters related to sports injuries from the preprocessed data and quantify the human motion state.

[0108] The risk feature extraction unit is used to: filter out injury-related sensitive features from motion parameters and convert them into quantitative indicators;

[0109] The injury risk assessment unit is used to: determine the level of injury risk under the current motion state based on quantitative indicators;

[0110] The early warning signal generation unit is used to generate a corresponding early warning signal based on the assessed damage risk level.

[0111] The early warning instruction transmission unit is used to: parse the generated early warning signal and transmit the parsed instruction data to the user terminal;

[0112] The user terminal personalization management unit is used to generate personalized sports and health suggestions based on the user's basic status, exercise history, and warning commands, and to display the generated sports and health suggestions.

[0113] The flexible sensor data acquisition and processing unit is also used for:

[0114] First, wear the intelligent flexible sensing device on the target part of the human body. The intelligent flexible sensing device includes mechanical parameter sensing device, motion state sensing device, physiological parameter sensing device and auxiliary device; the target part of the human body includes the lower limb joints, trunk area, upper limb joints and muscle and tendon groups.

[0115] The intelligent flexible sensing device is turned on after being worn, and then performs self-testing and calibration.

[0116] After the self-test and calibration are completed and qualified, the raw motion data is collected in real time. During the collection process, the intelligent flexible sensing device synchronizes time through an internal clock or an external synchronization signal.

[0117] After real-time data acquisition is completed, it is transmitted to the local processing unit in real time through low-power wireless communication technology. At the same time, the data packet format is used during the transmission process, which includes sensor ID, timestamp and measurement value.

[0118] The local processing unit performs data preprocessing on the received raw motion data, including data cleaning and outlier handling, signal filtering and noise reduction, and data normalization and calibration.

[0119] The final step is to complete the process of acquiring and preprocessing raw motion data.

[0120] Specifically, the intelligent flexible sensing device includes three core types of sensors: mechanical, motion state, and physiological parameters. It can capture human motion and physiological information from multiple dimensions, and specifically cover key parts such as lower limb joints, trunk, upper limb joints, and muscle and tendon groups. It can completely acquire human motion-related data, avoiding the limitations of monitoring a single part or parameter. After the device is worn, it undergoes a self-test and calibration process to identify equipment faults and correct initial errors in advance, ensuring the accuracy of subsequent data collection from the source. During data collection, time synchronization is achieved through an internal clock or external synchronization signal to ensure the temporal consistency of multi-sensor data, which facilitates subsequent motion timing analysis. It also adopts low-power wireless communication technology to balance real-time transmission and device battery life. The data includes sensor ID, timestamp, and measurement value, which can clearly identify the data source and collection time, avoiding data confusion. Fourth, the data preprocessing is comprehensive. The local processing unit effectively removes interference data and unifies the data format by cleaning outliers, filtering and reducing noise, and normalizing calibration, laying a high-quality data foundation for subsequent data applications.

[0121] The motion parameter analysis unit is also used for:

[0122] Based on biomechanical mechanisms, damage-related parameters were screened from the raw motion data, where the biomechanical mechanisms were extracted from the database;

[0123] Damage-related parameters include mechanical load parameters, kinematic parameters, dynamic parameters, and symmetry parameters;

[0124] The selected damage-related parameters are subjected to feature extraction, which includes time-domain feature extraction and periodic feature extraction.

[0125] Temporal feature extraction involves: extracting the key extreme values ​​of the damage-related parameters within the motion cycle, calculating the statistics of the damage-related parameters within the motion cycle, and finally capturing the changing trend of the damage-related parameters over time.

[0126] Periodic feature extraction involves: segmenting continuous motion into action cycles, extracting parameter features within each cycle after segmentation, calculating the mean and coefficient of variation of multiple cycles, and finally analyzing the parameter change trend of adjacent cycles.

[0127] The injury-related parameters after feature extraction are integrated into quantitative indicators under motion conditions, including load level quantification, motion quality quantification, and fatigue degree quantification.

[0128] Among them, load level quantification is as follows: the absolute load index is constructed by using the characteristic values ​​of mechanical parameters, and then normalized by combining human physiological parameters, and relative load index is constructed; movement quality quantification is as follows: the actual movement parameters are compared with the standard movement parameters, and then the variation of the parameters is quantified, and finally the difference of the parameters of the two limbs is calculated; fatigue level quantification is as follows: the change trend of parameters with exercise duration is quantified, and then the synchronous change of multiple parameters is quantified.

[0129] Ultimately, the key motion parameters were extracted, and the human motion state was quantified.

[0130] Specifically, the system relies on databases to extract biomechanical mechanisms and screen injury-related parameters, ensuring that the correlation between parameters and injuries has a theoretical basis, avoiding blind parameter selection, and improving the reliability of subsequent quantification. It covers mechanical load, kinematic, dynamic, and symmetry parameters, comprehensively covering the core dimensions related to injury from a biomechanical perspective, and can capture risk signals during exercise from multiple angles. Temporal feature extraction focuses on extreme values, statistics, and time trends, accurately capturing key parameter information within a single cycle. Cycle feature extraction, through cycle segmentation, multi-cycle statistics, and adjacent cycle analysis, takes into account both the repetitiveness and dynamic changes of movement, providing rich data support for quantification. Load level quantification combines absolute and relative indicators, taking into account both objective data and individual physiological differences. Movement quality quantification, through standard comparison, variation analysis, and bilateral difference calculation, can intuitively assess the standardization and symmetry of movements. Fatigue level quantification, based on duration trends and multi-parameter synchronicity, can dynamically monitor the process of exercise fatigue, providing accurate quantitative basis for exercise risk prevention and rehabilitation guidance.

[0131] To address the shortcomings of existing technologies, such as the lack of sensitivity and validity verification in extracting highly sensitive features for sports injuries, the absence of scientific quantification based on medical standards and individual circumstances, and the reliance on unweighted single indicators for risk assessment with unclear overall risk level determination criteria, please refer to [link to relevant documentation]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0132] The risk feature extraction unit is also used for:

[0133] First, based on the mechanism of exercise pathology, highly sensitive features are extracted from key exercise parameters;

[0134] High sensitivity features are typical injuries to different target parts of the human body, including ankle sprains, anterior cruciate ligament injuries of the knee, and lumbar muscle strain;

[0135] The extracted highly sensitive features undergo sensitivity verification. The sensitivity verification process is as follows:

[0136] S1: Clinical data comparison: Compare highly sensitive features with a database of historical sports injury cases, and confirm the correlation between the degree of feature abnormality and the probability of injury based on the comparison results;

[0137] S2: Dynamic threshold test: Resimulate the gradual changes in motion parameters and confirm the synchronicity between the characteristic values ​​and the pre-injury signals;

[0138] S3: Redundant Feature Removal: Finally, perform correlation analysis between features, and confirm the sensitivity verification process for highly sensitive features based on the correlation analysis results.

[0139] The highly sensitive feature data that has completed sensitivity verification will then undergo validity verification, which includes cross-group verification, cross-movement scenario verification, and timeliness verification.

[0140] Specifically, confirming the synchronicity between characteristic values ​​and pre-injury warning signals includes:

[0141] The time series of eigenvalues ​​and precursor signals of damage are divided to generate multiple continuous time periods;

[0142] For each time period, based on the characteristic values ​​and precursor signals of damage, obtain the characteristic value change trend parameters and the precursor signal change trend parameters for every two adjacent time periods;

[0143] The characteristic value change trend parameter includes the slope of the characteristic value and the absolute difference between the maximum and minimum values ​​of the characteristic value within the time period; the damage precursor signal change trend parameter includes the slope of the damage precursor signal and the absolute difference between the maximum and minimum values ​​of the damage precursor signal within the time period.

[0144] The trend of change of characteristic value parameters and the trend of change of damage precursor signal parameters are compared between every two adjacent time periods to obtain the trend comparison results.

[0145] When the trend comparison results show that the slope trend in the characteristic value change trend parameter is not exactly the same as the slope trend of the damage precursor signal, it is determined that the characteristic value and the damage precursor signal are not synchronized, and an asynchronous warning is issued.

[0146] When the trend comparison results show that the slope trend of the characteristic value change trend parameter is exactly the same as the slope trend of the damage precursor signal, the synchronization quality between the characteristic value and the damage precursor signal is determined.

[0147] The aforementioned technical solution first compares the slope trends of adjacent time periods to quickly identify obvious asynchrony, providing timely alerts and avoiding ineffective detailed analysis of asynchronous data. This significantly improves the efficiency of preliminary synchrony verification and provides an efficient pre-screening mechanism for the synchrony analysis process. For scenarios with consistent slope trends, normalization is introduced to eliminate dimensional differences in various parameters. Furthermore, multi-dimensional data, such as the average slope and the average of multiple absolute differences, are integrated to calculate synchrony quality assessment parameters. This overcomes the limitations of simply judging trend consistency, accurately quantifying synchrony quality from aspects such as numerical matching degree and overall change synergy, making synchrony determination more precise. By closely aligning with the correlation patterns between actual physiological signals and feature values, the accuracy of synchronicity quality assessment is significantly improved. Through comparison of synchronicity quality assessment parameters with preset thresholds, quantitative and graded early warning of synchronicity quality is achieved. When parameters fail to reach the threshold, a synchronicity anomaly alarm is triggered promptly, enabling early identification of potential synchronicity deficiencies. This lays a solid foundation for subsequent validation of the effectiveness of highly sensitive features, ensuring the reliability of the "feature-damage signal" correlation in the sports injury state early warning system. Ultimately, this enhances the accuracy of the entire system's sports injury risk warning, reduces misjudgments or missed judgments of injury risks due to synchronicity deviations, and more effectively provides scientific basis and technical support for sports injury prevention.

[0148] Specifically, determining the synchronization quality between the feature value and the pre-damage signal includes:

[0149] Retrieve the slope value of each feature value within each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data within those time periods;

[0150] Normalize the slope value of each feature value in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in the time period to obtain the normalized slope value of each feature value and the absolute difference between the maximum and minimum data values ​​in the time period.

[0151] Retrieve the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data in that time period;

[0152] Normalize the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in that time period to obtain the normalized slope value of each pre-injury signal and the absolute difference between the maximum and minimum data values ​​in that time period.

[0153] Synchronization quality assessment parameters are obtained by combining the normalized slope value corresponding to each feature value in each two adjacent time periods with the absolute difference between the maximum and minimum data values ​​in that time period, and the normalized slope value corresponding to each pre-injury signal with the absolute difference between the maximum and minimum data values ​​in that time period.

[0154] The synchronization quality assessment parameters are obtained using the following formula:

[0155]

[0156] Where S represents the synchronicity quality assessment parameter; It represents the average of the normalized slope values ​​corresponding to all feature values ​​within each two adjacent time periods; This represents the average of the normalized slope values ​​corresponding to all pre-injury signals within each two adjacent time periods; n represents the number of pre-injury signals. It represents the average absolute difference of the normalized i-th pre-injury signal within each two adjacent time periods; It represents the average of the absolute differences of all normalized eigenvalues ​​within each two adjacent time periods;

[0157] The synchronization quality assessment parameters are compared with preset assessment thresholds;

[0158] If the synchronization quality assessment parameter does not exceed the preset assessment threshold, the synchronization quality is determined to be non-compliant, and a synchronization anomaly alarm is triggered.

[0159] The aforementioned technical solution utilizes normalization processing to eliminate dimensional differences between different parameters, enabling the slope and absolute difference of eigenvalues ​​and pre-injury signals to be fused and analyzed under a unified dimension. This lays the foundation for accurate quantification of subsequent synchronization quality. The calculation of synchronization quality assessment parameters simultaneously considers the "difference of average slope" and the "average difference of multiple sets of absolute differences." From the two core dimensions of "consistency of change trend" and "synergy of numerical fluctuations," it comprehensively captures the synchronous correlation characteristics between eigenvalues ​​and pre-injury signals, breaking through the limitations of single-dimensional analysis and significantly improving the comprehensiveness and accuracy of synchronization quality assessment. By comparing the calculated synchronization quality assessment parameters with preset thresholds, it is possible to quantitatively determine whether the synchronization quality meets the requirements. When the parameters fail to meet the standards, a synchronization anomaly alarm is triggered in a timely manner, providing reliable "signal synchronization" support for subsequent sports injury status early warning. This effectively avoids misjudgment or missed judgment of injury risk caused by the synchronization deviation between eigenvalues ​​and pre-injury signals, significantly improving the reliability and accuracy of the sports injury status early warning system.

[0160] Meanwhile, existing technologies for verifying the "synchronicity between eigenvalues ​​and pre-injury signals" often focus on a single dimension (such as only timestamp alignment or simple trend matching). The aforementioned technical solution, through layered collaborative verification of "precise timeline alignment + dynamic correlation of changing trends + identification and correction of abnormal synchronization patterns," can not only identify explicit synchronization deviations but also uncover hidden problems such as "false synchronization" and "slightly delayed, correctable synchronization." This makes the verification results more adaptable to the dynamic changes in physiological signals under complex motion scenarios (such as motion mode switching and dynamic force exertion at multiple sites), significantly improving the anti-interference capability and scenario adaptability of synchronization verification—something that single-dimensional verification cannot achieve. Furthermore, the aforementioned technical solution achieves precise differentiation between "true synchronization" and "correctable synchronization." Even with signal transmission delays or complex environmental interference, it can still accurately determine synchronization, significantly reducing the false alarm rate and greatly improving the utilization rate of effective data. This overcomes the limitations of traditional synchronization verification's crude "either right or wrong" judgment. Furthermore, existing technologies for quantifying synchronicity quality often employ a single parameter (such as only slope or only amplitude difference) or a simple linear combination. The synchronicity quality assessment parameters of the aforementioned technical solution utilize a nonlinear fusion of an "exponential function (characterizing slope trend differences) + reciprocal form (integrating multiple sets of absolute difference differences)," allowing "consistency of change trends" and "synergy of numerical fluctuations" to mutually restrain and enhance each other. This provides a more nuanced and realistic reflection of the overall synchronicity quality between feature values ​​and pre-injury signals, avoiding "overall synchronicity misjudgment" caused by single-dimensional compliance, and significantly improving the consistency between the quantitative assessment results and the actual physiological signal correlation characteristics. Simultaneously, one of the core pain points of existing sports injury early warning systems is that "synchronicity deviation between features and injury signals leads to inaccurate warnings." The aforementioned technical solution optimizes the entire process from synchronicity verification to quality assessment, making the correlation between highly sensitive features and pre-injury signals more accurate. This, in turn, cascades improvements in the timeliness and accuracy of the entire system's warnings of sports injury risks, significantly reducing false alarms and missed alarms. This effectively solves the problem that "synchronicity improvements are only basic optimizations and cannot directly drive a qualitative change in the system's core early warning capabilities."

[0161] Transform highly sensitive features that have undergone validity verification into quantitative indicators;

[0162] The process for converting quantitative indicators is as follows:

[0163] S4: Hierarchical quantification of feature values: First, set the safety value, warning value and danger value of the feature according to the medical safety standard. The medical safety standard is extracted from the database. Then, the feature change rate is calculated in combination with the individual baseline. The duration or frequency of persistent abnormal features is quantified.

[0164] S5: Multi-feature integration and quantification: Based on the influence weight of different features on the injury, a comprehensive risk value is calculated. At the same time, based on the linkage relationship between features, the mechanism of multi-factor synergy leading to injury in real sports is simulated. Finally, the quantification standard is adjusted in real time according to the duration of sports and the degree of fatigue.

[0165] S6: Standardization of quantitative indicators: Map the comprehensive quantitative value to high risk, medium risk and low risk, then normalize the characteristic indicators of different types, and finally organize and output the quantitative indicators in the structure of characteristic name, current value, risk level and benchmark value.

[0166] Finally, the quantitative indicators of sensitive features are transformed.

[0167] Specifically, it directly focuses on typical injuries such as ankle sprains and anterior cruciate ligament injuries of the knee, accurately identifying core risk characteristics in key motion parameters, avoiding interference from irrelevant features, and improving the efficiency of risk identification. Sensitivity verification uses clinical data comparison to anchor the correlation between features and injuries, dynamic threshold testing to confirm the synchronicity with injury precursors, and redundancy removal to ensure feature conciseness and effectiveness. Effectiveness verification covers cross-groups, cross-scenarios, and timeliness, breaking through the limitations of single scenarios and ensuring the applicability of features under different conditions. Tiered quantification combines medical standards with individual baselines, taking into account both common safety and individual differences. Multi-feature integrated quantification considers weights and linkage relationships, conforming to the multi-factor synergistic mechanism of real injuries, and can also be adjusted in real time according to exercise duration and fatigue level, improving quantification accuracy. The indicators are normalized and have a clear structure, facilitating subsequent risk assessment and decision-making reference, providing an intuitive and practical basis for sports injury prevention and control.

[0168] The damage risk assessment unit is also used for:

[0169] After retrieving risk threshold standards related to the current type of movement and target body part and matching single-index risk thresholds, a virtual movement scene is constructed through a quantum simulator to simulate the tissue stress distribution under different combinations of mechanical parameters.

[0170] The risk threshold standards related to the current type of exercise and the target body part are retrieved from the database, and the quantitative indicators of sensitive features are matched with the risk thresholds. After the matching is completed, the single-indicator risk threshold is obtained.

[0171] Risk levels are determined based on the range of single-indicator risk thresholds, including low risk, medium risk, and high risk. Low risk means that the single-indicator risk threshold is within the safe range; medium risk means that the single-indicator risk threshold reaches the warning threshold but does not reach the danger threshold; and high risk means that the single-indicator risk threshold reaches or exceeds the danger threshold.

[0172] Weights are assigned to different dimensions based on the risk level determination results of single-indicator risk thresholds, including core risk dimensions, auxiliary risk dimensions, and temporary weight adjustments in special scenarios;

[0173] The risk level determination result of the single indicator risk threshold is combined with the corresponding assigned weight to calculate the comprehensive risk value. After the comprehensive risk value is calculated, the overall risk level is obtained.

[0174] If the overall risk value is less than 1.5, it is judged as low risk overall; if 1.5 ≤ overall risk value < 2.5, it is judged as medium risk overall; if the overall risk value is greater than or equal to 2.5, it is judged as high risk overall.

[0175] Finally, the level of injury risk under the current motion state is determined.

[0176] Specifically, the system retrieves risk thresholds corresponding to the current type of exercise and target body part from the database, avoiding the problem of insufficient adaptability of general thresholds and ensuring that the thresholds are highly consistent with the actual exercise scenario and the assessment site, laying a precise foundation for subsequent risk assessment. It clearly defines the threshold ranges for low, medium, and high risks, eliminating subjective judgment bias and making the classification of single-indicator risk levels more standardized and operable. It differentiates weights according to core and auxiliary risk dimensions, taking into account the degree of influence of different indicators on injuries, while also supporting temporary adjustments in special scenarios. This addresses the differences in risk factors brought about by changes in exercise scenarios, improving the adaptability of the assessment. The overall risk level is defined through specific numerical ranges, avoiding vague expressions and making the assessment results clear and easy to understand, facilitating the rapid acquisition of risk information and providing a direct and reliable basis for sports injury early warning and intervention decisions. The injury risk level judgment table is shown below:

[0177]

[0178] To address the issues in existing technologies, such as mismatch between early warning strategies and risk levels, incomplete content, single and unprioritized early warning transmission channels, poor data transmission adaptability, inefficient channel selection, and a lack of personalized and feasible suggestions based on user profiles, please refer to [link / reference needed]. Figure 1 and Figure 2 This embodiment provides the following technical solution:

[0179] The warning signal generation unit is also used for:

[0180] The overall risk level is mapped to the early warning strategy templates in the database, and different early warning strategies for different risk levels are obtained after the mapping is completed;

[0181] Early warning content is generated based on different early warning strategies. The early warning content includes core information, action instructions, and feedback instructions.

[0182] The core information includes risk characterization, risk positioning, and risk quantification; action instructions provide immediate corrective suggestions for instructions with medium risk overall, and send stop instructions for instructions with high overall risk; feedback instructions generate encouraging information and reinforce correct behavior when the risk level decreases or user actions improve.

[0183] The generated warning content is transmitted to the user through physical signals via different sensory channels, including tactile, auditory, and visual signals;

[0184] Among them, the tactile signal is: adjusting the vibration intensity of the intelligent flexible sensing device according to the risk level, and using different vibration modes to distinguish information; the auditory signal is: using different frequencies of sound, for medium and high risks, directly playing the generated text content through headphones or device speakers; the visual signal is: displaying the warning content on the user terminal with icons and text simultaneously, and using color to distinguish the risk level.

[0185] When multiple risks occur simultaneously or signal channels are occupied, high-risk warnings have the highest interruption priority.

[0186] Finally, based on the target terminal, signal mode, specific content, and duration, a warning signal corresponding to the damage risk level is obtained.

[0187] Specifically, by matching the overall risk level with the early warning strategy templates in the database, differentiated strategies are matched to different risk levels to avoid over- or under-warning, thus improving the targeting of early warnings. Core information clearly defines the nature, location, and degree of risk. Action instructions provide corrective suggestions and stop instructions for medium and high risks respectively. Feedback instructions encourage and reinforce correct behavior, forming a closed loop of notification, action, and feedback to enhance user motivation. It integrates multiple channels such as touch, hearing, and vision to adapt to user perception habits in different scenarios, improving the early warning reach rate. High-risk early warnings have the highest interruption priority, ensuring that key early warnings are delivered first when multiple risks occur concurrently or channels are occupied, avoiding the neglect of important information. It generates corresponding early warning signals by combining target terminals, signal modes, etc., taking into account device characteristics and user needs, ultimately achieving accurate, efficient, and personalized delivery of early warning signals, effectively improving the efficiency of risk response.

[0188] The early warning instruction transmission unit is also used for:

[0189] The generated early warning signal is broken down into basic instruction information, content data, and transmission control parameters;

[0190] The disassembled early warning signals are then reassembled into structured data, which is then reassembled as follows:

[0191] S7: Adopts a hierarchical structure: the top layer contains basic control fields, the middle layer contains core risk information, and the bottom layer contains content data;

[0192] S8: Unified Data Identification: Add standardized labels to each field;

[0193] S9: Compress redundant information: Simplify repetitive or minor content and retain differentiated data;

[0194] The structured data is adjusted according to the type, status, and communication environment of the user terminal;

[0195] The adjustment of structured data is as follows: adapt terminal characteristics according to the type of user terminal, select the corresponding transmission path according to the communication method, configure communication parameters, and finally encrypt the transmitted structured data.

[0196] After the structured data is adjusted, it is transmitted to the corresponding user terminal.

[0197] At the same time, when structured data is transmitted to the user terminal, the fastest transmission channel is automatically selected.

[0198] First, confirm the amount of structured data transmitted. The amount of data transmitted can be found in the data information database.

[0199] Real-time monitoring of the remaining capacity and signal strength of each transmission channel;

[0200] Select a transmission channel with a data transmission volume less than the remaining signal capacity as the first transmission channel for structured data.

[0201] When there is more than one first transmission channel, the transmission channel with the strongest channel signal strength is selected as the second transmission channel for structured data.

[0202] Choose either the first transmission channel or the second transmission channel as the communication channel for transmitting structured data to the user terminal.

[0203] Specifically, the warning signal is broken down into basic instructions, content data, and transmission control parameters. A hierarchical structure is then used to organize the logic, and standardized tags ensure data identifiability. Redundant content is compressed, reducing data volume and saving resources while maintaining a clear data structure for easy terminal parsing. Parameters are configured according to user terminal type and communication method to address compatibility issues across different terminals. Data encryption protects the transmission process, preventing information leakage and balancing compatibility and security. Real-time monitoring of channel capacity and signal strength prioritizes channels with sufficient capacity for transmission. When multiple channels meet the criteria, the channel with the strongest signal is selected to ensure data capacity and stable transmission, preventing transmission failures due to insufficient capacity or weak signal. The system automatically selects the fastest transmission channel, combining capacity and signal strength for dual filtering to ensure rapid data delivery while reducing the risk of transmission interruption. This ensures timely and stable delivery of warning signals to terminals, supporting users in receiving timely risk warnings.

[0204] The user terminal personalization management unit is also used for:

[0205] After receiving the warning signal, the user terminal first integrates the data to obtain a real-time user profile, which includes real-time warning instructions, basic user status, and historical motion data.

[0206] Personalized suggestions are generated based on real-time user profiles. These personalized suggestions include the generation of immediate response strategies, planning of short-term adjustment plans, and recommendations for long-term improvement plans.

[0207] The system includes real-time response strategy generation, which consists of high-risk and medium-risk warnings. High-risk warnings are mandatory recommendations, while medium-risk warnings are corrective recommendations. Short-term adjustment plan planning combines real-time warnings with user goals and plans subsequent actions. Long-term improvement plan recommendations are based on recurring error patterns, automatically generating long-term suggestions to address the root causes, while also considering the user's medical history.

[0208] The generated personalized suggestions undergo a feasibility check. Once the feasibility check is completed and the suggestions pass, they are displayed on the user's terminal.

[0209] Specifically, by integrating real-time alert commands, user basic status, and exercise history data to construct real-time user profiles, the system avoids the limitations of relying on a single data source. It comprehensively understands the user's current risk, individual conditions, and past exercise habits, providing accurate data for subsequent recommendations. Immediate response strategies are designed differently for high and medium risk levels. Short-term plans are combined with user goals to plan actions, while long-term plans address recurring error patterns and provide fundamental solutions related to injury history. This covers all dimensions of emergency response, short-term adjustments, and long-term improvement, meeting the needs of different scenarios. Feasibility checks screen qualified recommendations, avoiding ineffective suggestions that are out of touch with the user's actual abilities or environment, ensuring that recommendations are executable and easy to implement, improving user acceptance and execution effectiveness. Long-term recommendations are linked to injury history, and short-term plans are combined with user goals, fully considering individual differences. This avoids one-size-fits-all advice while addressing specific user issues, providing users with safe and personalized exercise guidance to help scientifically prevent injuries and manage long-term health.

[0210] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0211] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention.

Claims

1. A motion injury state early warning system based on intelligent flexible sensing devices, characterized in that, include: The flexible sensor data acquisition and processing unit is used to acquire raw motion data of the target part of the human body in real time during the movement, and to preprocess the acquired raw motion data. The motion parameter analysis unit is used to extract key motion parameters related to sports injuries from the preprocessed data and quantify the human motion state. The risk feature extraction unit is used to screen out injury-related sensitive features from motion parameters and convert them into quantitative indicators. First, based on the mechanism of motion pathology, highly sensitive features in key motion parameters are extracted; High sensitivity features are typical injuries to different target parts of the human body, including ankle sprains, anterior cruciate ligament injuries of the knee, and lumbar muscle strain; The injury risk assessment unit is used to determine the level of injury risk under the current motion state based on quantitative indicators. Among them, risk threshold standards related to the current type of movement and target parts of the human body are retrieved from the database, and the quantitative indicators of sensitive features are matched with the risk thresholds. After the matching is completed, a single indicator risk threshold is obtained. The early warning signal generation unit is used to generate a corresponding early warning signal based on the assessed damage risk level. The early warning instruction transmission unit is used to parse the generated early warning signal and transmit the parsed instruction data to the user terminal. The generated early warning signal is broken down into basic instruction information, content data, and transmission control parameters; The disassembled early warning signals are then reassembled into structured data; The user terminal personalization management unit is used to generate personalized sports and health suggestions based on the user's basic status, exercise history and warning instructions, and to display the generated sports and health suggestions. The risk feature extraction unit is also used for: The extracted highly sensitive features undergo sensitivity verification. The sensitivity verification process is as follows: S1: Clinical data comparison: Compare highly sensitive features with a database of historical sports injury cases, and confirm the correlation between the degree of feature abnormality and the probability of injury based on the comparison results; S2: Dynamic threshold test: Resimulate the gradual changes in motion parameters and confirm the synchronicity between the characteristic values ​​and the pre-injury signals; S3: Redundant Feature Removal: Finally, perform correlation analysis between features, and confirm the sensitivity verification process for highly sensitive features based on the correlation analysis results. The highly sensitive feature data that has completed sensitivity verification will then undergo validity verification, which includes cross-group verification, cross-movement scenario verification, and timeliness verification. Confirm the synchronicity between the characteristic values ​​and the precursory damage signals, including: The time series of eigenvalues ​​and precursor signals of damage are divided to generate multiple continuous time periods; For each time period, based on the characteristic values ​​and precursor signals of damage, obtain the characteristic value change trend parameters and the precursor signal change trend parameters for every two adjacent time periods; The characteristic value change trend parameter includes the slope of the characteristic value and the absolute difference between the maximum and minimum values ​​of the characteristic value within the time period; the damage precursor signal change trend parameter includes the slope of the damage precursor signal and the absolute difference between the maximum and minimum values ​​of the damage precursor signal within the time period. The trend of change of characteristic value parameters and the trend of change of damage precursor signal parameters are compared between every two adjacent time periods to obtain the trend comparison results. When the trend comparison results show that the slope trend in the characteristic value change trend parameter is not exactly the same as the slope trend of the damage precursor signal, it is determined that the characteristic value and the damage precursor signal are not synchronized, and an asynchronous warning is issued. When the trend comparison results show that the slope trend of the characteristic value change trend parameter is exactly the same as the slope trend of the damage precursor signal, the synchronization quality between the characteristic value and the damage precursor signal is determined.

2. The motion injury state early warning system based on intelligent flexible sensing device according to claim 1, characterized in that, The flexible sensor data acquisition and processing unit is also used for: First, wear the intelligent flexible sensing device on the target part of the human body. The intelligent flexible sensing device includes mechanical parameter sensing device, motion state sensing device, physiological parameter sensing device and auxiliary device; the target part of the human body includes the lower limb joints, trunk area, upper limb joints and muscle and tendon groups. The intelligent flexible sensing device is turned on after being worn, and then performs self-testing and calibration. After the self-test and calibration are completed and qualified, the raw motion data is collected in real time. During the collection process, the intelligent flexible sensing device synchronizes time through an internal clock or an external synchronization signal. After real-time data acquisition is completed, it is transmitted to the local processing unit in real time through low-power wireless communication technology. At the same time, the data packet format is used during the transmission process, which includes sensor ID, timestamp and measurement value. The local processing unit performs data preprocessing on the received raw motion data, including data cleaning and outlier handling, signal filtering and noise reduction, and data normalization and calibration. The final step is to complete the process of acquiring and preprocessing raw motion data.

3. The motion injury state early warning system based on intelligent flexible sensing device according to claim 2, characterized in that, The motion parameter analysis unit is also used for: Based on biomechanical mechanisms, damage-related parameters were screened from the raw motion data, where the biomechanical mechanisms were extracted from the database; Damage-related parameters include mechanical load parameters, kinematic parameters, dynamic parameters, and symmetry parameters; The selected damage-related parameters are subjected to feature extraction, which includes time-domain feature extraction and periodic feature extraction. Temporal feature extraction involves: extracting the key extreme values ​​of the damage-related parameters within the motion cycle, calculating the statistics of the damage-related parameters within the motion cycle, and finally capturing the changing trend of the damage-related parameters over time. Periodic feature extraction involves: segmenting continuous motion into action cycles, extracting parameter features within each cycle after segmentation, calculating the mean and coefficient of variation of multiple cycles, and finally analyzing the parameter change trend of adjacent cycles. The injury-related parameters after feature extraction are integrated into quantitative indicators under motion conditions, including load level quantification, motion quality quantification, and fatigue degree quantification. Among them, load level quantification is as follows: the absolute load index is constructed by using the characteristic values ​​of mechanical parameters, and then normalized by combining human physiological parameters, and relative load index is constructed; movement quality quantification is as follows: the actual movement parameters are compared with the standard movement parameters, and then the variation of the parameters is quantified, and finally the difference of the parameters of the two limbs is calculated; fatigue level quantification is as follows: the change trend of parameters with exercise duration is quantified, and then the synchronous change of multiple parameters is quantified. Ultimately, the key motion parameters were extracted, and the human motion state was quantified.

4. The motion injury state early warning system based on intelligent flexible sensing device according to claim 3, characterized in that, Determining the synchronization quality between eigenvalues ​​and pre-injury warning signals includes: Retrieve the slope value of each feature value within each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data within those time periods; Normalize the slope value of each feature value in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in the time period to obtain the normalized slope value of each feature value and the absolute difference between the maximum and minimum data values ​​in the time period. Retrieve the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum values ​​of the data in that time period; Normalize the slope value of each pre-injury signal in each two adjacent time periods and the absolute difference between the maximum and minimum data values ​​in that time period to obtain the normalized slope value of each pre-injury signal and the absolute difference between the maximum and minimum data values ​​in that time period. Synchronization quality assessment parameters are obtained by combining the normalized slope value corresponding to each feature value in each two adjacent time periods with the absolute difference between the maximum and minimum data values ​​in that time period, and the normalized slope value corresponding to each pre-injury signal with the absolute difference between the maximum and minimum data values ​​in that time period. The synchronization quality assessment parameters are compared with preset assessment thresholds; If the synchronization quality assessment parameter does not exceed the preset assessment threshold, the synchronization quality is determined to be non-compliant, and a synchronization anomaly alarm is triggered.

5. The motion injury state early warning system based on intelligent flexible sensing device according to claim 4, characterized in that, The risk feature extraction unit is also used for: Transform highly sensitive features that have undergone validity verification into quantitative indicators; The process for converting quantitative indicators is as follows: S4: Hierarchical quantification of feature values: First, set the safety value, warning value and danger value of the feature according to the medical safety standard. The medical safety standard is extracted from the database. Then, the feature change rate is calculated in combination with the individual baseline. The duration or frequency of persistent abnormal features is quantified. S5: Multi-feature integration and quantification: Based on the influence weight of different features on the injury, a comprehensive risk value is calculated. At the same time, based on the linkage relationship between features, the mechanism of multi-factor synergy leading to injury in real sports is simulated. Finally, the quantification standard is adjusted in real time according to the duration of sports and the degree of fatigue. S6: Standardization of quantitative indicators: Map the comprehensive quantitative value to high risk, medium risk and low risk, then normalize the characteristic indicators of different types, and finally organize and output the quantitative indicators in the structure of characteristic name, current value, risk level and benchmark value. Finally, the quantitative indicators of sensitive features are transformed.

6. The motion injury state early warning system based on intelligent flexible sensing device according to claim 5, characterized in that, The damage risk assessment unit is also used for: After retrieving risk threshold standards related to the current type of movement and target body part and matching single-index risk thresholds, a virtual movement scene is constructed through a quantum simulator to simulate the tissue stress distribution under different combinations of mechanical parameters. Risk levels are determined based on the range of single-indicator risk thresholds, including low risk, medium risk, and high risk. Low risk means that the single-indicator risk threshold is within the safe range; medium risk means that the single-indicator risk threshold reaches the warning threshold but does not reach the danger threshold; and high risk means that the single-indicator risk threshold reaches or exceeds the danger threshold. Weights are assigned to different dimensions based on the risk level determination results of single-indicator risk thresholds, including core risk dimensions, auxiliary risk dimensions, and temporary weight adjustments in special scenarios; The risk level determination result of the single indicator risk threshold is combined with the corresponding assigned weight to calculate the comprehensive risk value. After the comprehensive risk value is calculated, the overall risk level is obtained. If the overall risk value is less than 1.5, it is judged as low risk overall; if 1.5 ≤ overall risk value < 2.5, it is judged as medium risk overall; if the overall risk value is greater than or equal to 2.5, it is judged as high risk overall. Finally, the level of injury risk under the current motion state is determined.

7. The motion injury state early warning system based on intelligent flexible sensing device according to claim 6, characterized in that, The warning signal generation unit is also used for: The overall risk level is mapped to the early warning strategy templates in the database, and different early warning strategies for different risk levels are obtained after the mapping is completed; Early warning content is generated based on different early warning strategies. The early warning content includes core information, action instructions, and feedback instructions. The core information includes risk characterization, risk positioning, and risk quantification; action instructions provide immediate corrective suggestions for instructions with medium risk overall, and send stop instructions for instructions with high overall risk; feedback instructions generate encouraging information and reinforce correct behavior when the risk level decreases or user actions improve. The generated warning content is transmitted to the user through physical signals via different sensory channels, including tactile, auditory, and visual signals; Among them, the tactile signal is: adjusting the vibration intensity of the intelligent flexible sensing device according to the risk level, and using different vibration modes to distinguish information; the auditory signal is: using different frequencies of sound, for medium and high risks, directly playing the generated text content through headphones or device speakers; the visual signal is: displaying the warning content on the user terminal with icons and text simultaneously, and using color to distinguish the risk level. When multiple risks occur simultaneously or signal channels are occupied, high-risk warnings have the highest interruption priority. Finally, based on the target terminal, signal mode, specific content, and duration, a warning signal corresponding to the damage risk level is obtained; The warning instruction transmission unit is also used for: Structured data reorganized into: S7: Adopts a hierarchical structure: the top layer contains basic control fields, the middle layer contains core risk information, and the bottom layer contains content data; S8: Unified Data Identification: Add standardized labels to each field; S9: Compress redundant information: Simplify repetitive or minor content and retain differentiated data; The structured data is adjusted according to the type, status, and communication environment of the user terminal; The adjustment of structured data is as follows: adapt terminal characteristics according to the type of user terminal, select the corresponding transmission path according to the communication method, configure communication parameters, and finally encrypt the transmitted structured data. After the structured data is adjusted, it is transmitted to the corresponding user terminal.

8. The motion injury state early warning system based on intelligent flexible sensing device according to claim 7, characterized in that, The warning instruction transmission unit is also used for: At the same time, when structured data is transmitted to the user terminal, the fastest transmission channel is automatically selected. First, confirm the amount of structured data transmitted. The amount of data transmitted can be found in the data information database. Real-time monitoring of the remaining capacity and signal strength of each transmission channel; Select a transmission channel with a data transmission volume less than the remaining signal capacity as the first transmission channel for structured data. When there is more than one first transmission channel, the transmission channel with the strongest channel signal strength is selected as the second transmission channel for structured data. Choose either the first transmission channel or the second transmission channel as the communication channel for transmitting structured data to the user terminal; The user terminal personalization management unit is also used for: After receiving the warning signal, the user terminal first integrates the data to obtain a real-time user profile, which includes real-time warning instructions, basic user status, and historical motion data. Personalized suggestions are generated based on real-time user profiles. These personalized suggestions include the generation of immediate response strategies, planning of short-term adjustment plans, and recommendations for long-term improvement plans. The system includes real-time response strategy generation, which consists of high-risk and medium-risk warnings. High-risk warnings are mandatory recommendations, while medium-risk warnings are corrective recommendations. Short-term adjustment plan planning combines real-time warnings with user goals and plans subsequent actions. Long-term improvement plan recommendations are based on recurring error patterns, automatically generating long-term suggestions to address the root causes, while also considering the user's medical history. The generated personalized suggestions undergo a feasibility check. Once the feasibility check is completed and the suggestions pass, they are displayed on the user's terminal.

Images