Physical education quality dynamic evaluation system and method based on artificial intelligence
By collecting and analyzing student movement and teacher behavior data in real time, a dynamic evaluation model is constructed, which solves the consistency and personalization problems in traditional evaluation methods and realizes real-time, personalized feedback and optimization of physical education teaching.
Patent Information
- Application Number
- CN202511394394.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-28
- Publication Date
- 2026-01-06
AI Technical Summary
Existing physical education assessment methods rely on manual scoring, which lacks consistency and fairness, makes it difficult to provide real-time feedback on students' shortcomings, fails to provide personalized guidance, and the traditional assessment system is time-consuming and difficult to adapt to the actual classroom situation.
Using inertial motion capture sensors, human posture data fitting, non-embedded physiological signal measurement, video recording, and speech recognition technologies, the system collects student movement and teacher behavior data in real time. It then combines machine learning and deep learning models to analyze the data, build a dynamic evaluation model, and provide personalized feedback and learning paths.
It enables real-time, objective, and personalized assessment of physical education teaching, improves the accuracy of assessment results and the practicality of teaching, and promotes the two-way optimization of individualized instruction and teaching.
Smart Images
Figure CN121279862A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of artificial intelligence technology, and in particular to a dynamic evaluation system and method for physical education teaching quality based on artificial intelligence. Background Technology
[0002] With the rapid development of information technology, artificial intelligence (AI) technology has been widely applied in many fields and is gradually penetrating the education industry. In the specific field of physical education, the application of AI technology provides new possibilities for improving teaching quality and efficiency, such as providing personalized training suggestions through intelligent analysis of student performance data. However, current physical education assessment methods mostly remain at a relatively traditional level, relying mainly on manual scoring and subjective judgment. Traditional assessment methods are usually based on the subjective evaluation of teachers or judges, and different people may have different opinions on the same performance. This method, which relies on personal experience, makes it difficult to ensure the consistency and fairness of assessment results. Secondly, existing assessment methods often focus on the final score or the performance of certain specific skills, while ignoring the consideration of students' overall athletic ability and progress process, as well as the in-depth analysis of the effectiveness of teachers' teaching strategies. Furthermore, because manual assessment is time-consuming, it is difficult to provide immediate feedback, preventing students from understanding their shortcomings and making improvements in a timely manner. At the same time, teachers also find it difficult to quickly adjust their teaching plans to adapt to the actual classroom situation. Finally, the guidance provided under the current system is mostly general advice, failing to fully consider the individual differences and needs of each student, thus limiting the effectiveness of individualized instruction. Therefore, in order to overcome these challenges, it is urgent to develop a new type of dynamic evaluation system and method for physical education teaching quality based on artificial intelligence. This system can realize real-time and automated monitoring and evaluation of students' exercise status and teachers' teaching behavior, thereby providing more scientific and accurate teaching feedback and support, and promoting the development of physical education towards intelligence and personalization. Summary of the Invention
[0003] A method for dynamic evaluation of physical education teaching quality based on artificial intelligence includes the following steps; S1. Data Acquisition: Utilizing inertial motion capture sensor technology, human posture data fitting technology, and non-embedded physiological signal measurement technology, real-time data is collected on students' body data, movement posture data, and physiological signal data during physical activities. Through video recording and speech recognition technology, the teaching behavior and explanation content of teachers are collected. S2. Data Analysis: Clean, denoise, and convert the collected data. Use machine learning algorithms and deep learning models to extract features and recognize patterns in the processed data to analyze students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. S3. Evaluation Model Construction: Based on the data analysis results, a dynamic evaluation model for physical education teaching quality is constructed to evaluate the physical education teaching process; S4. Feedback: Based on the evaluation model results, provide teachers with suggestions for optimizing instructional design, improving classroom management, and adjusting teaching evaluation, while also providing students with personalized learning paths and resources.
[0004] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S1, inertial motion capture sensor technology, human posture data fitting technology, and non-embedded physiological signal measurement technology are used to collect students' body data, movement posture data, and physiological signal data in real time during physical activities. Through video recording and speech recognition technology, the teaching behavior and explanation content of teachers are collected. The specific steps are as follows: S11. Sensors collect body data: Students are equipped with inertial motion capture devices. The devices have built-in accelerometers and gyroscopes, covering the arms, legs and torso. The devices record acceleration and angular velocity data in real time and transmit them wirelessly to the data processing unit. The unit receives the data and performs preprocessing, including calibration and filtering. S12. Human posture data fitting: The human posture data fitting algorithm is used to iteratively fit the parameterized human body model with the real-time data collected by the inertial sensor to generate human posture estimation. S13. Non-embedded physiological signal measurement: Equip students with non-embedded physiological signal measurement devices to monitor physiological indicators including heart rate and blood pressure in real time. The data is transmitted wirelessly to the data processing unit for preprocessing and analysis to extract key physiological information. S14. Video and speech recognition: Install cameras and microphones in sports venues to capture teachers' teaching behaviors and explanations. The data is transmitted to the data processing unit via wired or wireless means for preprocessing and organization, and the teaching information is extracted and stored.
[0005] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S2, the collected data is cleaned, denoised, and format converted. Machine learning algorithms and deep learning models are used to extract features and recognize patterns in the processed data to analyze students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. The specific steps are as follows: S21. Data preprocessing and analysis preparation: Clean, denoise, and convert the format of the collected data to remove outliers and duplicate data, reduce sensor noise and environmental interference, and ensure that the data conforms to the unified format required for subsequent analysis. S22. Feature Extraction and Pattern Recognition: Using machine learning algorithms and deep learning models, feature extraction is performed on the preprocessed data, including extracting students' movement speed, strength, and coordination indicators, teachers' teaching style and interaction characteristics, and the fractal dimension of the movement sequence; through pattern recognition technology, the movement patterns of students in sports activities and the teaching behavior patterns of teachers are identified. S23. Analysis of athletic performance and skill mastery: Based on the extracted features and identified patterns, analyze students' athletic performance and skill mastery, assess students' athletic skill level, movement standardization, and progress rate, identify problems and areas for improvement during the athletic process, and analyze the impact of different sports and training plans on students' athletic performance. S24. Teaching effectiveness evaluation: Analyze teachers' teaching style, interaction methods, and clarity of explanation; identify the strengths and weaknesses in the teaching process; provide teaching feedback and suggestions; and store the analysis results in a database.
[0006] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S3, based on the data analysis results, a dynamic evaluation model for the quality of physical education teaching is constructed to evaluate the physical education teaching process. The specific steps are as follows: S31. Evaluation Indicator Determination and Weight Allocation: Determine the evaluation indicators for students' athletic performance, skill mastery, psychological state, and teachers' instructional design and classroom interaction, and allocate the weights for each indicator; S32. Dynamic Evaluation Model Design and Construction: A dynamic evaluation model is constructed using machine learning and deep learning technologies. The model integrates a comprehensive evaluation function for exercise efficiency: η=(α×A+β×S+θ×P) / FD, where η is the comprehensive evaluation index for exercise efficiency; A is the accuracy of movement, with a value range of [0,100]; S is the stability of movement, with a value range of [0,100]; P is the physiological load index, with a value range of [0,1]; α, β, and θ are weight coefficients, with values ranging from [0.2,0.5], and satisfying α+β+θ=1. Their optimal values are determined through model training; FD is the fractal dimension of the movement sequence, with a value range of [1,2]. The optimal range for FD values is between 1.2 and 1.5. S33. Model Validation and Optimization: Apply the model to actual teaching scenarios for validation. By comparing the model evaluation results with the actual situation, assess the accuracy and reliability of the model. Adjust and optimize the model based on the validation results.
[0007] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S4, based on the evaluation model results, teachers are provided with optimized instructional design, improved classroom management, and adjusted teaching evaluation. At the same time, students are provided with personalized learning paths and learning resources. The specific steps are as follows: S41. Personalized Teaching Suggestion Generation: Based on the output of the dynamic assessment model, it provides teachers with personalized teaching suggestions covering instructional design optimization, classroom management improvement, and teaching evaluation adjustment; S42. Classroom Management Optimization Plan Development: Based on the analysis of student classroom performance using the evaluation model, develop classroom management optimization plans for teachers, including suggestions on maintaining classroom discipline, improving student participation, and creating a positive classroom atmosphere. S43. Personalized Learning Path Planning for Students: Based on the analysis results of the assessment model on students' physical fitness, skill level, interests, etc., personalized learning paths are planned for students, including recommending suitable sports programs, training plans, and learning resources. This provides students with learning solutions that meet their characteristics and needs, and helps to effectively improve their sports skills and physical fitness. S44. Learning Resource Recommendation and Access Support: Based on students' learning paths and individual needs, relevant teaching videos, training guides, and online courses are recommended, and resource links, download methods, and usage instructions are provided.
[0008] An AI-based dynamic evaluation system for physical education teaching quality is provided, wherein the AI-based dynamic evaluation system for physical education teaching quality is used to implement any of the methods for dynamic evaluation of physical education teaching quality based on AI; the AI-based dynamic evaluation system for physical education teaching quality includes: a data acquisition module, a data analysis module, an evaluation model construction module, and a feedback module; The data acquisition module uses sensor technology to collect students' body, posture, and physiological data in real time during physical activities. At the same time, it captures teachers' teaching behaviors and explanations through video recording and speech recognition technology. Data analysis module: Cleans, denoises, and converts the format of the collected data; uses machine learning and deep learning models to extract features and identify patterns; and analyzes students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. Assessment Model Construction Module: Based on data analysis, a dynamic assessment model is constructed to comprehensively and objectively assess physical education teaching by taking into account students' physical fitness, skills, interests, and teachers' teaching strategies and styles. Feedback module: Based on the evaluation results, it provides teachers with personalized teaching suggestions, including design optimization, management improvement, and evaluation adjustment, and customizes learning paths and resources for students.
[0009] The beneficial effects of this invention are: 1. Real-time dynamic evaluation of physical education teaching process: Through multi-source data acquisition technology, including inertial motion capture, non-embedded physiological signal measurement and audio-visual recognition, comprehensive and real-time monitoring of teachers' and students' behavior and status can be achieved, providing a high-precision and timely data foundation for teaching quality evaluation; 2. Improved scientific rigor and adaptability of the evaluation model: By introducing a comprehensive evaluation function for exercise efficiency, combining multiple dimensions such as movement accuracy, exercise stability, physiological load index, and fractal dimension of exercise sequence, and optimizing the weight coefficients through machine learning, the evaluation model is made to have objectivity, interpretability, and personalized adjustment capabilities, significantly improving the accuracy and practicality of the evaluation results. 3. Two-way optimization of teaching and learning: Based on dynamic assessment results, teachers are provided with specific improvement suggestions in areas such as instructional design, classroom management, and teaching evaluation. At the same time, personalized learning paths and resource recommendations are generated for students that are tailored to their physical fitness, skill level, and interests, effectively promoting individualized instruction and mutual learning between teachers and students. Attached Figure Description
[0010] Figure 1 This is a flowchart illustrating a method for dynamic evaluation of physical education teaching quality based on artificial intelligence. Detailed Implementation
[0011] A method for dynamic evaluation of physical education teaching quality based on artificial intelligence includes the following steps; S1. Data Acquisition: Utilizing inertial motion capture sensor technology, human posture data fitting technology, and non-embedded physiological signal measurement technology, real-time data is collected on students' body data, movement posture data, and physiological signal data during physical activities. Through video recording and speech recognition technology, the teaching behavior and explanation content of teachers are collected. S2. Data Analysis: Clean, denoise, and convert the collected data. Use machine learning algorithms and deep learning models to extract features and recognize patterns in the processed data to analyze students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. S3. Evaluation Model Construction: Based on the data analysis results, a dynamic evaluation model for physical education teaching quality is constructed to evaluate the physical education teaching process; S4. Feedback: Based on the evaluation model results, provide teachers with suggestions for optimizing instructional design, improving classroom management, and adjusting teaching evaluation, while also providing students with personalized learning paths and resources.
[0012] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S1, inertial motion capture sensor technology, human posture data fitting technology, and non-embedded physiological signal measurement technology are used to collect students' body data, movement posture data, and physiological signal data in real time during physical activities. Through video recording and speech recognition technology, the teaching behavior and explanation content of teachers are collected. The specific steps are as follows: S11. Sensors collect body data: Students are equipped with inertial motion capture devices. The devices have built-in accelerometers and gyroscopes, covering the arms, legs and torso. The devices record acceleration and angular velocity data in real time and transmit them wirelessly to the data processing unit. The unit receives the data and performs preprocessing, including calibration and filtering. S12. Human posture data fitting: The human posture data fitting algorithm is used to iteratively fit the parameterized human body model with the real-time data collected by the inertial sensor to generate human posture estimation. S13. Non-embedded physiological signal measurement: Equip students with non-embedded physiological signal measurement devices to monitor physiological indicators including heart rate and blood pressure in real time. The data is transmitted wirelessly to the data processing unit for preprocessing and analysis to extract key physiological information. S14. Video and speech recognition: Install cameras and microphones in sports venues to capture teachers' teaching behaviors and explanations. The data is transmitted to the data processing unit via wired or wireless means for preprocessing and organization, and the teaching information is extracted and stored.
[0013] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S2, the collected data is cleaned, denoised, and format converted. Machine learning algorithms and deep learning models are used to extract features and recognize patterns in the processed data to analyze students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. The specific steps are as follows: S21. Data preprocessing and analysis preparation: Clean, denoise, and convert the format of the collected data to remove outliers and duplicate data, reduce sensor noise and environmental interference, and ensure that the data conforms to the unified format required for subsequent analysis. S22. Feature Extraction and Pattern Recognition: Using machine learning algorithms and deep learning models, feature extraction is performed on the preprocessed data, including extracting students' movement speed, strength, and coordination indicators, teachers' teaching style and interaction characteristics, and the fractal dimension of the movement sequence; through pattern recognition technology, the movement patterns of students in sports activities and the teaching behavior patterns of teachers are identified. S23. Analysis of athletic performance and skill mastery: Based on the extracted features and identified patterns, analyze students' athletic performance and skill mastery, assess students' athletic skill level, movement standardization, and progress rate, identify problems and areas for improvement during the athletic process, and analyze the impact of different sports and training plans on students' athletic performance. S24. Teaching effectiveness evaluation: Analyze teachers' teaching style, interaction methods, and clarity of explanation; identify the strengths and weaknesses in the teaching process; provide teaching feedback and suggestions; and store the analysis results in a database.
[0014] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S3, based on the data analysis results, a dynamic evaluation model for the quality of physical education teaching is constructed to evaluate the physical education teaching process. The specific steps are as follows: S31. Evaluation Indicator Determination and Weight Allocation: Determine the evaluation indicators for students' athletic performance, skill mastery, psychological state, and teachers' instructional design and classroom interaction, and allocate the weights for each indicator; S32. Dynamic Evaluation Model Design and Construction: A dynamic evaluation model is constructed using machine learning and deep learning technologies. The model integrates a comprehensive evaluation function for exercise efficiency: η=(α×A+β×S+θ×P) / FD, where η is the comprehensive evaluation index for exercise efficiency; A is the accuracy of movement, with a value range of [0,100]; S is the stability of movement, with a value range of [0,100]; P is the physiological load index, with a value range of [0,1]; α, β, and θ are weight coefficients, with values ranging from [0.2,0.5], and satisfying α+β+θ=1. Their optimal values are determined through model training; FD is the fractal dimension of the movement sequence, with a value range of [1,2]. The optimal range for FD values is between 1.2 and 1.5. S33. Model Validation and Optimization: Apply the model to actual teaching scenarios for validation. By comparing the model evaluation results with the actual situation, assess the accuracy and reliability of the model. Adjust and optimize the model based on the validation results.
[0015] Furthermore, an artificial intelligence-based method for dynamic evaluation of physical education teaching quality is proposed. In step S4, based on the evaluation model results, teachers are provided with optimized instructional design, improved classroom management, and adjusted teaching evaluation. At the same time, students are provided with personalized learning paths and learning resources. The specific steps are as follows: S41. Personalized Teaching Suggestion Generation: Based on the output of the dynamic assessment model, it provides teachers with personalized teaching suggestions covering instructional design optimization, classroom management improvement, and teaching evaluation adjustment; S42. Classroom Management Optimization Plan Development: Based on the analysis of student classroom performance using the evaluation model, develop classroom management optimization plans for teachers, including suggestions on maintaining classroom discipline, improving student participation, and creating a positive classroom atmosphere. S43. Personalized Learning Path Planning for Students: Based on the analysis results of the assessment model on students' physical fitness, skill level, interests, etc., personalized learning paths are planned for students, including recommending suitable sports programs, training plans, and learning resources. This provides students with learning solutions that meet their characteristics and needs, and helps to effectively improve their sports skills and physical fitness. S44. Learning Resource Recommendation and Access Support: Based on students' learning paths and individual needs, relevant teaching videos, training guides, and online courses are recommended, and resource links, download methods, and usage instructions are provided.
[0016] An AI-based dynamic evaluation system for physical education teaching quality is provided, wherein the AI-based dynamic evaluation system for physical education teaching quality is used to implement any of the methods for dynamic evaluation of physical education teaching quality based on AI; the AI-based dynamic evaluation system for physical education teaching quality includes: a data acquisition module, a data analysis module, an evaluation model construction module, and a feedback module; The data acquisition module uses sensor technology to collect students' body, posture, and physiological data in real time during physical activities. At the same time, it captures teachers' teaching behaviors and explanations through video recording and speech recognition technology. Data analysis module: Cleans, denoises, and converts the format of the collected data; uses machine learning and deep learning models to extract features and identify patterns; and analyzes students' athletic performance, skill mastery, and the effectiveness of teachers' teaching. Assessment Model Construction Module: Based on data analysis, a dynamic assessment model is constructed to comprehensively and objectively assess physical education teaching by taking into account students' physical fitness, skills, interests, and teachers' teaching strategies and styles. Feedback module: Based on the evaluation results, it provides teachers with personalized teaching suggestions, including design optimization, management improvement, and evaluation adjustment, and customizes learning paths and resources for students.
Claims
1. An artificial intelligence-based method for dynamically evaluating the quality of sports teaching, characterized in that, It comprises the following steps: S1. Data collection: using inertial motion capture sensor technology, human posture data fitting technology, non-embedded physiological signal measurement technology, real-time collection of students' body data, motion posture data and physiological signal data in sports activities, through video recording and speech recognition technology, collecting teachers' teaching behavior and explanation content; S2. Data analysis: data cleaning, denoising and format conversion are performed on the collected data, machine learning algorithms and deep learning models are used to extract features and identify patterns from the processed data, analyze students' sports performance, skill mastery and teachers' teaching effectiveness; S3. Evaluation model construction: based on the data analysis results, a dynamic evaluation model of physical education teaching quality is constructed to evaluate the physical education teaching process; S4. Feedback: based on the evaluation model results, teachers are provided with teaching design optimization, classroom management improvement, teaching evaluation adjustment, while students are provided with personalized learning paths and learning resources.
2. The method of claim 1, wherein the inertial motion capture sensor technology, human posture data fitting technology, and non-embedded physiological signal measurement technology are used to collect real-time body data, motion posture data, and physiological signal data of students during sports activities, and the video recording and speech recognition technology is used to collect the teaching behavior and explanation content of teachers, and the specific steps are as follows: S11. Sensor collection of body data: inertial motion capture equipment is fitted for students to record real-time acceleration and angular velocity data during exercise, which is transmitted to the data processing unit for calibration and filtering preprocessing; S12. Human posture data fitting: human posture data fitting algorithm is used to iteratively fit the parameterized human model with sensor data to generate human posture estimation; S13. Non-embedded physiological signal measurement: non-embedded physiological signal measurement equipment is used to monitor heart rate, blood pressure physiological indicators, which are transmitted to the data processing unit for preprocessing and analysis; S14. Video and speech recognition: teacher's teaching behavior and explanation content are captured by camera and microphone, which are transmitted to the data processing unit for preprocessing and arrangement.
3. The method of claim 1, wherein the collected data is cleaned, denoised and format converted, and machine learning algorithms and deep learning models are used to extract features and identify patterns from the processed data to analyze students' sports performance, skill mastery and teachers' teaching effectiveness, and the specific steps are as follows: S21. Data preprocessing and analysis preparation: clean, denoise and format convert the collected data to remove outliers and duplicate data; S22. Feature extraction and pattern recognition: using machine learning algorithms and deep learning models, the pre-processed data is extracted, including the student's movement speed, strength, coordination indicators, the teacher's teaching style, interaction mode features, and the fractal dimension of the movement sequence; through pattern recognition technology, the student's movement pattern and the teacher's teaching behavior pattern in sports activities are identified; S23. Movement performance and skill mastery analysis: based on the extracted features and identified patterns, the student's movement performance and skill mastery are analyzed, the student's movement skill level, action standardization, and progress speed are evaluated, the problems and areas for improvement in the movement process are identified, and the influence of different sports items and training plans on the student's movement performance is analyzed; S24. Teaching effectiveness evaluation: analyze the teacher's teaching style, interaction mode, and explanation clarity, identify the strengths and weaknesses in the teaching process, provide teaching feedback suggestions, and store the analysis results in the database.
4. The method of claim 1, wherein, based on the data analysis results, a dynamic evaluation model of sports teaching quality is constructed to evaluate the sports teaching process, and the specific steps are as follows: S31. Evaluation index determination and weight distribution: determine the evaluation indexes of student movement performance, skill mastery, psychological state, and teacher teaching design and classroom interaction, and distribute the weights; S32. Dynamic evaluation model design and construction: a dynamic evaluation model is constructed using machine learning and deep learning techniques, and the model integrates a movement efficiency comprehensive evaluation function: η=(α×A+β×S+θ×P) / FD, where η is the movement efficiency comprehensive evaluation index; A is the action accuracy, with a value range of [0, 100]; S is the movement stability, with a value range of [0, 100]; P is the physiological load index, with a value range of [0, 1]; α, β, θ are weight coefficients, with a value range of [0.2, 0.5], and satisfy α+β+θ=1, and the optimal value is determined by model training; FD is the fractal dimension of the movement sequence, with a value range of [1, 2], and the optimal interval is 1.2 to 1.5; S33. Model verification and optimization adjustment: the model is applied to actual teaching scenarios for verification, the accuracy and reliability of the model are evaluated by comparing the model evaluation results with the actual situation, and the model is adjusted and optimized according to the verification results.
5. The method of claim 1, wherein, based on the evaluation model results, teaching design optimization, classroom management improvement, and teaching evaluation adjustment are provided for teachers, and personalized learning paths and learning resources are provided for students, and the specific steps are as follows: S41. Generation of personalized teaching suggestions: based on the output results of the dynamic evaluation model, personalized teaching suggestions covering teaching design optimization, classroom management improvement, and teaching evaluation adjustment are provided for teachers; S42. Develop classroom management optimization plan: based on the analysis of students' classroom performance by the evaluation model, develop a classroom management optimization plan for teachers, including suggestions for maintaining classroom discipline, improving student participation, and creating a classroom atmosphere; S43. Individualized learning path planning for students: based on the analysis of students' physical fitness, skill level, and interest by the evaluation model, plan an individualized learning path for students, including recommending suitable sports projects, training plans, and learning resources; S44. Learning resource recommendation and access support: recommend relevant teaching videos, training guidance, and online courses based on students' learning paths and individualized needs, and provide resource links, download methods, and usage methods.
6. An artificial intelligence-based sports teaching quality dynamic evaluation system, characterized in that, The system for dynamically evaluating the quality of sports teaching based on artificial intelligence is used to implement the method for dynamically evaluating the quality of sports teaching based on artificial intelligence according to any one of claims 1-5; the system for dynamically evaluating the quality of sports teaching based on artificial intelligence comprises a data collection module, a data analysis module, an evaluation model construction module, and a feedback module; The data collection module uses sensing technology to collect real-time data on students' physical, postural, and physiological data during sports activities, and simultaneously captures teachers' teaching behavior and explanation content through video recording and voice recognition technology; The data analysis module cleans, denoises, and formats the collected data, uses machine learning and deep learning models to extract features and identify patterns, analyzes students' sports performance, skill mastery, and teachers' teaching effectiveness; The evaluation model construction module constructs a dynamic evaluation model based on data analysis, comprehensively considers students' physical fitness, skills, interests, and teachers' teaching strategies and styles, and objectively evaluates sports teaching; The feedback module provides individualized teaching suggestions for teachers based on the evaluation results, including design optimization, management improvement, and evaluation adjustment, and customizes learning paths and resources for students.