A smart digital human control system and method

Abstract

The present invention discloses a smart digital human control system and method, which relates to the field of educational information technology. The technical solution solves the problems of insufficient interactive intelligence, lack of emotional feedback and limited personalized teaching by dynamically adjusting the teaching mode. The system first generates a learning state index Xlz by collecting students' learning progress data and interactive feedback data in real time, and generates a comprehensive state value Fhxs by combining data such as the emotional change index Ex and the learning task difficulty value Nx, and dynamically adjusts the teaching mode; secondly, the emotion recognition technology provides emotional support by analyzing the emotional fluctuation trend of students, including encouraging feedback and interactive measures to reduce stress; finally, the system calculates the learning demand assessment value Lhx according to the student's learning progress rate Xjz, learning resistance index Zlx and learning interest value Xqx, and automatically adjusts the teaching difficulty to improve the personalized teaching effect, and achieves more accurate learning optimization and interactive experience.

Description

Technical Field

[0001] The present invention relates to the field of educational information technology, and in particular to a smart digital human control system and method. Background Art

[0002] The concept of digital human can be traced back to science fiction works in the 20th century. With the advancement of computer technology, the prototype of digital human has gradually been reflected in virtual reality and computer animation. Entering the 21st century, with the rapid development of artificial intelligence, deep learning, virtual reality, augmented reality and other technologies, digital human technology has matured rapidly. In the early days, digital human was mainly used in film and television animation and games. With the advancement of technology, digital human has not only significantly improved in visual fidelity, but also has autonomous interaction and intelligent behavior, becoming an important tool for many industries.

[0003] At the same time, digital humans are widely used in the education industry, including for creating virtual teachers. However, the digital human control system for creating virtual teachers has the following technical shortcomings in actual applications: 1. Insufficient interactive intelligence: Although current virtual teachers can provide basic teaching content, they have poor flexibility and accuracy in intelligent interaction when facing complex student questions or personalized needs, and it is difficult to achieve the deep interactive effect of real teachers. Especially in natural language processing and context understanding, the system often cannot correctly understand the diverse expressions of students.

[0004] 2. Lack of emotional feedback: Existing virtual teachers have limited emotional understanding and feedback, and cannot accurately identify and respond to students' emotional states, such as frustration, anxiety, or changes in interests. This lack of emotional understanding may lead to poor student experience in the virtual learning environment, which in turn affects learning outcomes and learning enthusiasm.

[0005] 3. Limited personalized teaching: Although virtual teachers have certain personalized teaching capabilities, they are limited in dynamically adjusting according to students' learning habits, learning progress, and knowledge mastery. The system's algorithm optimization in data analysis and real-time adjustment of teaching strategies is not perfect enough, resulting in personalized recommendations and adaptive teaching effects that are far less than expected. Summary of the invention

[0006] In view of the deficiencies in the prior art, the present invention provides a smart digital human control system and method, which solves the technical problems of insufficient interactive intelligence, lack of emotional feedback and limited personalized teaching in the background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a method for controlling a smart digital human, comprising the following steps: S1. Pre-set the interaction scenarios of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; S2. Real-time monitoring of students' learning progress-related data and interactive feedback-related data to generate a learning status data set; based on the learning status data set, determining students who need personalized adjustments from a number of students, and marking the students who need personalized adjustments; S3. According to the marked students obtained in step S2, the learning habit data and emotional feedback related data of the students are analyzed, a habit and emotional data set is constructed, and a comprehensive state value Fhxs of the student with sequence number i is generated and evaluated according to the habit and emotional data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; S4. After re-adjusting the teaching content, the secondary learning need assessment value Lhxc is calculated and evaluated again, and finally, according to the evaluation content of the secondary learning need assessment value Lhxc, it is determined whether the teaching mode of the students needs to be further adjusted, including increasing emotional support; S5. According to the adjusted teaching model, continue to monitor students' learning feedback and the virtual teacher's interaction strategy; enable the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

[0008] Preferably, S1 includes S11, S12 and S13, specifically: S11. First, design the interaction scenarios between virtual teachers and students based on teaching content, learning objectives and interaction needs. The interaction scenarios include one-on-one teaching mode, classroom discussion mode and question-answering mode; and each scenario is customized according to the age, knowledge level and course characteristics of the students, and the performance of the virtual teacher can adapt to the teaching needs in different scenarios; then obtain the relevant data of students' learning needs through historical data, and the acquisition of students' relevant data of needs is carried out through online questionnaires, learning platform records and teacher feedback; S12. Divide the course into several groups of teaching units according to the syllabus and student needs. Each teaching unit represents a specific knowledge point or course unit, including grammar, reading and mathematical formulas. The teaching units are divided according to the difficulty of the course, the logical order of the content and the differences in student needs. Modules are locked according to the current learning progress and needs of students. S13. Label all teaching units with serial numbers, specifically JC1, JC2...JCn, where JC1, JC2...JCn represent the first teaching unit, the second teaching unit...the nth teaching unit respectively, and n is the total number of teaching units.

[0009] Preferably, S2 includes S21 and S22, specifically: S21, extracting a learning state data set, and dimensionlessly processing the relevant data in the learning state data set, and obtaining a learning state index Xlz by the following formula: ; Where Kwc represents the course completion value in the learning status data set, Zzw represents the knowledge point mastery value in the learning status data set, Xxt represents the learning time in the learning status data set, Tws represents the number of questions asked in the learning status data set, Hcy represents the interactive participation in the learning status data set, and Zzd represents the concentration in the learning status data set.

[0010] Preferably, S22, by comparing and evaluating the preset learning goal Lmz with the student's current learning status index Xlz, determine the students who need personalized adjustment, and the specific evaluation content is as follows: When the learning status index Xlz≤the set learning goal Lmz, the corresponding students are identified as students who need personalized adjustment and marked; When the learning status index Xlz>learning target Lmz, it means that the corresponding students do not need personalized adjustments and are not marked.

[0011] Preferably, S3 includes S31, S32, S33 and S34, specifically: S31, collecting the learning habit data and emotional feedback related data of the students marked with personalized adjustment in S2, and constructing a habit and emotional data set after preprocessing and dimensionless processing; wherein the habit and emotional data set includes the students' emotional change index Ex and the learning task difficulty value Nx; then marking the students marked with personalized adjustment with serial numbers, including the students with serial number i; S32. By extracting the habit and emotion data set, the comprehensive state value Fhxs of the student with sequence number i is calculated by the following formula: ; Where Lrx represents the learning endurance index in the habit and emotion data set, and Cox represents the cognitive load index in the habit and emotion data set.

[0012] Preferably, S33, by presetting the first comprehensive state threshold W1 and the second comprehensive state threshold W2, the comprehensive state value Fhxs of the student with sequence number i is evaluated, and the first comprehensive state threshold W1 is greater than the second comprehensive state threshold W2, and the specific evaluation content is as follows: When the student comprehensive state value Fhxs ≥ the first comprehensive state threshold W1, the first teaching mode is generated, and the virtual teacher gives priority to adjusting the teaching content and provides personalized content; When the first comprehensive state threshold W1>comprehensive state value Fhxs≥second comprehensive state threshold W2, the second teaching mode is generated at this time, and the virtual teacher continues to execute according to the current standard teaching plan; When the second comprehensive state threshold W2>comprehensive state value Fhxs, the third teaching mode is generated, and the virtual teacher gives priority to emotional support and then makes teaching adjustments.

[0013] Preferably, S34, after dynamically adjusting the teaching mode of the student with sequence number i, real-time monitoring and collection of the learning rhythm and state-related data of the student with sequence number i, including the knowledge mastering progress value Kcz, the learning resistance index Zlx and the learning interest value Xqx; after dimensionless processing of the learning rhythm and state-related data of the student with sequence number i, the learning demand assessment value Lhx is calculated by the following formula: ; In the formula, a1, a2 and a3 are adjustment coefficients, and their specific values ​​are adjusted by the user; a1 is used to adjust the influence of the knowledge mastering progress value Kcz, a2 is used to adjust the influence of the learning interest value Xqx on the formula, and a3 is used to adjust the influence of the learning resistance index Zlx; The preset learning demand threshold E is compared with the learning demand assessment value Lhx to make an assessment and adjust the teaching content. The specific assessment content is as follows: If the learning need assessment value Lhx ≥ the learning need threshold E, it means that the learning need of the student with sequence number i is normal. At this time, the difficulty of the teaching content will be increased by 5% based on the original teaching content; If the learning needs assessment value Lhx is less than the learning needs threshold E, it means that the learning needs of the student with sequence number i are abnormal. At this time, the original teaching content should be maintained and the students should focus on strengthening the consolidation of basic knowledge. Preferably, S4 includes S41 and S42, specifically: S41. Through the online learning platform, based on the emotion recognition algorithm and behavior tracking equipment, the learning progress rate Xjz, emotion change value Xqg and concentration duration Xzz of the student with sequence number i are collected in real time and dimensionless processed, and the learning state index Xlzc is calculated by the following formula: ; S42, compare and evaluate the secondary learning state index Xlzc with the learning state data Xlz, and judge again whether the virtual teacher needs to adjust the teaching content. The specific contents are as follows: If the secondary learning state index Xlzc ≥ learning state data Xlz, it means that the student's learning state has improved under the adjusted teaching mode and has adapted to the new teaching content. At this time, the existing teaching mode and emotional support should be maintained; If the learning state index Xlzc is less than the learning state data Xlz, it means that the student's learning state has declined under the adjusted teaching mode and is not adapted to the new teaching content. At this time, the teaching mode should be adjusted, including simplifying knowledge points and increasing emotional support.

[0014] Preferably, S5, the virtual teacher continuously monitors the students' learning feedback data and the virtual teacher's interaction strategy based on the adjusted teaching model, including the students' learning progress, classroom emotional state and learning concentration; secondly, the virtual teacher analyzes the students' emotional fluctuations in real time, identifies the changing trend of the emotional state, and based on the results of emotion recognition, the virtual teacher provides emotional support to the students in a timely manner, wherein the emotional support includes positive feedback, encouraging prompts and interactive measures to reduce learning pressure.

[0015] Preferably, a smart digital human control system includes an interactive scene setting module, a learning state monitoring module, a learning habit and emotion analysis module, a teaching content dynamic adjustment module, and a learning feedback and emotion support module; The interactive scene setting module is used to pre-set the interactive scenes of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; The learning status monitoring module is used to monitor students' learning progress related data and interactive feedback related data in real time to generate a learning status data set; based on the learning status data set, students who need personalized adjustment are determined from a number of students, and students who need personalized adjustment are marked; The learning habit and emotion analysis module is used to analyze the learning habit data and emotion feedback related data of the students according to the marked students, construct the habit and emotion data set, and generate and evaluate the comprehensive state value Fhxs of the student with sequence number i according to the habit and emotion data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; The teaching content dynamic adjustment module is used to calculate and evaluate the secondary learning demand evaluation value Lhxc again after re-adjusting the teaching content, and finally determine whether the teaching mode of the students needs to be further adjusted according to the evaluation content of the secondary learning demand evaluation value Lhxc, including increasing emotional support; The learning feedback and emotional support module is used to continue monitoring students' learning feedback and the virtual teacher's interaction strategy according to the adjusted teaching model; enabling the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

[0016] The present invention provides a smart digital human control system and method, which has the following beneficial effects: (1) This intelligent digital human control system and method solves the problem of insufficient interactive intelligence by dynamically adjusting the students' teaching mode based on real-time monitoring of students' learning status, emotional feedback and learning habits. First, the system can provide personalized interactive scenes for students of different age groups and knowledge levels through the interactive scene setting module, and collect learning progress-related data and interactive feedback data in real time through the learning status monitoring module, and generate a learning status index Xlz through the data set. Then, the learning habit and emotion analysis module can accurately analyze the students' learning habit data and emotional feedback-related data, construct a data set containing a series of parameters such as the emotion change index Ex, the learning task difficulty value Nx, the learning endurance index Lrx and the cognitive load index Cox, generate a comprehensive status value Fhxs, and dynamically adjust the teaching mode according to the change of the comprehensive status value. This module solves the technical problem of insufficient intelligent interaction accuracy when the system is faced with complex student questions or personalized needs, and achieves a deeper interactive effect.

[0017] (2) This intelligent digital human control system and method achieves accurate emotional feedback through a learning feedback and emotional support module; the module uses emotion recognition technology to collect students' facial expressions, voices, and behavior data in real time, and analyzes the emotion change index Ex and emotion fluctuation trends; the system can provide positive emotional support for students in different emotional states, such as helping to relieve students' anxiety, frustration and other negative emotions through encouraging prompts and interactive measures to reduce learning pressure; the virtual teacher can dynamically adjust the teaching rhythm and content according to the changing trend of the emotional state to ensure that students receive appropriate support when their emotions fluctuate; this module effectively solves the shortcomings of existing virtual teachers in emotional understanding and feedback, improves students' experience in the virtual learning environment, and enhances their learning enthusiasm.

[0018] (3) A control system and method for an intelligent digital human, the system calculates a learning need assessment value Lhx according to the student's learning progress rate Xjz, learning resistance index Zlx and learning interest value Xqx, and compares and evaluates it in combination with a preset learning need threshold E; when the student's learning need assessment value Lhx is lower than the learning need threshold E, the system can automatically reduce the teaching difficulty, simplify the knowledge points, and strengthen the consolidation of basic knowledge; when the learning need assessment value Lhx is higher than the threshold, the system will increase the difficulty of the teaching content and enhance the learning challenge for the students; through these personalized adjustments, the system overcomes the limitations of traditional virtual teachers in dynamically adjusting teaching strategies, achieves more accurate personalized teaching optimization, and further improves students' learning effects and experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 A schematic diagram of the steps of a method for controlling a smart digital human according to the present invention; Figure 2 The figure is a schematic diagram of the framework structure of a smart digital human control system of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0021] See also Figure 1 The present invention provides a method for controlling an intelligent digital human, comprising the following steps: S1. Pre-set the interaction scenarios of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; S2. Real-time monitoring of students' learning progress-related data and interactive feedback-related data to generate a learning status data set; based on the learning status data set, determining students who need personalized adjustments from a number of students, and marking the students who need personalized adjustments; S3. According to the marked students obtained in step S2, the learning habit data and emotional feedback related data of the students are analyzed, a habit and emotional data set is constructed, and a comprehensive state value Fhxs of the student with sequence number i is generated and evaluated according to the habit and emotional data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; S4. After re-adjusting the teaching content, the secondary learning need assessment value Lhxc is calculated and evaluated again, and finally, according to the evaluation content of the secondary learning need assessment value Lhxc, it is determined whether the teaching mode of the students needs to be further adjusted, including increasing emotional support; S5. According to the adjusted teaching model, continue to monitor students' learning feedback and the virtual teacher's interaction strategy; enable the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

[0022] In this embodiment, S1 achieves precise targeting of students' personalized needs by presetting interactive scenarios and obtaining learning needs data, thereby effectively improving the pertinence of teaching; S2 monitors students' learning progress-related data and interactive feedback data in real time, generates a learning status data set, and accurately identifies students who need personalized adjustments; S3 generates and evaluates a comprehensive status value Fhxs based on learning habits and emotional feedback-related data, and can effectively optimize personalized teaching content and enhance students' learning experience by dynamically adjusting the teaching mode and calculating the learning needs assessment value Lhx; S4 calculates the learning needs assessment value Lhxc, and compares the learning state index Xlzc with the learning state index Xlz to accurately determine whether the teaching mode needs to be further adjusted, including increasing emotional support; S5 monitors students' emotional fluctuations based on emotion recognition technology, and the virtual teacher can adjust the teaching rhythm and content in real time, provide emotional support in a timely manner, and enhance students' learning enthusiasm and concentration. Example

[0023] S1 includes S11, S12 and S13, specifically: S11. First, design the interaction scenarios between virtual teachers and students based on teaching content, learning objectives and interaction needs. The interaction scenarios include one-on-one teaching mode, classroom discussion mode and question-answering mode; and each scenario is customized according to the age, knowledge level and course characteristics of the students, and the performance of the virtual teacher can adapt to the teaching needs in different scenarios; then obtain the relevant data of students' learning needs through historical data, and the acquisition of students' relevant data of needs is carried out through online questionnaires, learning platform records and teacher feedback; S12. Divide the course into several groups of teaching units according to the syllabus and student needs. Each teaching unit represents a specific knowledge point or course unit, including grammar, reading and mathematical formulas. The teaching units are divided according to the difficulty of the course, the logical order of the content and the differences in student needs. Modules are locked according to the current learning progress and needs of students. S13. Label all teaching units with serial numbers, specifically JC1, JC2...JCn, where JC1, JC2...JCn represent the first teaching unit, the second teaching unit...the nth teaching unit respectively, and n is the total number of teaching units.

[0024] In this embodiment, through the implementation of S1, the present invention can accurately customize personalized teaching interaction scenarios for students; in S11, the virtual teacher designs a one-to-one teaching mode, a classroom discussion mode and a question-and-answer mode according to the students' age, knowledge level and course characteristics, thereby improving students' learning participation and adaptability; at the same time, students' learning needs-related data are collected through online questionnaires, learning platform records and teacher feedback channels to ensure the accuracy of teaching content; S12 divides several groups of teaching units according to the syllabus and student needs, covering grammar, reading and mathematical formulas, and classifies them according to course difficulty, content logical order and student needs differences, so that the teaching content is structured and targeted; S13 marks all teaching units with serial numbers, effectively manages and schedules different modules, and ensures the orderliness and controllability of the teaching process; finally, through this step, the teaching interaction scenario can be effectively optimized, making the virtual teacher's teaching more personalized and flexible, greatly improving the students' learning effect and the intelligent interaction performance of the system. Example

[0025] S21, extracting the learning state data set, and after dimensionless processing of the relevant data in the learning state data set, the learning state index Xlz is calculated by the following formula: ; Where Kwc represents the course completion value in the learning status data set, Zzw represents the knowledge point mastery value in the learning status data set, Xxt represents the learning time in the learning status data set, Tws represents the number of questions asked in the learning status data set, Hcy represents the interactive participation in the learning status data set, and Zzd represents the concentration in the learning status data set.

[0026] S22. Compare and evaluate the preset learning goal Lmz with the student’s current learning status index Xlz to identify students who need personalized adjustments. The specific evaluation content is as follows: When the learning status index Xlz≤the set learning goal Lmz, the corresponding students are identified as students who need personalized adjustment and marked; When the learning status index Xlz>learning target Lmz, it means that the corresponding students do not need personalized adjustments and are not marked.

[0027] In this embodiment, through the implementation of S21 and S22, the present invention can accurately calculate and evaluate the learning status of students, and then determine the students who need personalized adjustment; first, S21 extracts the learning status data set and merges it for dimensionless processing, integrating multiple key parameters such as course completion value Kwc, knowledge point mastery value Zzw, learning time Xxt, number of questions Tws, interactive participation Hcy and concentration Zzd, comprehensively reflecting the students' learning progress, knowledge mastery, learning initiative and concentration, thereby forming a comprehensive learning status index Xlz; S22 compares the preset learning goal Lmz with the student's learning status index Xlz to evaluate whether the student has achieved the expected learning goal. When the learning status index Xlz is lower than the learning goal Lmz, the system can automatically identify and mark the students who need personalized adjustment; this process effectively improves the accuracy of personalized teaching, ensuring that the virtual teacher can adjust the teaching content in a targeted manner, thereby improving students' learning effect and participation. Example

[0028] S3 includes S31, S32, S33 and S34, specifically: S31, collecting the learning habit data and emotional feedback related data of the students marked with personalized adjustment in S2, and constructing a habit and emotional data set after preprocessing and dimensionless processing; wherein the habit and emotional data set includes the students' emotional change index Ex and the learning task difficulty value Nx; then marking the students marked with personalized adjustment with serial numbers, including the students with serial number i; S32. By extracting the habit and emotion data set, the comprehensive state value Fhxs of the student with sequence number i is calculated by the following formula: ; Where Lrx represents the learning endurance index in the habit and emotion data set, and Cox represents the cognitive load index in the habit and emotion data set.

[0029] S33, by presetting the first comprehensive state threshold W1 and the second comprehensive state threshold W2, the comprehensive state value Fhxs of the student with the serial number i is evaluated, and the first comprehensive state threshold W1 is greater than the second comprehensive state threshold W2. The specific evaluation content is as follows: When the student comprehensive state value Fhxs ≥ the first comprehensive state threshold W1, the first teaching mode is generated, and the virtual teacher gives priority to adjusting the teaching content and provides personalized content; When the first comprehensive state threshold W1>comprehensive state value Fhxs≥second comprehensive state threshold W2, the second teaching mode is generated at this time, and the virtual teacher continues to execute according to the current standard teaching plan; When the second comprehensive state threshold W2>comprehensive state value Fhxs, the third teaching mode is generated, and the virtual teacher gives priority to emotional support and then makes teaching adjustments.

[0030] S34. After dynamically adjusting the teaching mode of the student with sequence number i, real-time monitoring and collection of the learning rhythm and state-related data of the student with sequence number i, including the knowledge mastery progress value Kcz, the learning resistance index Zlx and the learning interest value Xqx; after dimensionless processing of the learning rhythm and state-related data of the student with sequence number i, the learning demand assessment value Lhx is calculated by the following formula: ; In the formula, a1, a2 and a3 are adjustment coefficients, and their specific values ​​are adjusted by the user; a1 is used to adjust the influence of the knowledge mastering progress value Kcz, a2 is used to adjust the influence of the learning interest value Xqx on the formula, and a3 is used to adjust the influence of the learning resistance index Zlx; The preset learning demand threshold E is compared with the learning demand assessment value Lhx to make an assessment and adjust the teaching content. The specific assessment content is as follows: If the learning need assessment value Lhx ≥ the learning need threshold E, it means that the learning need of the student with sequence number i is normal. At this time, the difficulty of the teaching content will be increased by 5% based on the original teaching content; If the learning needs assessment value Lhx is less than the learning needs threshold E, it means that the learning needs of the student with sequence number i are abnormal. At this time, the original teaching content should be maintained and the students should focus on strengthening the consolidation of basic knowledge. In this embodiment, through the implementation of S3, the present invention can accurately evaluate and dynamically adjust the teaching mode of students, and improve the personalized teaching effect; in S31, the learning habit data and emotional feedback data of students, such as the emotional change index Ex and the learning task difficulty value Nx, are collected, and the comprehensive state value Fhxs is calculated in combination with the learning endurance index Lrx and the cognitive load index Cox. The collection of these parameters reflects the learning tolerance and cognitive pressure of students in different tasks, ensuring that the adjustment of the teaching mode is based on the actual learning state of the students; S32 realizes a comprehensive evaluation of the current state of the students through the calculation of the comprehensive state value Fhxs, and generates the first teaching mode, the second teaching mode and the third teaching mode through the preset first comprehensive state threshold W1 and the second comprehensive state threshold W2, so that the virtual teacher can flexibly adjust the teaching content or emotional support; S34 collects the knowledge mastery progress value Kcz, the learning resistance index Zlx and the learning interest value Xqx, and calculates the learning demand evaluation value Lhx through these parameters, so as to further adjust the teaching content. The knowledge mastery progress value Kcz represents the student's mastery of the current knowledge, and the learning resistance index Zlx represents the student's mastery of the current knowledge. The learning interest value Xqx reflects the resistance encountered by students in learning, and reflects the students' interest in the learning content. The collection of all these data ensures the pertinence and effectiveness of teaching adjustments, and ultimately enables the system to flexibly adjust the teaching difficulty or provide more support according to the students' learning needs, greatly improving the students' learning effects. Example

[0031] S41. Through the online learning platform, based on the emotion recognition algorithm and behavior tracking equipment, the learning progress rate Xjz, emotion change value Xqg and concentration duration Xzz of the student with sequence number i are collected in real time and dimensionless processed, and the learning state index Xlzc is calculated by the following formula: ; S42, compare and evaluate the secondary learning state index Xlzc with the learning state data Xlz, and judge again whether the virtual teacher needs to adjust the teaching content. The specific contents are as follows: If the secondary learning state index Xlzc ≥ learning state data Xlz, it means that the student's learning state has improved under the adjusted teaching mode and has adapted to the new teaching content. At this time, the existing teaching mode and emotional support should be maintained; If the learning state index Xlzc is less than the learning state data Xlz, it means that the student's learning state has declined under the adjusted teaching mode and is not adapted to the new teaching content. At this time, the teaching mode should be adjusted, including simplifying knowledge points and increasing emotional support.

[0032] In this embodiment, through the implementation of S41 and S42, the present invention can effectively monitor and adjust the learning status of students in real time, and improve the accuracy and flexibility of personalized teaching; in S41, based on the emotion recognition algorithm and behavior tracking equipment, the learning progress rate Xjz, the emotion change value Xqg and the concentration duration Xzz are collected, the learning progress rate Xjz reflects the progress of the students in completing the learning tasks, the emotion change value Xqg represents the changing trend of the students' emotions during the learning process, and the concentration duration Xzz reflects the duration of the students' concentration. These lower-level parameters can comprehensively evaluate the changes in the students' state in learning, and ensure that the secondary learning state index Xlzc accurately reflects the students' performance under the new teaching mode; S42 compares the sub-learning state index Xlzc with the learning state index Xlz, which can effectively determine whether students adapt to the adjusted teaching mode and ensure that students can gradually restore their learning effects under the new teaching mode; through this process, the system can continuously and dynamically optimize the teaching content to ensure that students get the best learning experience and support. Example

[0033] S5. Based on the adjusted teaching model, the virtual teacher continuously monitors the students' learning feedback data and the virtual teacher's interaction strategy, including students' learning progress, classroom emotional state and learning concentration; secondly, the virtual teacher analyzes students' emotional fluctuations in real time, identifies the changing trend of emotional state, and based on the results of emotion recognition, provides students with emotional support in a timely manner, including positive feedback, encouraging reminders and interactive measures to reduce learning pressure.

[0034] In this embodiment, through the implementation of S5, the present invention can further enhance the intelligent interaction and emotional support capabilities of the virtual teacher; based on the adjusted teaching mode, the virtual teacher continuously monitors the students' learning feedback data, including learning progress Xjz, classroom emotional state Xqg and learning concentration Xzz, to ensure that the teaching content is closely matched with the students' current state; at the same time, the virtual teacher analyzes the students' emotional fluctuations, identifies the changing trend of the emotional state, and provides emotional support in a timely manner based on the emotion recognition results. The emotional support forms include positive feedback, encouraging prompts and interactive measures to reduce learning pressure. These functions can effectively alleviate students' emotional pressure in learning and improve their learning enthusiasm and concentration; finally, through continuous monitoring and emotional feedback adjustment, the virtual teacher can dynamically optimize the teaching strategy to ensure that students maintain a positive emotional state and efficient learning results during the learning process. Example

[0035] See also Figure 2, an intelligent digital human control system, including an interactive scene setting module, a learning status monitoring module, a learning habit and emotion analysis module, a teaching content dynamic adjustment module and a learning feedback and emotion support module; The interactive scene setting module is used to pre-set the interactive scenes of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; The learning status monitoring module is used to monitor students' learning progress related data and interactive feedback related data in real time to generate a learning status data set; based on the learning status data set, students who need personalized adjustment are determined from a number of students, and students who need personalized adjustment are marked; The learning habit and emotion analysis module is used to analyze the learning habit data and emotion feedback related data of the students according to the marked students, construct the habit and emotion data set, and generate and evaluate the comprehensive state value Fhxs of the student with sequence number i according to the habit and emotion data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; The teaching content dynamic adjustment module is used to calculate and evaluate the secondary learning demand evaluation value Lhxc again after re-adjusting the teaching content, and finally determine whether the teaching mode of the students needs to be further adjusted according to the evaluation content of the secondary learning demand evaluation value Lhxc, including increasing emotional support; The learning feedback and emotional support module is used to continue monitoring students' learning feedback and the virtual teacher's interaction strategy according to the adjusted teaching model; enabling the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

[0036] In this embodiment, through the synergy of each module, a significant teaching optimization effect is achieved; the interactive scene setting module can flexibly set the interactive scene of the virtual teacher according to the students' learning needs and course content, and divide the teaching units to ensure the accurate labeling and execution of personalized response strategies; the learning status monitoring module effectively generates a learning status data set by real-time monitoring of students' learning progress related data and interactive feedback data, and accurately identifies students who need personalized adjustment; the learning habit and emotion analysis module generates a comprehensive state value Fhxs by analyzing students' learning habit data and emotional feedback related data, and makes the teaching mode more personalized and dynamically adjusted by evaluating and adjusting the teaching mode and calculating the learning needs assessment value Lhx; the teaching content dynamic adjustment module determines whether the teaching mode needs to be further adjusted by comparing the learning needs assessment value Lhxc with the learning state index Xlz, so as to ensure that students obtain appropriate emotional support and teaching optimization in the learning process; the learning feedback and emotional support module continuously monitors students' emotional fluctuations through emotion recognition technology, and provides students with timely emotional support and learning rhythm adjustment to enhance learning effects and enthusiasm.

[0037] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for controlling an intelligent digital human, characterized in that: The following steps are involved: S1. Pre-set the interaction scenarios of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; S2. Real-time monitoring of students' learning progress-related data and interactive feedback-related data to generate a learning status data set; based on the learning status data set, determining students who need personalized adjustments from a number of students, and marking the students who need personalized adjustments; S3. According to the marked students obtained in step S2, the learning habit data and emotional feedback related data of the students are analyzed, a habit and emotional data set is constructed, and a comprehensive state value Fhxs of the student with sequence number i is generated and evaluated according to the habit and emotional data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; S4. After re-adjusting the teaching content, the secondary learning need assessment value Lhxc is calculated and evaluated again, and finally, according to the evaluation content of the secondary learning need assessment value Lhxc, it is determined whether the teaching mode of the students needs to be further adjusted, including increasing emotional support; S5. According to the adjusted teaching model, continue to monitor students' learning feedback and the virtual teacher's interaction strategy; enable the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

2. The intelligent digital human control method according to claim 1, characterized in that: S1 includes S11, S12 and S13, specifically: S11. First, design the interaction scenarios between virtual teachers and students according to the teaching content, learning objectives and interaction needs; the interaction scenarios include one-on-one teaching mode, classroom discussion mode and question-answering mode; and each scenario is customized according to the age, knowledge level and course characteristics of the students, and the performance of the virtual teacher can adapt to the teaching needs in different scenarios; then obtain the relevant data of students' learning needs through historical data, and the acquisition of relevant data of students' learning needs is carried out through online questionnaires, learning platform records and teacher feedback; S12. Divide the course into several groups of teaching units according to the syllabus and student needs. Each teaching unit represents a specific knowledge point or course unit, including grammar, reading and mathematical formulas. The teaching units are divided according to the difficulty of the course, the logical order of the content and the differences in student needs. Modules are locked according to the current learning progress and needs of students. S13. Label all teaching units with serial numbers, specifically JC1, JC2...JCn, where JC1, JC2...JCn represent the first teaching unit, the second teaching unit...the nth teaching unit respectively, and n is the total number of teaching units.

3. The intelligent digital human control method according to claim 1, characterized in that: S2 includes S21 and S22, specifically: S21, extracting the learning state data set, and after dimensionless processing of the relevant data in the learning state data set, the learning state index Xlz is calculated by the following formula: ； Where Kwc represents the course completion value in the learning status data set, Zzw represents the knowledge point mastery value in the learning status data set, Xxt represents the learning time in the learning status data set, Tws represents the number of questions asked in the learning status data set, Hcy represents the interactive participation in the learning status data set, and Zzd represents the concentration in the learning status data set.

4. The intelligent digital human control method according to claim 3, characterized in that: S22. Compare and evaluate the preset learning goal Lmz with the student’s current learning status index Xlz to identify students who need personalized adjustments. The specific evaluation content is as follows: When the learning status index Xlz≤the set learning goal Lmz, the corresponding students are identified as students who need personalized adjustment and marked; When the learning status index Xlz>learning target Lmz, it means that the corresponding students do not need personalized adjustments and are not marked.

5. The intelligent digital human control method according to claim 4, characterized in that: S3 includes S31, S32, S33 and S34, specifically: S31, collecting the learning habit data and emotional feedback related data of the students marked with personalized adjustment in S2, and constructing a habit and emotional data set after preprocessing and dimensionless processing; wherein the habit and emotional data set includes the students' emotional change index Ex and the learning task difficulty value Nx; then marking the students marked with personalized adjustment with serial numbers, including the students with serial number i; S32. By extracting the habit and emotion data set, the comprehensive state value Fhxs of the student with sequence number i is calculated by the following formula: ； Where Lrx represents the learning endurance index in the habit and emotion data set, and Cox represents the cognitive load index in the habit and emotion data set.

6. The intelligent digital human control method according to claim 5, characterized in that: S33, by presetting the first comprehensive state threshold W1 and the second comprehensive state threshold W2, the comprehensive state value Fhxs of the student with the serial number i is evaluated, and the first comprehensive state threshold W1 is greater than the second comprehensive state threshold W2. The specific evaluation content is as follows: When the student comprehensive state value Fhxs ≥ the first comprehensive state threshold W1, the first teaching mode is generated, and the virtual teacher gives priority to adjusting the teaching content and provides personalized content; When the first comprehensive state threshold W1>comprehensive state value Fhxs≥second comprehensive state threshold W2, the second teaching mode is generated at this time, and the virtual teacher continues to execute according to the current standard teaching plan; When the second comprehensive state threshold W2>comprehensive state value Fhxs, the third teaching mode is generated, and the virtual teacher gives priority to emotional support and then makes teaching adjustments.

7. The intelligent digital human control method according to claim 6, characterized in that: S34. After dynamically adjusting the teaching mode of the student with sequence number i, real-time monitoring and collection of the learning rhythm and state-related data of the student with sequence number i, including the knowledge mastery progress value Kcz, the learning resistance index Zlx and the learning interest value Xqx; after dimensionless processing of the learning rhythm and state-related data of the student with sequence number i, the learning demand assessment value Lhx is calculated by the following formula: ； In the formula, a1, a2 and a3 are adjustment coefficients, and their specific values ​​are adjusted by the user; a1 is used to adjust the influence of the knowledge mastering progress value Kcz, a2 is used to adjust the influence of the learning interest value Xqx on the formula, and a3 is used to adjust the influence of the learning resistance index Zlx; The preset learning demand threshold E is compared with the learning demand assessment value Lhx to make an assessment and adjust the teaching content. The specific assessment content is as follows: If the learning need assessment value Lhx ≥ the learning need threshold E, it means that the learning need of the student with sequence number i is normal. At this time, the difficulty of the teaching content will be increased by 5% based on the original teaching content; If the learning needs assessment value Lhx is less than the learning needs threshold E, it means that the learning needs of the student with serial number i are abnormal. At this time, the original teaching content should be maintained and attention should be focused on strengthening the consolidation of basic knowledge.

8. The intelligent digital human control method according to claim 3, characterized in that: S4 includes S41 and S42, specifically: S41. Through the online learning platform, based on the emotion recognition algorithm and behavior tracking equipment, the learning progress rate Xjz, emotion change value Xqg and concentration duration Xzz of the student with sequence number i are collected in real time and dimensionless processed, and the learning state index Xlzc is calculated by the following formula: ； S42, compare and evaluate the secondary learning state index Xlzc with the learning state index Xlz, and judge again whether the virtual teacher needs to adjust the teaching content. The specific contents are as follows: If the secondary learning state index Xlzc ≥ learning state data Xlz, it means that the student's learning state has improved under the adjusted teaching mode and has adapted to the new teaching content. At this time, the existing teaching mode and emotional support should be maintained; If the learning state index Xlzc is less than the learning state data Xlz, it means that the student's learning state has declined under the adjusted teaching mode and is not adapted to the new teaching content. At this time, the teaching mode should be adjusted, including simplifying knowledge points and increasing emotional support.

9. The intelligent digital human control method according to claim 1, characterized in that: S5. Based on the adjusted teaching model, the virtual teacher continuously monitors the students' learning feedback data and the virtual teacher's interaction strategy, including students' learning progress, classroom emotional state and learning concentration; secondly, the virtual teacher analyzes students' emotional fluctuations in real time, identifies the changing trend of emotional state, and based on the results of emotion recognition, provides students with emotional support in a timely manner, including positive feedback, encouraging reminders and interactive measures to reduce learning pressure.

10. An intelligent digital human control system, applied to an intelligent digital human control method according to any one of claims 1 to 9, characterized in that: It includes interactive scene setting module, learning status monitoring module, learning habit and emotion analysis module, teaching content dynamic adjustment module and learning feedback and emotional support module; The interactive scene setting module is used to pre-set the interactive scenes of the virtual teacher, obtain the data related to the students' learning needs, divide the teaching units into several groups and lock them, and mark the individual response strategies of different teaching units and virtual teachers with serial numbers; The learning status monitoring module is used to monitor students' learning progress related data and interactive feedback related data in real time to generate a learning status data set; based on the learning status data set, students who need personalized adjustment are determined from a number of students, and students who need personalized adjustment are marked; The learning habit and emotion analysis module is used to analyze the learning habit data and emotion feedback related data of the students according to the marked students, construct the habit and emotion data set, and generate and evaluate the comprehensive state value Fhxs of the student with sequence number i according to the habit and emotion data set. Based on the evaluation result of the comprehensive state value Fhxs, the teaching mode of the student with sequence number i is dynamically adjusted, and then the learning need evaluation value Lhx of the student with sequence number i is calculated and evaluated again. The virtual teacher adjusts the teaching content according to the evaluation content of the learning need evaluation value Lhx; The teaching content dynamic adjustment module is used to calculate and evaluate the secondary learning demand evaluation value Lhxc again after re-adjusting the teaching content, and finally determine whether the teaching mode of the students needs to be further adjusted according to the evaluation content of the secondary learning demand evaluation value Lhxc, including increasing emotional support; The learning feedback and emotional support module is used to continue monitoring students' learning feedback and the virtual teacher's interaction strategy according to the adjusted teaching model; enabling the virtual teacher to provide emotional support in a timely manner through emotion recognition and adjust the learning rhythm and content.

Images