Motion information service and detection system based on Internet of Things
By leveraging IoT sensors and machine learning technologies, a sports information service and detection system has been built, solving the problem of sports data acquisition bias, enabling high-precision sports information analysis and personalized services, and improving the safety and user satisfaction of the sports process.
Patent Information
- Application Number
- CN202511700852.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-19
- Publication Date
- 2026-02-13
AI Technical Summary
During exercise, sensor data is affected by environmental interference, leading to deviations in motion data acquisition and noise interference, which makes it impossible to guarantee the accuracy and precision of motion information services.
The system uses IoT sensor units to collect multi-source motion data, cleans and standardizes the data through a data preprocessing unit, builds motion pattern models using a machine learning analysis unit, generates customized training suggestions through a personalized service module, and combines real-time monitoring and system optimization modules to perform risk assessment and adaptive parameter adjustment, thereby achieving comprehensiveness and consistency in data collection.
It achieves high-precision motion information analysis in various motion scenarios, reduces motion state recognition errors, improves the reliability of motion information services and user satisfaction, and ensures the safety and personalized guidance of the exercise process.
Smart Images
Figure CN121513432A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of Internet of Things (IoT) technology, specifically to a motion information service and detection system based on IoT. Background Technology
[0002] The Internet of Things (IoT) originated in the media field. It refers to connecting any object to a network through information sensing devices according to agreed protocols. Objects exchange and communicate information through information transmission media to achieve intelligent identification, positioning, tracking, and monitoring functions.
[0003] Currently, due to the presence of various dynamic factors during exercise, the data collected by the equipped sensor units is subject to environmental interference when performing real-time exercise data detection. This makes it impossible to detect data acquisition deviations during exercise in real time. When sensor data conflicts and noise interference occur, the error in exercise state analysis will be large, and the accuracy of exercise information services cannot be guaranteed.
[0004] Therefore, a motion information service and detection system based on the Internet of Things is proposed to solve the above problems. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a motion information service and detection system based on the Internet of Things, which solves the problem mentioned in the background technology that the motion state analysis has large errors and cannot guarantee the accuracy of motion information services.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a motion information service and detection system based on the Internet of Things, comprising: The motion data acquisition module uses IoT sensor units to collect multi-source motion data, cleans and standardizes the raw data through a data preprocessing unit, and outputs the initial motion data through a data transmission unit. The motion information processing module receives the initial motion data, identifies key motion features through the feature extraction unit, constructs a motion pattern model using the machine learning analysis unit, and outputs the processed motion information through the anomaly detection unit. The personalized service module receives the processed motion information, generates customized training suggestions through the recommendation engine unit, displays real-time feedback through the user interaction unit, and outputs user behavior data through the service response unit. The real-time monitoring module receives the processed motion information, calculates the motion risk index through the risk assessment unit, executes early warning actions through the alarm triggering unit, and outputs a monitoring report through the log management unit. The system optimization module receives the user behavior data and monitoring reports, analyzes the system performance through the performance evaluation unit, adjusts the sensor parameters through the parameter adaptation unit, and outputs optimization instructions to the motion data acquisition module to iteratively update the data acquisition strategy.
[0007] Preferably, the IoT sensor unit includes an accelerometer, a gyroscope, a heart rate sensor, and a location tracker; The accelerometer collects motion acceleration data, the gyroscope measures motion attitude angular velocity, the heart rate sensor monitors the user's heart rate changes in real time, and the location tracker obtains motion location information based on GPS and Bluetooth technologies. The data preprocessing unit includes a noise filtering subunit and a data normalization subunit. The noise filtering subunit uses a wavelet transform algorithm to eliminate environmental interference, and the data normalization subunit converts the original data into a uniform dimension through linear scaling.
[0008] Preferably, the feature extraction unit includes a motion trajectory parsing subunit and a biometric feature recognition subunit; The motion trajectory analysis sub-unit reconstructs the motion path using the Kalman filter algorithm and outputs velocity, displacement, and acceleration feature vectors. The biometric identification subunit extracts heart rate variability, fatigue index, and metabolic rate features from heart rate sensor data using a deep learning model. The machine learning analysis unit uses a fusion model of convolutional neural network (CNN) and long short-term memory network (LSTM) to construct a motion pattern model from the input feature vector, and outputs motion efficiency score, abnormal behavior probability, and predictive injury risk.
[0009] Preferably, the recommendation engine unit includes a personalized training plan generation subunit and a real-time feedback adaptation subunit; The personalized training plan generation subunit generates customized training suggestions, including training intensity, frequency, and recovery strategies, based on user historical data, motion target input, and environmental factors, using a decision tree algorithm. The real-time feedback adaptation subunit utilizes augmented reality (AR) technology to display visual guidance on the mobile terminal through the user interaction unit, and receives user voice and gesture input to dynamically adjust suggestions; The service response unit integrates an API interface to synchronize user behavior data to a cloud database.
[0010] Preferably, the risk assessment unit includes an exercise load calculation subunit and a risk prediction subunit; The exercise load calculation subunit analyzes heart rate and acceleration data through a physiological model and outputs an instantaneous load index; The risk prediction subunit uses a Bayesian network algorithm, combined with ambient temperature, humidity and user health records, to calculate the sports risk index and classify it into low, medium and high risk levels; The alarm triggering unit includes an audio alarm subunit and a vibration feedback subunit. The audio alarm subunit broadcasts a voice warning when the risk index exceeds the threshold, and the vibration feedback subunit applies tactile feedback through a wearable device.
[0011] Preferably, the log management unit includes a data compression subunit and an encrypted storage subunit; The data compression subunit uses the LZ77 algorithm to compress monitoring report data, reducing storage space usage; The encrypted storage sub-unit securely stores data locally and in the cloud using the AES-256 encryption protocol. The performance evaluation unit includes an efficiency index calculation subunit and a fault diagnosis subunit. The efficiency index calculation subunit calculates the system performance score based on data acquisition frequency, processing latency, and service response time. The fault diagnosis subunit identifies sensor faults and network interruptions through abnormal log patterns.
[0012] Preferably, the parameter adaptive unit includes a sensor sensitivity adjustment subunit and a sampling rate optimization subunit; The sensor sensitivity control subunit dynamically adjusts the sensor gain based on ambient light and motion intensity, while the sampling rate optimization subunit automatically sets the data sampling frequency according to the user's motion type through reinforcement learning algorithms. The system optimization module also includes an energy management unit, which uses dynamic voltage regulation technology to reduce sensor power consumption and feeds back optimization instructions to the IoT sensor unit in real time.
[0013] Preferably, the motion data acquisition module further includes a multi-source data fusion subunit and a spatiotemporal alignment subunit; The multi-source data fusion subunit is integrated within the data preprocessing unit. It receives acceleration data, heart rate data, and location data from IoT sensor units and uses the DS evidence theory algorithm to perform data fusion in order to eliminate sensor conflicts and improve data consistency. The spatiotemporal alignment subunit is connected to the data transmission unit. Through timestamp correction and spatial coordinate mapping algorithms, it synchronizes the fused data with the GPS information output by the position tracker to generate high-precision initial motion data. The data fusion processing includes a dynamic weight allocation mechanism that automatically adjusts the fusion weights based on sensor confidence levels to ensure that the output data is suitable for the feature extraction unit of the motion information processing module.
[0014] Preferably, the user interaction unit includes a voice recognition subunit and a gesture control subunit; The speech recognition subunit integrates a natural language processing (NLP) engine to parse user voice commands, while the gesture control subunit captures user gestures through a camera and inertial sensors and converts them into control signals. The personalized service module also integrates a social sharing unit, which connects to social media platforms via API, allowing users to share training data and feedback.
[0015] Preferably, the system optimization module further includes a cloud integration subunit and a distributed processing subunit; The cloud integration subunit is embedded in the log management unit and uploads the monitoring report and user behavior data to the cloud server through a data synchronization protocol. The distributed processing subunit is integrated into the performance evaluation unit and uses the Apache Kafka framework to implement real-time data stream processing, including data warehouse storage and microservice architecture calls. The cloud server provides external application access through the API gateway unit, and supports the recommendation engine unit of the personalized service module to dynamically call historical cloud data for training and suggestion optimization. The distributed processing subunit also includes a load balancing mechanism that automatically allocates computing resources based on the amount of data to improve the efficiency of the parameter adaptive unit of the system optimization module.
[0016] Beneficial effects Compared with existing technologies, the present invention provides a motion information service and detection system based on the Internet of Things, which has the following beneficial effects: 1. In this invention, when performing real-time acquisition and processing of motion data, the multi-source data acquisition function is integrated through the Internet of Things sensor unit, and data preprocessing standards are set for different types of motion to ensure the comprehensiveness and consistency of data acquisition in various motion scenarios. At the same time, the sensor data is processed using data fusion and spatiotemporal alignment technology, which can detect and eliminate conflicts and noise interference in the data acquisition process in real time, ensuring the accuracy of motion information analysis and further reducing the error of motion state recognition.
[0017] 2. In this invention, when performing motion anomaly detection and risk assessment, the feature extraction unit and machine learning analysis unit calculate the motion posture deviation value and risk index in real time, dynamically determine whether the user's motion state is abnormal, so that the system can promptly detect motion trajectory deviation and potential injury risks, and when an anomaly is detected, it can provide immediate feedback and correction suggestions through the alarm triggering unit of the real-time monitoring module to ensure safety and standardized movements during the exercise process.
[0018] 3. In this invention, when providing personalized sports services, the recommendation engine unit and user interaction unit adaptively generate multi-dimensional training suggestions, and dynamically adjust the service content based on the user's real-time sports data and historical behavior. This enables the system to achieve personalized and high-precision sports guidance, reduce the mismatch between sports suggestions and the user's actual needs, and further improve the reliability of sports services and user satisfaction. Attached Figure Description
[0019] Figure 1 This is a schematic diagram of the overall system architecture of the present invention. Detailed Implementation
[0020] 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.
[0021] Please see Figure 1 The IoT-based motion information service and detection system includes: The motion data acquisition module uses IoT sensor units to collect multi-source motion data, cleans and standardizes the raw data through a data preprocessing unit, and outputs the initial motion data through a data transmission unit. The motion information processing module receives initial motion data, identifies key motion features through the feature extraction unit, constructs a motion pattern model using the machine learning analysis unit, and outputs the processed motion information through the anomaly detection unit. The personalized service module receives the processed exercise information, generates customized training suggestions through the recommendation engine unit, displays real-time feedback through the user interaction unit, and outputs user behavior data through the service response unit. The real-time monitoring module receives processed motion information, calculates the motion risk index through the risk assessment unit, executes early warning actions through the alarm triggering unit, and outputs a monitoring report through the log management unit. The system optimization module receives user behavior data and monitoring reports, analyzes system performance through the performance evaluation unit, adjusts sensor parameters through the parameter adaptation unit, and outputs optimization instructions to the motion data acquisition module to iteratively update the data acquisition strategy. IoT sensor units include accelerometers, gyroscopes, heart rate sensors, and location trackers; Among them, the accelerometer collects motion acceleration data, the gyroscope measures the angular velocity of motion posture, the heart rate sensor monitors the user's heart rate changes in real time, and the location tracker obtains motion location information based on GPS and Bluetooth technologies. The data preprocessing unit includes a noise filtering subunit and a data normalization subunit. The noise filtering subunit uses wavelet transform algorithm to eliminate environmental interference, and the data normalization subunit converts the original data into a uniform dimension through linear scaling. Wavelet transform algorithm: a signal processing technique that eliminates noise by decomposing a signal into different frequency subbands, suitable for non-stationary signals; Data normalization formula:
[0022] in, For raw sensor data, , These are the minimum and maximum values of the original data. , To normalize the range and ensure that the output data has a consistent dimension; This formula is applied to the data normalization sub-unit to linearly scale accelerometer and gyroscope data to a uniform range, eliminating dimensional differences. The feature extraction unit includes a motion trajectory parsing subunit and a biometric feature recognition subunit; The motion trajectory analysis sub-unit reconstructs the motion path using the Kalman filter algorithm and outputs velocity, displacement, and acceleration feature vectors. Kalman filter algorithm: a recursive filtering technique that estimates motion trajectories through a prediction-update step, suitable for noisy environments; Kalman filter state equation:
[0023]
[0024] in, For state vectors, Here is the state transition matrix. To control the input, Let be the error covariance matrix. For process noise covariance; This formula is applied in the motion trajectory analysis sub-unit. The original displacement data is input, and the feature vector is output, with the dimensions normalized to the velocity unit. The biometric identification subunit extracts heart rate variability, fatigue index, and metabolic rate features from heart rate sensor data using a deep learning model. The machine learning analysis unit uses a fusion model of convolutional neural network (CNN) and long short-term memory network (LSTM). It inputs feature vectors to construct a motion pattern model and outputs motion efficiency scores, abnormal behavior probabilities, and predictive injury risks. A fusion model of Convolutional Neural Network (CNN) and Long Short-Term Memory Network (LSTM): CNN processes spatial features, LSTM processes time series, and the combination improves the accuracy of motion pattern recognition. LSTM output formula:
[0025] in, In hidden state, In cellular state, For output gate, forget gate, and input gate; Input heart rate variability data, output fatigue index, and construct an exercise pattern model; The recommendation engine unit includes a personalized training plan generation subunit and a real-time feedback adaptation subunit; The personalized training plan generation subunit generates customized training suggestions, including training intensity, frequency, and recovery strategies, based on user historical data, motion target input, and environmental factors, using a decision tree algorithm. Decision tree algorithm: a machine learning method that uses a tree structure to segment data and generate training suggestions; Decision tree splitting formula:
[0026] in, For the dataset, For the proportion of classes, Number of categories; In the personalized training plan generation sub-unit, this formula takes the user's historical data as input and outputs training intensity suggestions, with the units normalized to power units. The real-time feedback adaptation subunit utilizes augmented reality (AR) technology to display visual guidance on the mobile terminal through the user interaction unit, and receives user voice and gesture input to dynamically adjust suggestions; The service response unit integrates API interfaces to synchronize user behavior data to the cloud database; The risk assessment unit includes an exercise load calculation subunit and a risk prediction subunit; The exercise load calculation subunit analyzes heart rate and acceleration data through a physiological model and outputs an instantaneous load index; Physiological model: Mathematical equations simulate human response, the relationship between heart rate and load; Exercise load index formula:
[0027] in, This is the instantaneous load index. Heart rate, For acceleration, These are weighting coefficients, determined through calibration; Input heart rate and acceleration data, output load index, and normalize the dimensions to power units; The risk prediction subunit uses a Bayesian network algorithm, combined with ambient temperature, humidity and user health records, to calculate the sports risk index and classify it into low, medium and high risk levels; Bayesian network algorithm: a probabilistic graphical model that assesses risk based on conditional probability; Bayesian risk probability formula:
[0028] in, For risk conditional probability, Environmental factors; Input ambient temperature and user health data, output risk level; The alarm triggering unit includes an audio alarm subunit and a vibration feedback subunit. The audio alarm subunit broadcasts a voice warning when the risk index exceeds the threshold, and the vibration feedback subunit applies tactile feedback through a wearable device. The log management unit includes a data compression subunit and an encrypted storage subunit; The data compression subunit uses the LZ77 algorithm to compress monitoring report data, reducing storage space usage; The encrypted storage sub-unit securely stores data locally and in the cloud using the AES-256 encryption protocol. The performance evaluation unit includes an efficiency index calculation subunit and a fault diagnosis subunit. The efficiency index calculation subunit calculates the system performance score based on data acquisition frequency, processing latency, and service response time. The fault diagnosis subunit identifies sensor faults and network interruptions through abnormal log patterns. The parameter adaptive unit includes a sensor sensitivity control subunit and a sampling rate optimization subunit; The sensor sensitivity control subunit dynamically adjusts the sensor gain based on ambient light and motion intensity, while the sampling rate optimization subunit automatically sets the data sampling frequency according to the user's motion type through reinforcement learning algorithms. Reinforcement learning algorithms: an AI method that optimizes decision-making through reward mechanisms; Reinforcement learning reward function:
[0029] in, As a reward value, For state, For action, Due to data error, For sensor power consumption; Input motion type, output optimized sampling rate; The system optimization module also includes an energy management unit, which uses dynamic voltage regulation technology to reduce sensor power consumption and feeds back optimization instructions to the IoT sensor unit in real time. The motion data acquisition module also includes a multi-source data fusion subunit and a spatiotemporal alignment subunit; The multi-source data fusion subunit is integrated into the data preprocessing unit. It receives acceleration data, heart rate data, and location data from IoT sensor units and uses the DS evidence theory algorithm to perform data fusion in order to eliminate sensor conflicts and improve data consistency. DS Evidence Theory Algorithm: An uncertainty reasoning method that fuses multi-source sensor data and reduces conflicts through evidence combination rules; Multi-source data fusion formula:
[0030] in, The confidence level of the merged events, , For different sensors, the evidence quality function is used. Conflict factors are dynamically weighted. This formula is implemented in the multi-source data fusion subunit, which takes acceleration, heart rate and position data as input and outputs high-precision initial motion data. The spatiotemporal alignment subunit is connected to the data transmission unit. Through timestamp correction and spatial coordinate mapping algorithms, it synchronizes the fused data with the GPS information output by the position tracker to generate high-precision initial motion data. The data fusion processing includes a dynamic weight allocation mechanism that automatically adjusts the fusion weights based on sensor confidence levels to ensure that the output data is suitable for the feature extraction unit of the motion information processing module. The user interaction unit includes a voice recognition subunit and a gesture control subunit; The speech recognition subunit integrates a natural language processing (NLP) engine to parse user voice commands, while the gesture control subunit captures user gestures through a camera and inertial sensors and converts them into control signals. Natural Language Processing (NLP) Engine: An AI model for parsing user voice commands, based on the Transformer architecture; NLP engine parsing formulas:
[0031] in, For word sequence probability, The current word; Input voice commands, output control signals, gesture control subunit; The personalized service module also integrates a social sharing unit, which connects to social media platforms via API, allowing users to share training data and feedback; The system optimization module also includes a cloud integration subunit and a distributed processing subunit; The cloud integration subunit is embedded within the log management unit, and uploads monitoring reports and user behavior data to the cloud server through a data synchronization protocol. The distributed processing subunit is integrated into the performance evaluation unit and uses the Apache Kafka framework to implement real-time data stream processing, including data warehouse storage and microservice architecture calls. Apache Kafka framework: a distributed stream processing platform for real-time data management; Among them, the cloud server provides access to external applications through the API gateway unit, and supports the recommendation engine unit of the personalized service module to dynamically call historical data in the cloud for training and suggestion optimization; The distributed processing subunit also includes a load balancing mechanism that automatically allocates computing resources based on the amount of data to improve the efficiency of parameter adaptive adjustment in the system optimization module.
[0032] Load balancing formula:
[0033] in, For data processing throughput, For data volume, For processing time, The number of nodes; In the distributed processing subunit, monitoring report data is input, and optimized resource allocation is output. The operation steps of the IoT-based motion information service and detection system are as follows: Step 1: Multi-source motion data acquisition In this stage, the system initiates the data acquisition process, acquiring user motion data in real time through IoT sensor units, including accelerometers, gyroscopes, heart rate sensors, and location trackers. The data acquisition process follows the principle of multi-source data acquisition to ensure coverage of speed, location, and physiological information.
[0034] Detailed Explanation of the Principle: The sensor unit is automatically activated according to the preset motion type, and the data preprocessing unit immediately intervenes. Among them, the noise filtering subunit uses wavelet transform algorithm to clean the raw data, and the data normalization subunit standardizes the data through linear scaling formula. At the same time, the multi-source data fusion subunit applies DS evidence theory algorithm to dynamically allocate weights and generate high-precision initial motion data. This step ensures the comprehensiveness and consistency of data acquisition and lays the foundation for subsequent processing.
[0035] Step 2: Data Preprocessing and Spatiotemporal Alignment The initial data collected needs further processing to eliminate spatiotemporal bias. The system integrates the data through a data preprocessing unit and a spatiotemporal alignment subunit to ensure the accuracy of subsequent analysis.
[0036] Detailed Explanation of the Principle: The multi-source data fusion subunit receives acceleration, heart rate, and location data, and fuses them using the DS evidence theory formula, dynamically adjusting the weights. Subsequently, the spatiotemporal alignment subunit synchronizes the fused data with GPS information through timestamp correction and spatial coordinate mapping algorithms. Timestamp correction ensures that all sensor data are aligned at a unified time point, while spatial coordinate mapping transforms the location data to a standard coordinate system. This step eliminates sensor conflicts and noise interference, outputs standardized initial motion data, and reduces motion state recognition error to within 5%.
[0037] Step 3: Motion Feature Extraction and Pattern Recognition The system enters the analysis phase, where the motion information processing module receives preprocessed data, extracts key features, and constructs a motion pattern model. This step utilizes machine learning algorithms to achieve accurate state recognition.
[0038] Detailed Explanation of the Principle: The feature extraction unit is divided into a motion trajectory analysis subunit and a biometric recognition subunit. The motion trajectory analysis subunit applies the Kalman filter algorithm to reconstruct the motion path and output feature vectors. Example formula: Kalman filter state equation. The biometric recognition subunit extracts heart rate variability and fatigue index from heart rate data through a deep learning model. The machine learning analysis unit inputs these feature vectors to construct a motion pattern model. The LSTM network processes the time series data. This step achieves high-precision motion state analysis and reduces the error rate.
[0039] Step 4: Personalized Service Generation and User Interaction Based on the analysis results, the system generates customized services and provides real-time feedback to users through the interaction module. The personalized service module ensures that the suggestions match the user's needs.
[0040] Detailed Explanation of Principles: The recommendation engine unit receives processed motion information, the personalized training plan generation sub-unit uses a decision tree algorithm, combining user historical data, motion goals, and environmental factors. Formula example: decision tree segmentation criteria. The real-time feedback adaptation sub-unit displays visual guidance on the mobile terminal through augmented reality technology. The user interaction unit supports voice recognition and gesture control. The service response unit synchronizes user behavior data to the cloud database, ensuring a service response time of less than 500ms, further improving user satisfaction and service reliability.
[0041] Step 5: Real-time risk monitoring and alarm triggering The system continuously monitors exercise status, assesses risks, and intervenes promptly. The real-time monitoring module ensures exercise safety.
[0042] Detailed Explanation of Principles: The risk assessment unit receives the output of the motion information processing module. The motion load calculation subunit uses a physiological model to calculate the instantaneous load index. The risk prediction subunit uses a Bayesian network algorithm, combined with environmental data and user health records, to calculate the motion risk index. Example formula: Bayesian probability. When the risk exceeds the threshold, the alarm triggering unit executes an early warning. The sound alarm subunit broadcasts a voice warning. The vibration feedback subunit applies tactile feedback through wearable devices. The log management unit simultaneously compresses and encrypts the monitoring report. This step ensures motion safety, with a response latency of less than 100ms.
[0043] Step Six: System Optimization and Iterative Updates Finally, the system performs self-optimization based on user feedback and performance data, and the system optimization module achieves closed-loop iteration to improve overall efficiency.
[0044] Detailed Explanation of Principles: The performance evaluation unit receives user behavior data and monitoring reports; the efficiency index calculation subunit calculates the system performance score; the fault diagnosis subunit identifies anomalies; the parameter adaptation unit applies reinforcement learning algorithms to adjust sensor parameters; the sampling rate optimization subunit sets the data sampling frequency; and the energy management unit reduces power consumption through dynamic voltage regulation. The distributed processing subunit utilizes the Apache Kafka framework to implement data stream processing, and a load balancing mechanism automatically allocates resources. Optimization instructions are fed back to the motion data acquisition module for iterative updates, further improving system efficiency.
[0045] 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 a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.
[0046] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. A motion information service and detection system based on the Internet of Things, characterized in that: include: The motion data acquisition module uses IoT sensor units to collect multi-source motion data, cleans and standardizes the raw data through a data preprocessing unit, and outputs the initial motion data through a data transmission unit. The motion information processing module receives the initial motion data, identifies key motion features through the feature extraction unit, constructs a motion pattern model using the machine learning analysis unit, and outputs the processed motion information through the anomaly detection unit. The personalized service module receives the processed motion information, generates customized training suggestions through the recommendation engine unit, displays real-time feedback through the user interaction unit, and outputs user behavior data through the service response unit. The real-time monitoring module receives the processed motion information, calculates the motion risk index through the risk assessment unit, executes early warning actions through the alarm triggering unit, and outputs a monitoring report through the log management unit. The system optimization module receives the user behavior data and monitoring reports, analyzes the system performance through the performance evaluation unit, adjusts the sensor parameters through the parameter adaptation unit, and outputs optimization instructions to the motion data acquisition module to iteratively update the data acquisition strategy.
2. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The IoT sensor unit includes an accelerometer, a gyroscope, a heart rate sensor, and a location tracker; The accelerometer collects motion acceleration data, the gyroscope measures motion attitude angular velocity, the heart rate sensor monitors the user's heart rate changes in real time, and the location tracker obtains motion location information based on GPS and Bluetooth technologies. The data preprocessing unit includes a noise filtering subunit and a data normalization subunit. The noise filtering subunit uses a wavelet transform algorithm to eliminate environmental interference, and the data normalization subunit converts the original data into a uniform dimension through linear scaling.
3. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The feature extraction unit includes a motion trajectory parsing subunit and a biometric feature recognition subunit; The motion trajectory analysis sub-unit reconstructs the motion path using the Kalman filter algorithm and outputs velocity, displacement, and acceleration feature vectors. The biometric identification subunit extracts heart rate variability, fatigue index, and metabolic rate features from heart rate sensor data using a deep learning model. The machine learning analysis unit uses a fusion model of convolutional neural network (CNN) and long short-term memory network (LSTM) to construct a motion pattern model from the input feature vector, and outputs motion efficiency score, abnormal behavior probability, and predictive injury risk.
4. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The recommendation engine unit includes a personalized training plan generation subunit and a real-time feedback adaptation subunit. The personalized training plan generation subunit generates customized training suggestions, including training intensity, frequency, and recovery strategies, based on user historical data, motion target input, and environmental factors, using a decision tree algorithm. The real-time feedback adaptation subunit utilizes augmented reality (AR) technology to display visual guidance on the mobile terminal through the user interaction unit, and receives user voice and gesture input to dynamically adjust suggestions; The service response unit integrates an API interface to synchronize user behavior data to a cloud database.
5. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The risk assessment unit includes an exercise load calculation subunit and a risk prediction subunit; The exercise load calculation subunit analyzes heart rate and acceleration data through a physiological model and outputs an instantaneous load index; The risk prediction subunit uses a Bayesian network algorithm, combined with ambient temperature, humidity and user health records, to calculate the sports risk index and classify it into low, medium and high risk levels; The alarm triggering unit includes an audio alarm subunit and a vibration feedback subunit. The audio alarm subunit broadcasts a voice warning when the risk index exceeds the threshold, and the vibration feedback subunit applies tactile feedback through a wearable device.
6. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The log management unit includes a data compression subunit and an encrypted storage subunit; The data compression subunit uses the LZ77 algorithm to compress monitoring report data, reducing storage space usage; The encrypted storage sub-unit securely stores data locally and in the cloud using the AES-256 encryption protocol. The performance evaluation unit includes an efficiency index calculation subunit and a fault diagnosis subunit. The efficiency index calculation subunit calculates the system performance score based on data acquisition frequency, processing latency, and service response time. The fault diagnosis subunit identifies sensor faults and network interruptions through abnormal log patterns.
7. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The parameter adaptive unit includes a sensor sensitivity control subunit and a sampling rate optimization subunit; The sensor sensitivity control subunit dynamically adjusts the sensor gain based on ambient light and motion intensity, while the sampling rate optimization subunit automatically sets the data sampling frequency according to the user's motion type through reinforcement learning algorithms. The system optimization module also includes an energy management unit, which uses dynamic voltage regulation technology to reduce sensor power consumption and feeds back optimization instructions to the IoT sensor unit in real time.
8. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The motion data acquisition module also includes a multi-source data fusion subunit and a spatiotemporal alignment subunit; The multi-source data fusion subunit is integrated within the data preprocessing unit. It receives acceleration data, heart rate data, and location data from IoT sensor units and uses the DS evidence theory algorithm to perform data fusion in order to eliminate sensor conflicts and improve data consistency. The spatiotemporal alignment subunit is connected to the data transmission unit. Through timestamp correction and spatial coordinate mapping algorithms, it synchronizes the fused data with the GPS information output by the position tracker to generate high-precision initial motion data. The data fusion processing includes a dynamic weight allocation mechanism that automatically adjusts the fusion weights based on sensor confidence levels to ensure that the output data is suitable for the feature extraction unit of the motion information processing module.
9. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The user interaction unit includes a voice recognition subunit and a gesture control subunit; The speech recognition subunit integrates a natural language processing (NLP) engine to parse user voice commands, while the gesture control subunit captures user gestures through a camera and inertial sensors and converts them into control signals. The personalized service module also integrates a social sharing unit, which connects to social media platforms via API, allowing users to share training data and feedback.
10. The motion information service and detection system based on the Internet of Things according to claim 1, characterized in that: The system optimization module also includes a cloud integration subunit and a distributed processing subunit; The cloud integration subunit is embedded in the log management unit and uploads the monitoring report and user behavior data to the cloud server through a data synchronization protocol. The distributed processing subunit is integrated into the performance evaluation unit and uses the Apache Kafka framework to implement real-time data stream processing, including data warehouse storage and microservice architecture calls. The cloud server provides external application access through the API gateway unit, and supports the recommendation engine unit of the personalized service module to dynamically call historical cloud data for training and suggestion optimization. The distributed processing subunit also includes a load balancing mechanism that automatically allocates computing resources based on the amount of data to improve the efficiency of the parameter adaptive unit of the system optimization module.