A method and system for dynamic monitoring of motion behavior based on artificial intelligence

Abstract

This invention discloses an artificial intelligence-based dynamic monitoring method and system for motion behavior, belonging to the field of dynamic monitoring technology. After user authorization, it acquires the user's acceleration, angular velocity, and pressure data, and synchronizes them with timestamps. A timestamp set is constructed, and the position and angular posture of a single user at a single timestamp are calculated. The motion state index of a single user at a single timestamp is calculated, and combined with pressure data, the motion-force index value of a single user at a single timestamp is calculated. Based on the motion-force index value, a motion-force threshold Boolean matrix is ​​constructed, and the similarity between different users is calculated. The number of times a single user exhibits abnormal motion behavior across all timestamps is calculated, a preset threshold is established, and the analysis and output of warnings are performed. This invention achieves high-precision dynamic monitoring and anomaly detection of motion behavior, enabling timely identification of potential motion risks, improving motion safety, and optimizing motion behavior management.

Description

Technical Field

[0001] This invention relates to the field of dynamic monitoring technology, specifically to a method and system for dynamic monitoring of motion behavior based on artificial intelligence. Background Technology

[0002] In traditional methods, researchers primarily rely on optical tracking systems, wearable sensors, and biomechanical measurement devices to collect human motion data and then assess motion behavior through mathematical modeling and statistical analysis. In recent years, with the rapid development of artificial intelligence technology, motion behavior monitoring methods based on deep learning and machine learning have gradually become a research hotspot. By integrating sensor data and computer vision technology, existing research has been able to achieve real-time monitoring of motion trajectories, posture changes, and force conditions, playing an important role in multiple fields such as sports training, rehabilitation medicine, and health management. Furthermore, the widespread adoption of smart wearable devices has further promoted the personalization and portability of motion behavior monitoring technology, making real-time, non-invasive data collection and analysis possible.

[0003] Despite the progress made in motion behavior monitoring, existing technologies still have many shortcomings. First, traditional monitoring systems based on optical tracking and fixed mechanical measurements are expensive and limited to fixed scenarios, making it difficult to meet the motion monitoring needs of daily life. Second, while existing smart wearable devices can collect motion data, they often lack the ability to fuse multimodal information, resulting in inaccurate comprehensive analysis of motion state and force conditions. In addition, existing methods for detecting abnormal motion behavior mainly rely on preset rules or simple pattern matching, which cannot fully adapt to individual differences, leading to large errors in abnormal identification. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for dynamic monitoring of motion behavior based on artificial intelligence, so as to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] An AI-based dynamic monitoring method for motion behavior includes the following steps: Step S1: After user authorization, acquire the user's acceleration data, angular velocity data, and pressure data, and synchronize and align the timestamps; Step S2: Construct a timestamp set; calculate the position and angular posture of a single user at a single timestamp based on the acceleration data, angular velocity data, and pressure data; Step S3: Calculate the motion state index of a single user at a single timestamp based on the position and angular posture; calculate the motion-force index value of a single user at a single timestamp based on the motion state index and pressure data; Step S4: Construct a motion-force threshold Boolean matrix based on the motion-force threshold Boolean matrix; calculate the similarity between different users based on the motion-force threshold Boolean matrix; calculate the number of times a single user exhibits abnormal motion behavior across all timestamps, preset a threshold, analyze the data, and output an early warning.

[0007] As a preferred embodiment of the AI-based dynamic monitoring method for motion behavior described in this invention, after user authorization, a smart wearable device and a motion assist device are deployed on the user's body surface. The smart wearable device includes a smart bracelet, and the motion assist device includes a pressure patch. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect the user's pressure data during exercise.

[0008] Using the Bluetooth Mesh protocol, a communication connection is established between the smart bracelet and the pressure patch, and the acceleration data, angular velocity data and pressure data are synchronized and aligned on the timestamp.

[0009] As a preferred embodiment of the motion behavior dynamic monitoring method based on artificial intelligence described in this invention, a timestamp set is constructed, denoted as TY = {TY...} t |t∈[1,T]}, where TY t This represents the t-th timestamp, where T represents the total number of timestamps. The user will be assigned a time stamp at timestamp TY. t The acceleration, angular velocity, and pressure data are denoted as AD(TY) respectively. t ), SCD(TY) t ) and SD(TY) t ).

[0010] Based on the user's timestamp TY t The acceleration data AD(TY) below t ), angular velocity data SCD(TY) t ), calculate timestamp TY t The user's position and angle are as follows:

[0011] v(TY t )=v(TY t-1 )+AD(TY t )×ΔTY t,t-1 ;

[0012] s(TY t ) = s(TY t-1 )+v(TY t )×ΔTY t,t-1 ;

[0013] θ(TY t )=θ(TY t-1 )+SCD(TY t )×ΔTY t,t-1 ;

[0014] Among them, v(TY) t This indicates that the user is at the timestamp TY. t The speed of the downward movement, v(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The speed of ΔTY t,t-1 Indicates timestamp TY t With timestamp TY t-1 The interval between them, s(TY) t This indicates that the user is at the timestamp TY. t The position below, s(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The position below, θ(TY) t This indicates that the user is at the timestamp TY. t The angle and attitude below, θ(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The initial angle and attitude are zero, the initial velocity v(TY0) is zero, and the initial angle and attitude θ(TY0) is zero degrees.

[0015] As a preferred embodiment of the AI-based dynamic monitoring method for motion behavior described in this invention, based on the user's timestamp TY... t The position below s(TY) t ) and users at timestamp TY t The angle and attitude θ (TY) t ), calculate the user's timestamp TY t The motion state index is calculated using the following formula:

[0016] MSI(TY t )=α×||s(TY t )-s(TY t-1 )||+β×|θ(TY t )-θ(TY t-1 )|;

[0017] Among them, MSI(TY) t This indicates that the user is at the timestamp TY. t The motion state index is given by α, which represents the preset position weight factor (unit is 1 / m), β, which represents the preset angle and attitude weight factor (unit is 1 / degree), and ||·|| represents the norm.

[0018] It should be noted that ||s(TY) t )-s(TY t-1 || represents the change in displacement, |θ(TY) t )-θ(TY t-1 | Represents the change in posture angle, as indicated by the user's timestamp TY. t The Motion State Index (MSI) under the following conditions t This can reflect the user's range of motion; the larger the value, the more drastic the change in motion. The position s(TY) here... t A is spatial location information, which is usually a three-dimensional coordinate. Since the trajectory is a three-dimensional vector, it is more reasonable to use the vector norm ||·|| to calculate the displacement.

[0019] Based on the user's timestamp TY t The Motion State Index (MSI) under the following conditions t ), combined with the user's timestamp TY t The pressure data SD(TY) t ), calculate the user's timestamp TY t The force index value under motion is calculated using the following formula:

[0020]

[0021] Among them, MFI(TY) t This indicates that the user is at the timestamp TY. t The motion-force index value is given below, where γ represents the preset motion state index influence factor, δ represents the preset pressure data influence factor, and SD... ref This indicates the preset maximum pressure value that the human body can withstand.

[0022] It should be noted that if the motion-force index value MFI(TY) t A high MFI (Motor-Force Index) value indicates vigorous exercise and significant force exertion, potentially representing high-intensity activities such as jumping or sprinting. t A low MSI (Motion State Index) indicates small range of motion and low force, possibly indicating a stationary state or slow walking; a low MSI (Motion State Index) indicates a low MSI (Motion State Index) indicates a low MSI (Motion State Index) indicating ... t High but pressure data SD(TY) tLow: May indicate a suspended state, such as jumping or falling; if the pressure data SD(TY) t High but the MSI (Type II) of physical condition is high. t Low: may indicate static force, such as standing or squatting.

[0023] As a preferred embodiment of the AI-based dynamic monitoring method for motion behavior described in this invention, based on the user's timestamp TY... t Motion under the force index MFI (TY) t ), will the i-th user at timestamp TY t The motion-force index value under these conditions is denoted as MFI. i (TY t And construct the motion-force overthreshold Boolean matrix, as follows:

[0024]

[0025] Where B[i,t] represents the Boolean value in the i-th row and t-th column of the motion-force threshold Boolean matrix, that is, the value of the i-th user at timestamp TY. t Whether the value below is greater than or equal to the threshold value of the motion-force index is determined by the following Boolean value B[i,t]:

[0026]

[0027] Where τ represents the preset motion-force index threshold, if the i-th user is at timestamp TY t Motion-Force Index (MFI) under the following conditions i (TY t If the value of the motion-force index is greater than or equal to the threshold value τ, then it is determined that the i-th user is at timestamp TY. t Abnormal motor behavior exists below.

[0028] Based on the motion-force overthreshold Boolean matrix, the similarity between the i-th user and the (i+1)-th user is calculated using the following formula:

[0029]

[0030] Where S(B[i],B[i+1]) represents the similarity between the i-th user and the (i+1)-th user, B[i+1,t] represents the Boolean value in the (i+1)-th row and t-th column of the motion-force threshold Boolean matrix, I(·) represents the indicator function, which is 1 if B[i,t]≠B[i+1,t] and 0 if B[i,t]=B[i+1,t], and T represents the total number of timestamps.

[0031] It should be noted that the similarity value ranges from [0, 1]. When the superthreshold vectors of two users are completely identical (i.e., ... ), with a similarity of 1, when two users are completely different (i.e. The similarity is 0.

[0032] A preset similarity threshold (the similarity threshold belongs to [0, 1]) is set. If the similarity S(B[i], B[i+1]) between the i-th user and the (i+1)-th user is greater than or equal to the similarity threshold, then it is determined that the i-th user and the (i+1)-th user are similar.

[0033] The formula for calculating the number of times the i-th user exhibits abnormal movement behavior across all T timestamps is as follows: Where, N i This represents the number of times the i-th user exhibits abnormal movement behavior.

[0034] If the number of times the i-th user exhibits abnormal movement behavior is greater than or equal to a preset threshold, then the i-th user's movement behavior is determined to be highly abnormal. In this case, all users similar to the i-th user are retrieved, an alert is issued to the staff, and acceleration data, angular velocity data, and pressure data are acquired in real time for dynamic monitoring.

[0035] An AI-based dynamic monitoring system for motion behavior includes: a data acquisition module, a position and angle attitude calculation module, a motion state index and motion-force index value calculation module, and a matrix construction and analysis module.

[0036] The data acquisition module, after user authorization, acquires the user's acceleration data, angular velocity data, and pressure data, and performs timestamp synchronization and alignment.

[0037] The position and angle attitude calculation module: constructs a timestamp set; and calculates the position and angle attitude of a single user at a single timestamp based on acceleration data, angular velocity data, and pressure data.

[0038] The motion state index and motion-force index calculation module calculates the motion state index of a single user at a single timestamp based on position and angle posture; and calculates the motion-force index value of a single user at a single timestamp based on the motion state index and pressure data.

[0039] The matrix construction and analysis module: constructs a motion-force overthreshold Boolean matrix based on the motion-force index value; calculates the similarity between different users based on the motion-force overthreshold Boolean matrix; calculates the number of times a single user exhibits abnormal motion behavior across all timestamps, sets a preset threshold, analyzes and outputs a warning.

[0040] Furthermore, the data acquisition module includes a data acquisition unit.

[0041] The data acquisition unit, after user authorization, deploys a smart wearable device and a sports assistive device on the user's body surface. The smart wearable device includes a smart bracelet, and the sports assistive device includes a pressure patch. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect the user's pressure data during exercise. A communication connection is established between the smart bracelet and the pressure patch using the Bluetooth Mesh protocol, and the acceleration data, angular velocity data, and pressure data are synchronized and aligned in terms of timestamps.

[0042] Furthermore, the position and angle attitude calculation module includes a position and angle attitude calculation unit.

[0043] The position and angle attitude calculation unit: constructs a timestamp set, and calculates the user's position and angle attitude at a single timestamp based on the user's acceleration and angular velocity data at a single timestamp.

[0044] Furthermore, the motion state index and motion-force index value calculation module includes a motion state index calculation unit and a motion-force index value calculation unit.

[0045] The motion state index calculation unit calculates the user's motion state index at a single timestamp based on the user's position and angle posture at that single timestamp.

[0046] The motion-force index calculation unit calculates the motion-force index value of a single user at a single time point based on the motion state index of a single user at a single time point and the pressure data of a single user at a single time point.

[0047] Furthermore, the matrix construction and analysis module includes a matrix construction unit and an analysis unit.

[0048] The matrix construction unit: Based on the motion-force index value of a single user at a single timestamp, a motion-force overthreshold Boolean matrix is ​​constructed.

[0049] The analysis unit calculates the similarity between different users based on the motion-force threshold Boolean matrix; it presets a similarity threshold, and if the similarity between different users is greater than or equal to the similarity threshold, it determines that the users are similar; it calculates the number of times a user exhibits abnormal motion behavior across all timestamps; if the number of times a user exhibits abnormal motion behavior is greater than or equal to a preset threshold, it determines that the user's motion behavior is highly abnormal, it then obtains all other users similar to the user, issues an alert to staff, and acquires acceleration data, angular velocity data, and pressure data in real time for dynamic monitoring.

[0050] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: The AI-based dynamic motion behavior monitoring method and system provided by this invention utilizes a smart bracelet and pressure patches to collect the user's acceleration, angular velocity, and pressure data. Data synchronization and alignment are achieved using the Bluetooth Mesh protocol, ensuring data temporal consistency and improving the accuracy of subsequent calculations. Based on the collected data, the system calculates the user's position and angular posture at different timestamps and constructs a timestamp set for subsequent dynamic analysis. This process enables the system to accurately locate the user's motion state, providing a foundation for subsequent index calculations. The motion state index is calculated using position and angular posture, and then combined with pressure data to calculate the motion-force index value. This process reflects the relationship between the user's motion behavior and force conditions, thus aiding in the assessment of motion risk. Based on the calculated motion-force index value, a motion-force threshold Boolean matrix is ​​constructed, and the similarity of motion behavior between different users is calculated. Simultaneously, the number of abnormal motion behaviors is counted. If the number of abnormal behaviors exceeds a set threshold, the user is determined to have highly abnormal motion behavior. Further analysis of its similarity to other users is conducted, ultimately issuing an early warning to staff and monitoring subsequent data in real time. This invention achieves high-precision dynamic monitoring and anomaly detection of movement behavior, enabling timely identification of potential movement risks, improving movement safety, and optimizing movement behavior management. Attached Figure Description

[0051] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0052] Figure 1 This is a schematic diagram illustrating the steps of a motion behavior dynamic monitoring method based on artificial intelligence according to the present invention;

[0053] Figure 2 This is a schematic diagram of the structure of a motion behavior dynamic monitoring system based on artificial intelligence according to the present invention. Detailed Implementation

[0054] 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.

[0055] Please see Figure 1 In this first embodiment: a method for dynamic monitoring of motion behavior based on artificial intelligence is provided, which includes the following steps:

[0056] Step S1: After user authorization, obtain the user's acceleration data, angular velocity data, and pressure data, and synchronize and align the timestamps.

[0057] Specifically, after user authorization, smart wearable devices and sports assistive devices are deployed on the user's body surface. The smart wearable devices include smart bracelets, and the sports assistive devices include pressure patches. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect the user's pressure data during exercise.

[0058] Furthermore, a communication connection is established between the smart bracelet and the pressure patch using the Bluetooth Mesh protocol, and the acceleration data, angular velocity data, and pressure data are synchronized and aligned on the timestamp.

[0059] In this invention, the combination of smart wearable devices and motion-assistive devices enables synchronous acquisition of multi-source data, ensuring data temporal consistency and improving the accuracy of subsequent calculations. The Bluetooth Mesh protocol is used to establish a communication connection, improving data transmission stability and low power consumption, thus enhancing the system's real-time performance and applicability. This step provides accurate foundational data for subsequent motion state calculations, ensuring the system can accurately capture key features during motion.

[0060] Step S2: Construct a timestamp set; based on acceleration data, angular velocity data, and pressure data, calculate the position and angular attitude of a single user at a single timestamp.

[0061] Specifically, construct a timestamp set, denoted as TY = {TY t |t∈[1,T]}, where TY t This represents the t-th timestamp, where T represents the total number of timestamps. The user will be assigned a time stamp at timestamp TY. t The acceleration, angular velocity, and pressure data are denoted as AD(TY) respectively. t ), SCD(TY) t ) and SD(TY) t ).

[0062] Furthermore, based on the user's timestamp TY t The acceleration data AD(TY) below t ), angular velocity data SCD(TY) t ), calculate timestamp TY t The user's position and angle are as follows:

[0063] v(TY t )=v(TY t-1)+AD(TY t )×ΔTY t,t-1 ;

[0064] s(TY t ) = s(TY t-1 )+v(TY t )×ΔTY t,t-1 ;

[0065] θ(TY t )=θ(TY t-1 )+SCD(TY t )×ΔTY t,t-1 ;

[0066] Among them, v(TY) t This indicates that the user is at the timestamp TY. t The speed of the downward movement, v(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The speed of ΔTY t,t-1 Indicates timestamp TY t With timestamp TY t-1 The interval between them, s(TY) t This indicates that the user is at the timestamp TY. t The position below, s(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The position below, θ(TY) t This indicates that the user is at the timestamp TY. t The angle and attitude below, θ(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The initial angle and attitude are zero, the initial velocity v(TY0) is zero, and the initial angle and attitude θ(TY0) is zero degrees.

[0067] In this invention, the user's position and angular posture are calculated through integration, enabling the system to accurately reconstruct the motion trajectory and improve the precision of motion behavior analysis. By combining data from multiple timestamps, the user's motion trajectory can be dynamically tracked, providing necessary information support for subsequent motion state calculations. By constructing a timestamp set, time-series processing of the data is achieved, making it easier to quantify and analyze the changing trends of motion states, thus improving the system's predictive capabilities.

[0068] Step S3: Based on position and angle posture, calculate the motion state index of a single user at a single timestamp; based on the motion state index and combined with pressure data, calculate the motion-force index value of a single user at a single timestamp.

[0069] Specifically, based on the user's timestamp TY t The position below s(TY) t ) and users at timestamp TYt The angle and attitude θ (TY) t ), calculate the user's timestamp TY t The motion state index is calculated using the following formula:

[0070] MSI(TY t )=α×||s(TY t )-s(TY t-1 )||+β×|θ(TY t )-θ(TY t-1 )|;

[0071] Among them, MSI(TY) t This indicates that the user is at the timestamp TY. t The motion state index is given by α, which represents the preset position weight factor (unit is 1 / m), β, which represents the preset angle and attitude weight factor (unit is 1 / degree), and ||·|| represents the norm.

[0072] Furthermore, based on the user's timestamp TY t The Motion State Index (MSI) under the following conditions t ), combined with the user's timestamp TY t The pressure data SD(TY) t ), calculate the user's timestamp TY t The force index value under motion is calculated using the following formula:

[0073]

[0074] Among them, MFI(TY) t This indicates that the user is at the timestamp TY. t The motion-force index value is given below, where γ represents the preset motion state index influence factor, δ represents the preset pressure data influence factor, and SD... ref This indicates the preset maximum pressure value that the human body can withstand.

[0075] In this invention, by calculating the motion state index, the system can accurately measure the user's motion intensity and posture stability at a specific time point, providing a high-precision indicator for anomaly detection. Combining pressure data with the calculation of the motion-force index value enables a quantitative analysis of the force situation during motion, allowing the system to not only analyze the motion state but also consider the impact of the external environment on motion safety, such as overload pressure or impact force. This indexation processing method provides more intuitive data support for anomaly detection, helping the system quickly identify potential risks and improve the accuracy and adaptability of monitoring.

[0076] Step S4: Based on the motion-force index value, construct the motion-force threshold Boolean matrix; based on the motion-force threshold Boolean matrix, calculate the similarity between different users; calculate the number of times a single user exhibits abnormal motion behavior across all timestamps, preset a threshold, analyze and output an early warning.

[0077] Specifically, based on the user's timestamp TY t Motion under the force index MFI (TY) t ), will the i-th user at timestamp TY t The motion-force index value under these conditions is denoted as MFI. i (TY t And construct the motion-force overthreshold Boolean matrix, as follows:

[0078]

[0079] Where B[i,t] represents the Boolean value in the i-th row and t-th column of the motion-force threshold Boolean matrix, that is, the value of the i-th user at timestamp TY. t Whether the value below is greater than or equal to the threshold value of the motion-force index is determined by the following Boolean value B[i,t]:

[0080]

[0081] Where τ represents the preset motion-force index threshold, if the i-th user is at timestamp TY t Motion-Force Index (MFI) under the following conditions i (TY t If the value of the motion-force index is greater than or equal to the threshold value τ, then it is determined that the i-th user is at timestamp TY. t Abnormal motor behavior exists below.

[0082] Furthermore, based on the motion-force overthreshold Boolean matrix, the similarity between the i-th user and the (i+1)-th user is calculated using the following formula:

[0083]

[0084] Where S(B[i],B[i+1]) represents the similarity between the i-th user and the (i+1)-th user, B[i+1,t] represents the Boolean value in the (i+1)-th row and t-th column of the motion-force threshold Boolean matrix, I(·) represents the indicator function, which is 1 if B[i,t]≠B[i+1,t] and 0 if B[i,t]=B[i+1,t], and T represents the total number of timestamps.

[0085] Furthermore, a preset similarity threshold (the similarity threshold belongs to [0, 1]) is set. If the similarity S(B[i], B[i+1]) between the i-th user and the (i+1)-th user is greater than or equal to the similarity threshold, then it is determined that the i-th user and the (i+1)-th user are similar.

[0086] The formula for calculating the number of times the i-th user exhibits abnormal movement behavior across all T timestamps is as follows: Where, N i This represents the number of times the i-th user exhibits abnormal movement behavior.

[0087] Furthermore, if the number of times the i-th user exhibits abnormal movement behavior is greater than or equal to a preset threshold, then the i-th user's movement behavior is determined to be highly abnormal. In this case, all users similar to the i-th user are identified, and an alert is issued to the staff. Acceleration data, angular velocity data, and pressure data are also acquired in real time for dynamic monitoring.

[0088] In this invention, by constructing a motion-force overthreshold Boolean matrix, motion behavior data is converted into a structured binary matrix, making anomaly detection more efficient and adaptable to large-scale data processing. Calculating the similarity between users enables the system to detect not only abnormal movements of individual users but also potential behavioral patterns among multiple users, aiding in team motion analysis or group health management. Setting anomaly thresholds for motion behavior and combining this with analysis of similar user data improves the accuracy of early warnings, reduces false alarm rates, and enhances the system's intelligence. Through real-time monitoring and early warning mechanisms, the system can respond instantly when motion anomalies occur, providing timely safety assurance for users.

[0089] Please see Figure 2 In this second embodiment, an artificial intelligence-based dynamic monitoring system for motion behavior is provided. The system includes: a data acquisition module, a position and angle attitude calculation module, a motion state index and motion-force index value calculation module, and a matrix construction and analysis module.

[0090] The data acquisition module, after user authorization, acquires the user's acceleration data, angular velocity data, and pressure data, and performs timestamp synchronization and alignment.

[0091] The position and angle attitude calculation module: constructs a timestamp set; and calculates the position and angle attitude of a single user at a single timestamp based on acceleration data, angular velocity data, and pressure data.

[0092] The motion state index and motion-force index calculation module calculates the motion state index of a single user at a single timestamp based on position and angle posture; and calculates the motion-force index value of a single user at a single timestamp based on the motion state index and pressure data.

[0093] The matrix construction and analysis module: constructs a motion-force overthreshold Boolean matrix based on the motion-force index value; calculates the similarity between different users based on the motion-force overthreshold Boolean matrix; calculates the number of times a single user exhibits abnormal motion behavior across all timestamps, sets a preset threshold, analyzes and outputs a warning.

[0094] Furthermore, the data acquisition module includes a data acquisition unit.

[0095] The data acquisition unit, after user authorization, deploys a smart wearable device and a sports assistive device on the user's body surface. The smart wearable device includes a smart bracelet, and the sports assistive device includes a pressure patch. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect the user's pressure data during exercise. A communication connection is established between the smart bracelet and the pressure patch using the Bluetooth Mesh protocol, and the acceleration data, angular velocity data, and pressure data are synchronized and aligned in terms of timestamps.

[0096] Furthermore, the position and angle attitude calculation module includes a position and angle attitude calculation unit.

[0097] The position and angle attitude calculation unit: constructs a timestamp set, and calculates the user's position and angle attitude at a single timestamp based on the user's acceleration and angular velocity data at a single timestamp.

[0098] Furthermore, the motion state index and motion-force index value calculation module includes a motion state index calculation unit and a motion-force index value calculation unit.

[0099] The motion state index calculation unit calculates the user's motion state index at a single timestamp based on the user's position and angle posture at that single timestamp.

[0100] The motion-force index calculation unit calculates the motion-force index value of a single user at a single time point based on the motion state index of a single user at a single time point and the pressure data of a single user at a single time point.

[0101] Furthermore, the matrix construction and analysis module includes a matrix construction unit and an analysis unit.

[0102] The matrix construction unit: Based on the motion-force index value of a single user at a single timestamp, a motion-force overthreshold Boolean matrix is ​​constructed.

[0103] The analysis unit calculates the similarity between different users based on the motion-force threshold Boolean matrix; it presets a similarity threshold, and if the similarity between different users is greater than or equal to the similarity threshold, it determines that the users are similar; it calculates the number of times a user exhibits abnormal motion behavior across all timestamps; if the number of times a user exhibits abnormal motion behavior is greater than or equal to a preset threshold, it determines that the user's motion behavior is highly abnormal, it then obtains all other users similar to the user, issues an alert to staff, and acquires acceleration data, angular velocity data, and pressure data in real time for dynamic monitoring.

[0104] 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.

[0105] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for dynamic monitoring of motion behavior based on artificial intelligence, characterized in that, The method includes the following steps: Step S1: After user authorization, obtain the user's acceleration data, angular velocity data, and pressure data, and synchronize and align the timestamps. Step S2: Construct a timestamp set; based on acceleration data, angular velocity data, and pressure data, calculate the position and angular attitude of a single user at a single timestamp; Step S3: Based on the user's t-th timestamp TY t The position below s(TY) t ) and users at timestamp TY t The angle and attitude θ (TY) t ), calculate the user's timestamp TY t The motion state index is calculated using the following formula: MSI(TY t )=α×||s(TY t )-s(TY t-1 )||+β×|θ(TY t )-θ(TY t-1 )|; Among them, MSI(TY) t This indicates that the user is at the timestamp TY. t The motion state index is given by α, where α represents the preset position weight factor, β represents the preset angle and attitude weight factor, ||·|| represents the norm, and s(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The position below, θ(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The angle and posture below; Based on the user's timestamp TY t The Motion State Index (MSI) under the following conditions t ), combined with the user's timestamp TY t The pressure data SD(TY) t ), calculate the user's timestamp TY t The force index value under motion is calculated using the following formula: Among them, MFI(TY) t This indicates that the user is at the timestamp TY. t The motion-force index value is given below, where γ represents the preset motion state index influence factor, δ represents the preset pressure data influence factor, and SD... ref This indicates the preset maximum pressure value that the human body can withstand; Step S4: Based on the user's timestamp TY t Motion under the force index MFI (TY) t ), will the i-th user at timestamp TY t The motion-force index value under these conditions is denoted as MFI. i (TY t And construct the motion-force overthreshold Boolean matrix, as follows: Where B[i,t] represents the Boolean value in the i-th row and t-th column of the motion-force threshold Boolean matrix, that is, the value of the i-th user at timestamp TY. t Whether the value below is greater than or equal to the threshold value of the motion-force index is determined by the following Boolean value B[i,t]: Where τ represents the preset motion-force index threshold, if the i-th user is at timestamp TY t Motion-Force Index (MFI) under the following conditions i (TY t If the value of the motion-force index is greater than or equal to the threshold value τ, then it is determined that the i-th user is at timestamp TY. t The system detects abnormal motion behavior; it calculates the similarity between different users based on the motion-force threshold Boolean matrix; it calculates the number of times a single user exhibits abnormal motion behavior across all timestamps, sets a preset threshold, analyzes the data, and outputs an alert.

2. The method for dynamic monitoring of motion behavior based on artificial intelligence according to claim 1, characterized in that, The specific implementation process of step S1 includes: After user authorization, smart wearable devices and sports assistive devices are deployed on the user's body surface. The smart wearable devices include a smart bracelet, and the sports assistive devices include a pressure patch. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect the user's pressure data during exercise. Using the Bluetooth Mesh protocol, a communication connection is established between the smart bracelet and the pressure patch, and the acceleration data, angular velocity data and pressure data are synchronized and aligned on the timestamp.

3. The method for dynamic monitoring of motion behavior based on artificial intelligence according to claim 2, characterized in that, The specific implementation process of step S2 includes: Construct a timestamp set, denoted as TY = {TY t |t∈[1,T]}, where TY t This represents the t-th timestamp, where T represents the total number of timestamps. The user will be assigned a time stamp at timestamp TY. t The acceleration, angular velocity, and pressure data are denoted as AD(TY) respectively. t ), SCD(TY) t ) and SD(TY) t ); Based on the user's timestamp TY t The acceleration data AD(TY) below t ), angular velocity data SCD(TY) t ), calculate timestamp TY t The user's position and angle are as follows: v(TY t )=v(TY t-1 )+AD(TY t )×ΔTY t,t-1 ; s(TY t )=s(TY t-1 )+v(TY t )×ΔTY t,t-1 ; θ(TY t )=θ(TY t-1 )+SCD(TY t )×ΔTY t,t-1 ; Among them, v(TY) t This indicates that the user is at the timestamp TY. t The speed of the downward movement, v(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The speed of ΔTY t,t-1 Indicates timestamp TY t With timestamp TY t-1 The interval between them, s(TY) t This indicates that the user is at the timestamp TY. t The position below, s(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The position below, θ(TY) t This indicates that the user is at the timestamp TY. t The angle and attitude below, θ(TY) t-1 This indicates that the user is at the timestamp TY. t-1 The initial angle and attitude are zero, the initial velocity v(TY0) is zero, and the initial angle and attitude θ(TY0) is zero degrees.

4. The method for dynamic monitoring of motion behavior based on artificial intelligence according to claim 3, characterized in that, Based on the motion-force overthreshold Boolean matrix, the similarity between the i-th user and the (i+1)-th user is calculated using the following formula: Where S(B[i],B[i+1]) represents the similarity between the i-th user and the (i+1)-th user, B[i+1,t] represents the Boolean value in the (i+1)-th row and t-th column of the motion-force threshold Boolean matrix, I(·) represents the indicator function, which is 1 if B[i,t]≠B[i+1,t] and 0 if B[i,t]=B[i+1,t], and T represents the total number of timestamps; A preset similarity threshold is set. If the similarity S(B[i],B[i+1]) between the i-th user and the (i+1)-th user is greater than or equal to the similarity threshold, then the i-th user and the (i+1)-th user are considered to be similar. The formula for calculating the number of times the i-th user exhibits abnormal movement behavior across all T timestamps is as follows: Where, N i This indicates the number of times the i-th user exhibits abnormal movement behavior; If the number of times the i-th user exhibits abnormal movement behavior is greater than or equal to a preset threshold, then the i-th user's movement behavior is determined to be highly abnormal. In this case, all users similar to the i-th user are retrieved, an alert is issued to the staff, and acceleration data, angular velocity data, and pressure data are acquired in real time for dynamic monitoring.

5. A motion behavior dynamic monitoring system based on artificial intelligence, executing the motion behavior dynamic monitoring method based on artificial intelligence as described in any one of claims 1-4, characterized in that, The system includes: a data acquisition module, a position and angle attitude calculation module, a motion state index and motion-force index value calculation module, and a matrix construction and analysis module; The data acquisition module: after user authorization, acquires the user's acceleration data, angular velocity data, and pressure data, and performs timestamp synchronization and alignment; The position and angle attitude calculation module: constructs a timestamp set; calculates the position and angle attitude of a single user at a single timestamp based on acceleration data, angular velocity data, and pressure data; The motion state index and motion-force index calculation module calculates the motion state index of a single user at a single timestamp based on position and angle posture; and calculates the motion-force index of a single user at a single timestamp based on the motion state index and pressure data. The matrix construction and analysis module: constructs a motion-force overthreshold Boolean matrix based on the motion-force index value; calculates the similarity between different users based on the motion-force overthreshold Boolean matrix; calculates the number of times a single user exhibits abnormal motion behavior across all timestamps, sets a preset threshold, analyzes and outputs a warning.

6. The motion behavior dynamic monitoring system based on artificial intelligence according to claim 5, characterized in that: The data acquisition module includes a data acquisition unit; The data acquisition unit: after user authorization, deploys smart wearable devices and sports assistive devices on the user's body surface. The smart wearable devices include smart bracelets, and the sports assistive devices include pressure patches. The smart bracelet has a built-in accelerometer sensor and a gyroscope. The accelerometer sensor is used to collect the user's acceleration data during exercise, and the gyroscope is used to acquire the user's angular velocity data during exercise. The pressure patch is used to collect pressure data from the user during exercise; Using the Bluetooth Mesh protocol, a communication connection is established between the smart bracelet and the pressure patch, and the acceleration data, angular velocity data and pressure data are synchronized and aligned on the timestamp.

7. The motion behavior dynamic monitoring system based on artificial intelligence according to claim 6, characterized in that: The position and angle attitude calculation module includes a position and angle attitude calculation unit; The position and angle attitude calculation unit: constructs a timestamp set, and calculates the user's position and angle attitude at a single timestamp based on the user's acceleration and angular velocity data at a single timestamp.

8. The motion behavior dynamic monitoring system based on artificial intelligence according to claim 7, characterized in that: The motion state index and motion-force index value calculation module includes a motion state index calculation unit and a motion-force index value calculation unit; The motion state index calculation unit calculates the motion state index of a user at a single timestamp based on the user's position and angle posture at a single timestamp. The motion-force index calculation unit calculates the motion-force index value of a single user at a single time point based on the motion state index of a single user at a single time point and the pressure data of a single user at a single time point.

9. The motion behavior dynamic monitoring system based on artificial intelligence according to claim 8, characterized in that: The matrix construction and analysis module includes a matrix construction unit and an analysis unit; The matrix construction unit: constructs a motion-force overthreshold Boolean matrix based on the motion-force index value of a single user at a single timestamp; The analysis unit calculates the similarity between different users based on the motion-force overthreshold Boolean matrix. A preset similarity threshold is set. If the similarity between different users is greater than or equal to the similarity threshold, then the users are considered to be similar. Calculate the number of times a user exhibits abnormal movement behavior across all timestamps; If a user exhibits abnormal movement behavior more than or equal to a preset threshold, the user's movement behavior is determined to be highly abnormal. In this case, all other users with similar behavior are identified, and an alert is issued to staff. Acceleration, angular velocity, and pressure data are also acquired in real time for dynamic monitoring.

Images