Teaching simulation system based on VR glasses

Abstract

The invention discloses a teaching simulation system based on VR glasses, particularly relates to the technical field of education, and comprises a data acquisition module, a data processing and analysis module, a teaching content adjustment module, a feedback generation module and a learning terminal. The teaching simulation system based on the VR glasses comprises a data acquisition module, a processing analysis module, a teaching content adjustment module, a feedback generation module, a learning terminal module and the like. The data acquisition module acquires multi-aspect data of students through various sensors and a VR auxiliary acquisition module; the data processing and analysis server processes the data by using technologies such as deep learning and calculates the attention index and emotion intensity of the student so as to determine subsequent operation; the teaching content adjusting module dynamically adjusts the teaching difficulty and mode according to the analysis result; the feedback generation module provides emotional feedback; the learning terminal comprises VR glasses and related equipment for assisting teaching. According to the system, various advanced technologies are integrated, and a highly personalized, immersive, intelligent and adaptive teaching environment is created for students.

Description

Technical Field

[0001] The present invention relates to the technical field of education and teaching, and particularly relates to a teaching simulation system based on VR glasses. Background Art

[0002] In the current education and teaching environment, traditional teaching methods are difficult to meet the personalized learning needs of students, and in the scenario of remote teaching, students lack an immersive learning experience and accurate analysis and adaptive adjustment of their learning status.

[0003] The patent with the publication number CN106981225A discloses a teaching scenario simulation system and method based on VR glasses. The system includes: an image acquisition device, a data management terminal, a network server, and VR glasses. The network server includes a scenario reproduction module. The image acquisition device collects three-dimensional images of students and the classroom during the teaching process in real time and sends the three-dimensional image data to the data management terminal. The network server establishes three-dimensional models of students and the classroom according to the three-dimensional image data according to a preset model establishment rule. The scenario reproduction module generates a scenario of a teacher giving a lesson to students in the classroom according to the three-dimensional models of students and the virtual classroom. At the student terminals in different positions, for example, students at home wear VR glasses. The VR glasses receive the scenario sent by the data management terminal and display it, making the students feel as if they are really having classes in the classroom with their classmates, which can enable students to be more focused on remote learning and improve the students' class efficiency.

[0004] Although the above patent document has certain improvements, it has not been able to comprehensively integrate technologies such as deep learning, machine learning, context awareness, emotion computing, and emerging VR technologies to achieve an efficient, intelligent, and teaching process that fits the actual situation of students. Summary of the Invention

[0005] The main purpose of the present invention is to provide a teaching simulation system based on VR glasses, which can effectively solve the problems in the background art.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] A teaching simulation system based on VR glasses includes:

[0008] A data acquisition module, which is used to collect students' behavior data, voice expression data, and physiological signal data;

[0009] A data processing and analysis module, which is equipped with deep learning and machine learning algorithm models and is used to process and analyze the collected data, and calculate the attention index of students and the intensity of students' positive emotions;

[0010] A teaching content adjustment module for adjusting teaching content, difficulty, and teaching methods according to the results obtained by the data processing and analysis module;

[0011] A feedback generation module that generates emotional feedback based on the emotional state and learning situation of students;

[0012] A learning terminal, which includes VR glasses, for receiving teaching content and feedback information, and is connected to the data acquisition module as one of the front-end devices for data acquisition.

[0013] Preferably, the data acquisition module specifically includes: a camera for collecting students' facial expressions, a microphone array for collecting students' voice information, a heart rate sensor for collecting students' heart rates, and a VR-assisted acquisition module for identifying and collecting objects and actions related to learning in the real environment by students.

[0014] Preferably, the data processing and analysis module calculates the attention index of students according to the following formula:

[0015]

[0016] where P is the attention index of the student, t f is the time of focused behavior, and t d is the time of distracted behavior;

[0017] The data processing and analysis module calculates the positive emotion intensity of students according to the following formula:

[0018] ES p = w v × v p + w f × f p + w a × a p

[0019] where ES p is the positive emotion intensity of the student, v p is the positive emotional feature value of the voice, f p is the positive emotional feature value of the facial expression, a p is the positive emotional feature value of the body movement, and w v , w f , w a are the corresponding weights respectively.

[0020] Preferably, the data processing and analysis module also has:

[0021] A distributed data processing architecture based on federated learning, which allows data from different schools or educational institutions to be jointly trained and model updated while protecting privacy;

[0022] An innovative adaptive algorithm adjustment mechanism that automatically adjusts the hyperparameters of deep learning and machine learning algorithms according to the learning history and learning style of individual students;

[0023] A 3D model construction and analysis sub-module optimized for VR teaching scenarios, which uses the collected data to construct 3D models of students and teaching scenarios, and combines the results of context awareness and sentiment analysis to dynamically adjust the display effect and interaction method of the models in the VR environment.

[0024] Preferably, the teaching content adjustment module includes: a multi-dimensional difficulty adjustment strategy, a multi-mode teaching strategy switching system, a personalized teaching path planner, and a teaching interaction enhancement unit based on VR interaction.

[0025] Preferably, the feedback generation module includes: a dynamic virtual teacher image generation sub-module based on emotional state, a cross-modal emotional feedback generation sub-module, an emotional feedback effect evaluation sub-module, and a feedback presentation optimization sub-module integrated into the VR scene.

[0026] Preferably, the learning terminal further includes an ultra-high-definition display sub-module, an intelligent interaction sub-module, and an adaptive learning progress storage and reminder sub-module.

[0027] Preferably, the data processing and analysis server further includes an abnormal data detection and repair sub-module, which uses the outlier detection algorithm of machine learning to detect the collected data, identifies data points that may be incorrect or abnormal, and uses a combination of data interpolation and model prediction to repair the abnormal data.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] 1. The teaching simulation system based on VR glasses of the present invention has many significant advantages. At the level of data collection and analysis, it is equipped with a variety of advanced devices such as high-precision cameras and high-sensitivity microphone arrays, combined with special image processing chips, adaptive gain control technology, etc., which can comprehensively and accurately collect data such as students' learning behaviors, voices, and physiological signals, and accurately analyze the students' attention index and emotional intensity with the help of unique calculation formulas and efficient data processing servers, laying a solid foundation for the adjustment of subsequent teaching strategies. At the same time, it not only ensures data security and the generalization ability of the model, but also greatly improves the accuracy and efficiency of analysis. The abnormal data detection and repair sub-module further ensures the reliability of the data.

[0030] 2. The present invention also performs excellently in the teaching implementation and feedback links. The multi-dimensional difficulty adjustment strategy, multi-mode teaching strategy switching system, personalized teaching path planner, and teaching interaction enhancement unit of the teaching content adjustment module are closely combined with the learning situation, progress, style, and interaction performance of students to customize teaching content and paths for students, effectively avoiding repetitive or jumping learning, greatly improving learning effects and enthusiasm, and also cultivating students' teamwork ability. A series of functions of the feedback generation module, such as dynamic virtual teacher images, cross-modal emotional feedback, effect evaluators, and feedback presentation optimization integrating VR scenarios, can provide rich, accurate, and effective emotional feedback according to students' emotions and sensory preferences, significantly enhancing the affinity and pertinence of teaching, and creating a highly personalized, immersive, and intelligent adaptive high-quality teaching environment for students. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a block diagram of the work flow module of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0032] To make the technical means, creative features, achieved purposes, and functions of the present invention easy to understand, the present invention will be further described below in conjunction with specific embodiments.

[0033] As Figure 1 shown, the present invention discloses a teaching simulation system based on VR glasses, which specifically includes:

[0034] A data acquisition module for collecting students' behavioral data, voice expression data, and physiological signal data;

[0035] A data processing and analysis module, which is equipped with deep learning and machine learning algorithm models for processing and analyzing the collected data, and calculating students' attention index and students' positive emotion intensity;

[0036] A teaching content adjustment module for adjusting teaching content, difficulty, and teaching methods according to the results obtained by the data processing and analysis module;

[0037] A feedback generation module for generating emotional feedback based on students' emotional states and learning situations;

[0038] A learning terminal, which includes VR glasses for receiving teaching content and feedback information, and is connected to the data acquisition module as one of the front-end devices for data acquisition.

[0039] Among them, the data acquisition module specifically includes:

[0040] A camera for collecting students' facial expressions, which can collect high-resolution images of students' facial expressions at a rate of at least 50 frames per second. The camera has intelligent focusing and intelligent light adjustment functions to ensure clear and accurate capture of the subtle muscle movements and expression changes on students' faces under different lighting conditions. It is equipped with a dedicated image processing chip that uses a deep learning-based image enhancement algorithm to perform real-time preprocessing on the collected images, automatically removing interference elements in the images, such as shadows, reflected light, etc., and further improving the accuracy of facial expression recognition;

[0041] A microphone array for collecting students' voice information, which can collect students' voices from different directions and has echo cancellation and noise reduction functions to ensure that the collected voice signals are clear and distinguishable. At the same time, through the collaborative work of multiple microphones, it can achieve the positioning of students' voices and accurately judge the direction of the sound source;

[0042] A heart rate sensor for collecting students' heart rates, which can simultaneously monitor the heart rate changes of students in different frequency bands. By analyzing the heart rate data in different frequency bands, it can further judge the physiological and psychological states of students. For example, by analyzing the ratio of low-frequency and high-frequency heart rate data, it can judge whether a student is in a relaxed state or a tense state;

[0043] A VR-assisted acquisition module for identifying and collecting learning-related objects and actions of students in the real environment, which uses the spatial perception and marker recognition capabilities of VR glasses to identify learning-related objects and actions in the real environment of students, further enriching the dimension of learning behavior data collection.

[0044] Furthermore, the data processing and analysis module calculates the attention index of students according to the following formula:

[0045]

[0046] where P is the attention index of the student, t f is the time of focused behavior, and t d is the time of distracted behavior;

[0047] The data processing and analysis module calculates the positive emotion intensity of students according to the following formula:

[0048] ES p =w v ×v p +w f ×f p +w a ×a p

[0049] where ES p is the positive emotion intensity of the student, v pis the positive emotional feature value of speech, f p is the positive emotional feature value of facial expression, a p is the positive emotional feature value of body movement, w v 、w f 、w a are the corresponding weights respectively;

[0050] Furthermore, in the present invention, the positive emotional feature values and weights of speech, facial expression, and body movement are obtained in the following manner:

[0051] Positive emotional feature value of speech (v p ):

[0052] First, the collected student speech signal is preprocessed, including operations such as noise reduction and filtering, to improve the speech quality.

[0053] Then, using speech emotion analysis technology, the prosodic features of speech (such as changes in pitch, speech rate, and volume), timbre features (such as formant frequencies, etc.), and semantic features (analyzing the meaning of words and sentences in speech through natural language processing technology) are extracted.

[0054] These features are input into a speech emotion classification model based on deep learning, which is trained with a large amount of speech data labeled with different emotional states. The model outputs a value between 0 and 1, representing the positive emotional degree of the speech, that is, v p . For example, if the speech contains more high-pitched tones, a fast speech rate, and semantically positive words, the model may output a higher v p value;

[0055] Positive emotional feature value of facial expression (f p ):

[0056] The student facial images collected by a high-precision camera are preprocessed based on deep learning to remove interference elements and enhance facial features.

[0057] Using a facial expression recognition algorithm, the key facial feature points (such as the position and shape changes of parts such as eyes, mouth, and eyebrows) are recognized.

[0058] These features are input into a convolutional neural network model trained with a large amount of labeled facial expression images, which can recognize different facial expression categories (such as happy, smiling, excited, etc.) and correspondingly output a positive emotional feature value f p . For example, when it is recognized that the student's face shows an obvious smiling expression, the f p value will be relatively high;

[0059] Positive emotional feature value of body movement (ap ):

[0060] Collect the body movement data of students through cameras or other motion capture devices, analyze and process the data, and extract the amplitude, speed, frequency of the movement, and the type of movement (for example, positive body movements may include raising hands, nodding, leaning forward, etc., and negative ones may include lowering the head, lying on the table, shaking the body frequently, etc.).

[0061] Use machine learning algorithms to classify and quantify these features, and the training data is a large number of samples labeled with body movements of different emotional tendencies. Obtain the positive emotion feature value a of the body movement according to the classification and quantification results p . For example, if a student raises their hand frequently and forcefully to participate in the interaction, it will correspond to a higher a p value;

[0062] Weights (w v , w f , w a ):

[0063] The weights are determined by methods based on big data analysis and machine learning. First, collect a large amount of multimodal data (speech, facial expressions, body movements) of students in different learning scenarios, as well as the corresponding learning effect and emotional state evaluation results.

[0064] Divide these data into a training set and a validation set, and use models such as multiple linear regression and neural networks for training to learn the relative importance of speech, facial expressions, and body movements in the overall positive emotion judgment under different circumstances.

[0065] After continuous training and optimization, obtain the weights w v , w f , w a suitable for this teaching system. For example, in some courses mainly focused on oral communication, it may be found that the speech emotion features are more critical for the overall emotion judgment, and at this time the value of w v will be relatively high; while in some courses that focus on practical operations, the weight w a of body movements may be more prominent;

[0066] Through the above methods, the positive emotion feature values and weights of speech, facial expressions, and body movements required in the formula can be accurately obtained, so as to effectively calculate the positive emotion intensity of students and provide a reliable basis for subsequent teaching strategy adjustment and emotional feedback.

[0067] Specifically, the deep learning and machine learning algorithm models of the data processing and analysis module include convolutional neural networks and recurrent neural networks, which are used to analyze the facial expressions and voices of students. Among them, the convolutional neural network adopts a novel residual connection structure. By increasing the number and depth of residual blocks, it can effectively solve the problem of gradient disappearance and improve the model's ability to extract subtle expression and voice features. The recurrent neural network adopts a hybrid structure of gated recurrent unit (GRU) and long short-term memory network (LSTM), combining the advantages of both to perform more accurate sentiment tendency and semantic understanding on the voice sequence of students.

[0068] It should be noted that the data processing and analysis module also has:

[0069] A distributed data processing architecture based on federated learning, which allows data from different schools or educational institutions to be jointly trained and model updated while protecting privacy. That is, the local data of each institution is trained locally, and only the gradient information of model parameter updates is aggregated and exchanged, avoiding the transmission of raw data, which not only ensures data security but also improves the generalization ability of the model. And this architecture adopts dynamic network topology optimization technology, which can automatically adjust the network connection method according to the number of participating institutions and data traffic to improve data transmission efficiency;

[0070] An innovative adaptive algorithm adjustment mechanism, which automatically adjusts the hyperparameters of deep learning and machine learning algorithms, such as learning rate, regularization parameter, etc., according to the learning history and learning style of individual students, enabling the algorithm to quickly adapt to the characteristics of different students, improving the accuracy and efficiency of analysis. At the same time, this mechanism also introduces a meta-learning algorithm, which can learn and adapt to new student groups and learning scenarios faster, further enhancing the adaptive ability;

[0071] A 3D model construction and analysis sub-module optimized for VR teaching scenarios, which uses the collected data to construct 3D models of students and teaching scenarios, and combines the results of context awareness and sentiment analysis to dynamically adjust the display effect and interaction method of the model in the VR environment. This sub-module can also automatically generate virtual teaching aids and scene elements according to the teaching content and students' cognitive levels. For example, it can generate virtual experimental instruments and dynamic demonstration scenarios in physics courses to enhance the intuitiveness of teaching.

[0072] Furthermore, the teaching content adjustment module includes:

[0073] The multi-dimensional difficulty adjustment strategy can not only adjust the difficulty level of teaching content according to students' learning situation, but also adjust the presentation form of teaching content. For example, for knowledge points with higher difficulty, abstract concepts can be converted into visual 3D animations or virtual reality scenes to make it easier for students to understand. In addition, an intelligent content recommendation algorithm is used in the adjustment process to recommend relevant extended learning materials and cases based on students' interests and knowledge weaknesses.

[0074] The multi-mode teaching strategy switching system can seamlessly switch between multiple teaching modes such as theoretical explanation, example demonstration, experimental simulation, group discussion, role-playing, etc., and automatically select the most suitable transition method according to students' real-time feedback and emotional state during the switching process to ensure the continuity and fluency of teaching. At the same time, the system also has an automatic teaching mode evaluation function, which can evaluate and optimize the switched teaching mode according to students' learning effects;

[0075] Personalized teaching path planner, which customizes a unique teaching path for each student based on their learning progress and mastery level, avoiding unnecessary repetitive learning or skipping learning in the learning process, ensuring the systematicness and coherence of the knowledge system. The planner can also dynamically adjust the teaching path according to the student's learning speed and ability changes, realizing truly personalized teaching;

[0076] The teaching interaction enhancement unit based on VR interaction can use the handle and gesture recognition function of VR glasses to design special interactive teaching links, such as virtual experiment operation, 3D model assembly, etc., so that students can better understand knowledge in practice. At the same time, the subsequent teaching content and difficulty can be adjusted according to the students' performance in the interaction. The unit also supports multi-person collaborative interactive teaching. Students can complete learning tasks with other classmates in a VR environment and cultivate teamwork ability.

[0077] In this embodiment, the teaching content adjustment module uses a reinforcement learning algorithm to train the teaching strategy adjustment model, optimizes the teaching strategy selection according to the students' learning effects and feedback, and uses a priority-based experience playback mechanism in the reinforcement learning algorithm to give priority to training learning experiences that have a greater impact on the teaching effect, thereby accelerating the convergence of the model. At the same time, in the design of the reward function, not only the improvement of academic performance is considered, but also comprehensive considerations of students' emotional experience, learning engagement and other factors, making the reward more comprehensive and reasonable.

[0078] The feedback generation module in the present invention includes:

[0079] The dynamic virtual teacher image generation sub-module based on emotional state can adjust the appearance, expression, movements, and dressing style of the virtual teacher in real time according to the emotional state of the students, making the image of the virtual teacher more amiable and targeted. For example, when the students are in an excited state, the virtual teacher will wear more colorful clothes and make more lively movements. Moreover, the system can also generate corresponding professional images and dressing styles for the virtual teacher according to different subjects and teaching contents, enhancing the professionalism of teaching;

[0080] The cross-modal emotional feedback generation sub-module can not only provide feedback through voice, text, expressions, etc., but also offer a more rich emotional feedback experience for students by combining various modalities such as environmental sound effects and tactile feedback (such as the vibration of VR glasses). The system can also automatically adjust the priority of the feedback modalities according to the sensory preferences of the students. For example, it will increase the proportion of visual feedback for visual learners;

[0081] The emotional feedback effect evaluation sub-module can evaluate the impact of emotional feedback on students in real time. According to the students' responses to the feedback, such as subsequent learning attitudes and behavioral changes, it can dynamically adjust the content and method of emotional feedback. And this evaluator adopts a machine learning classification algorithm, which can automatically classify and analyze the students' responses to improve the accuracy of the evaluation;

[0082] The feedback presentation optimization sub-module integrated into the VR scene skillfully integrates emotional feedback into the VR teaching scene. This sub-module can also reasonably arrange the display position and method of feedback information according to the layout of the VR scene and the key points of the teaching content. For example, it will display prompt information near the knowledge points that the students are concerned about to improve the effectiveness of feedback.

[0083] The learning terminal in the present invention further includes:

[0084] The ultra-high-definition display sub-module supports high-resolution display of teaching contents and the image of the virtual teacher. At the same time, it has a blue light protection function to reduce the harm to the students' eyes caused by long-term learning. And this display system also supports HDR (High Dynamic Range) display technology, which can present the details and colors of teaching contents more vividly;

[0085] The intelligent interaction sub-module supports various interaction methods, such as gesture recognition, eye tracking, voice interaction, etc., providing a more natural and convenient learning interaction experience for students. The system also has an interaction behavior analysis function, which can record and analyze the students' interaction behaviors to provide data support for teaching improvement;

[0086] The adaptive learning progress storage and reminder sub-module can automatically store the learning progress of students, remind students to continue learning at appropriate times according to their learning habits and plans, recommend corresponding learning resources and review plans to students based on their learning progress, and the system can also automatically adjust the recommendation strategy according to the changes in students' learning achievements and abilities to provide more accurate learning support.

[0087] In addition, the data processing and analysis server in the present invention further includes an abnormal data detection and repair sub-module. This sub-module uses the outlier detection algorithm of machine learning to detect the collected data, identify data points that may be incorrect or abnormal, and repair the abnormal data by combining data interpolation and model prediction.

[0088] The specific implementation manner of the present invention is as follows:

[0089] 1. System initialization;

[0090] When starting the system, calibrate and initialize each device and module. Calibrate the sensors to ensure accurate data collection, load the trained models, set the teaching parameters and initial teaching strategies, and at the same time initialize the display and interaction settings of the VR glasses to ensure that students can smoothly enter the teaching environment.

[0091] 2. Data collection and transmission;

[0092] During the teaching process, the data collection module continuously collects data. The VR glasses transmit the interaction information and the collected environmental data in real time, and send all the data to the data processing and analysis server quickly and stably through wired or wireless means. For example, the camera and microphone transmit data once per second, and the heart rate sensor and brain wave sensor transmit data every 0.5 seconds to ensure the timeliness and effectiveness of the data.

[0093] 3. Data analysis and decision-making;

[0094] After receiving the data, the data processing and analysis server immediately processes and analyzes it. Use deep learning and machine learning algorithms to identify students' expressions, speech emotions and behavioral intentions, combine the results of context awareness and emotion computing, and judge the students' states through calculation formulas. If it is found that a student is inattentive and anxious during the learning of chemical equation balancing, and it is determined through multi-source data fusion analysis that it is due to the difficulty in understanding the concept of redox reactions, immediately decide to adjust the teaching strategy.

[0095] 4. Teaching adjustment and feedback;

[0096] The teaching content adjustment module adjusts the teaching content and methods according to the decision, such as displaying the concept of redox reaction in the form of animation in the VR scene and adding an interactive Q&A session. The feedback generation module generates corresponding emotional feedback, such as the virtual teacher encouraging the students in a gentle tone and popping up a prompt box in the VR scene to display the problem-solving ideas. At the same time, using the tactile feedback function of the VR glasses, a slight vibration prompt is given when the student answers correctly to enhance the feedback effect.

[0097] 5. Continuous monitoring and optimization;

[0098] The system continuously monitors the students' learning status, updates the data analysis results, and optimizes the subsequent teaching according to the students' responses to the adjustment and feedback. For example, if the students' learning efficiency improves and their emotions are stable after receiving the adjustment, the system records the effectiveness of the strategy, gives priority to using it in subsequent similar situations, and further adjusts and optimizes the teaching process.

[0099] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. What is described in the above embodiments and the specification only illustrates the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention will have various changes and improvements, and these changes and improvements all fall within the scope of the present invention claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A teaching simulation system based on VR glasses, characterized in that: include: Data collection module, used to collect students' behavior data, speech expression data and physiological signal data; The data processing and analysis module is equipped with deep learning and machine learning algorithm models to process and analyze the collected data and calculate the students' attention index and the intensity of their positive emotions; The teaching content adjustment module is used to adjust the teaching content, difficulty and teaching method according to the results obtained by the data processing and analysis module; Feedback generation module, which generates emotional feedback based on students' emotional state and learning situation; The learning terminal includes VR glasses, which are used to receive teaching content and feedback information, and are connected to the data acquisition module as one of the front-end devices for data acquisition.

2. The teaching simulation system based on VR glasses according to claim 1, characterized in that: The data acquisition module specifically includes: a camera for collecting students' facial expressions, a microphone array for collecting students' voice information, a heart rate sensor for collecting students' heart rate, and a VR-assisted acquisition module for identifying and collecting students' learning-related objects and actions in a real environment.

3. The teaching simulation system based on VR glasses according to claim 1, characterized in that: The data processing and analysis module calculates the student's attention index according to the following formula: Among them, P is the student's attention index, t f is the focused behavior time, t d Time for distracting behaviors; The data processing and analysis module calculates the positive emotion intensity of the students according to the following formula: ES p =w v ×v p +w f ×f p +w a ×a p Among them, ES p is the positive emotion intensity of students, v p is the positive emotional feature value of speech, f p is the positive emotional feature value of facial expression, a p is the positive emotional feature value of body movements, w v 、w f 、w a are the corresponding weights respectively.

4. The teaching simulation system based on VR glasses according to claim 3 is characterized in that: The data processing and analysis module also has: A distributed data processing architecture based on federated learning, which allows data from different schools or educational institutions to be jointly trained and model updated while protecting privacy; Innovative adaptive algorithm adjustment mechanism, which automatically adjusts the hyperparameters of deep learning and machine learning algorithms based on individual students’ learning history and learning style; The 3D model construction and analysis submodule is optimized for VR teaching scenarios. It uses the collected data to build 3D models of students and teaching scenarios, and combines the results of situational awareness and sentiment analysis to dynamically adjust the display effect and interaction mode of the model in the VR environment.

5. The teaching simulation system based on VR glasses according to claim 1 is characterized in that: The teaching content adjustment module includes: a multi-dimensional difficulty adjustment strategy, a multi-mode teaching strategy switching system, a personalized teaching path planner, and a teaching interaction enhancement unit based on VR interaction.

6. The teaching simulation system based on VR glasses according to claim 1, characterized in that: The feedback generation module includes: a dynamic virtual teacher image generation submodule based on emotional state, a cross-modal emotional feedback generation submodule, an emotional feedback effect evaluation submodule, and a feedback presentation optimization submodule integrated into VR scenes.

7. The teaching simulation system based on VR glasses according to claim 1 is characterized in that: The learning terminal also includes an ultra-high-definition display submodule, an intelligent interaction submodule, and an adaptive learning progress storage and reminder submodule.

8. The teaching simulation system based on VR glasses according to claim 1, characterized in that: The data processing and analysis server also includes an abnormal data detection and repair submodule, which uses a machine learning outlier detection algorithm to detect the collected data, identify data points that may have errors or anomalies, and repair the abnormal data using a method that combines data interpolation and model prediction.

Images