Humanoid robot control method and system based on AI intelligent teaching

Abstract

The invention relates to the technical field of robot control, in particular to a humanoid robot control method and system based on AI intelligent teaching. A humanoid robot control method based on AI intelligent teaching comprises the following steps: S1, obtaining scheduling related data in a classroom, and carrying out data preprocessing on the scheduling related data to obtain preprocessed scheduling related data; s2, acquiring a personnel spacing value according to the preprocessed scheduling related data, and performing robot path planning and control according to the personnel spacing value; and S3, when the personnel spacing value is greater than or equal to a first preset distance, performing first path planning and control of the robot. By setting a three-stage switching mechanism, the robot can dynamically adjust the service strategy according to the real-time personnel spacing and the help demand degree, so that the autonomous thinking space of students is guaranteed, and effective assistance can be provided for auditing personnel and teachers at a proper time.

Description

Technical Field

[0001] The present invention relates to the field of robot control technology, and in particular to a humanoid robot control method and system based on AI intelligent teaching. Background Art

[0002] With the rapid development of artificial intelligence and robotics, the application prospects of intelligent teaching robots in the education field have attracted much attention. Existing teaching robots mostly focus on pre-set routes or single-scenario explanations. They lack the ability to dynamically adjust their paths based on the real-time classroom environment, making it difficult to meet the needs of multi-role, multi-demand teaching scenarios. Furthermore, existing technologies often focus only on the students' own needs for assistance or maintain a fixed distance from the teacher. They are unable to flexibly address the comprehensive needs of students, auditors, and teachers in the classroom, nor can they effectively plan and control based on multiple environmental factors such as real-time personnel distribution, physical obstacles, and crowd density. Summary of the Invention

[0003] In order to overcome the shortcoming of being unable to flexibly respond to the comprehensive needs of students, auditors and teachers in the classroom, the present invention provides a humanoid robot control method and system based on AI intelligent teaching.

[0004] The technical implementation scheme of the present invention is: a humanoid robot control method based on AI intelligent teaching, comprising the following steps:

[0005] S1: Acquire scheduling-related data in the classroom, and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data;

[0006] S2: Obtaining a personnel spacing value based on the pre-processed scheduling-related data, and performing path planning and control of the robot based on the personnel spacing value;

[0007] S3: When the personnel distance value is greater than or equal to a first preset distance, performing first path planning and control of the robot;

[0008] S4: When the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performing second path planning and control of the robot;

[0009] S5: When the personnel distance value is less than the second preset distance, the third path planning and control of the robot is performed.

[0010] Preferably, the obtaining of scheduling-related data in the classroom and performing data preprocessing on the scheduling-related data to obtain the preprocessed scheduling-related data include: obtaining scheduling-related data in the classroom, the scheduling-related data including personnel type data and personnel location data, performing normalization and data cleaning preprocessing on the scheduling-related data to obtain the preprocessed scheduling-related data, the personnel type data including teachers, students and auditors; the personnel location data including teachers' location data and auditors' location data.

[0011] Preferably, the personnel spacing value is obtained based on the pre-processed scheduling-related data, and the robot's path planning and control are performed based on the personnel spacing value, including: obtaining the personnel spacing value based on the teacher's position data and the auditor's position data, when the personnel spacing value is greater than or equal to a first preset distance, performing the robot's first path planning and control; when the personnel spacing value is less than the first preset distance and greater than or equal to a second preset distance, performing the robot's second path planning and control; when the personnel spacing value is less than the second preset distance, performing the robot's third path planning and control, wherein the first preset distance is greater than the second preset distance.

[0012] Preferably, when the interpersonal distance value is greater than or equal to a first preset distance, performing first path planning and control of the robot includes: when the interpersonal distance value is greater than or equal to the first preset distance, in the independent evaluation stage, obtaining the help needs of students in the classroom, taking students whose help needs are greater than the first preset need threshold as nodes to be processed, obtaining first adjustment-related data for the nodes to be processed, the first adjustment-related data including the closest distance between the robot and the teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed, and the crowd density between each node to be processed; based on the help needs of the students and the first adjustment-related data, using a first edge weight calculation formula to obtain first edge weights between each node to be processed; using a Dijkstra algorithm based on the first edge weights to obtain a first path plan for the robot; controlling the robot according to the first path plan; and simultaneously detecting the duration and number of eye contact interactions of the auditor in the independent evaluation stage. When the duration of the auditor in the independent evaluation stage is greater than the first preset duration and the number of eye contact interactions is less than a preset number, automatically generating a phased teaching briefing, the number of eye contact interactions being the number of eye contact interactions between the auditor and the teacher.

[0013] Preferably, the first edge weight calculation formula is used to obtain the first edge weight between each node to be processed according to the student's help need and the first adjustment-related data, including: wherein the first edge weight calculation formula is:

[0014]

[0015] Where, is the first edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; α1, α2, α3, α4, α5 are the weight adjustment coefficients of the first edge weight calculation formula; ε1 and ε2 are adjustment parameters; It is the first auxiliary adjustment value.

[0016] Preferably, when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, the robot is subjected to second path planning and control, including: when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, it is in the observation and feedback stage, obtaining the help needs of students and the help needs of auditors in the classroom, taking students whose help needs are greater than the first preset need threshold and auditors whose help needs are greater than the second preset need threshold as nodes to be processed, obtaining second adjustment-related data of the nodes to be processed, the second adjustment-related data including the closest distance between the robot and the teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed and the crowd density between each node to be processed, according to the help needs of the students, the help needs of the auditors and the second adjustment-related data, using a second edge weight calculation formula to obtain the second edge weight between each node to be processed, and using the Dijkstra algorithm according to the second edge weight to obtain the second path planning of the robot, controlling the robot according to the second path planning, and at the same time, the robot pushes teaching content to the auditors.

[0017] Preferably, the second edge weight calculation formula is used to obtain the second edge weight between each node to be processed based on the student's help need, the auditor's help need, and the second adjustment-related data, including: wherein the second edge weight calculation formula is:

[0018]

[0019] Where, is the second edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the help demand of the observer at node i; is the help demand of the observer at node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; β1, β2, β3, β4, β5 are the weight adjustment coefficients of the second edge weight calculation formula; δ1, δ2, θ are adjustment parameters; is the second auxiliary adjustment value.

[0020] Preferably, when the personnel distance value is less than the second preset distance, the robot's third path planning and control are performed, including: when the personnel distance value is less than the second preset distance, it is in the close collaboration stage, obtaining the teacher's help need, the student's help need and the auditor's help need, and taking the students whose help need is greater than the first preset need threshold, the auditor's help need is greater than the second preset need threshold, and the teacher's help need is greater than the third preset need threshold as the nodes to be processed, obtaining the third adjustment-related data of the nodes to be processed, the third adjustment-related data including the physical distance between each node to be processed, the static obstacle complexity between each node to be processed and the crowd density between each node to be processed, according to the teacher's help need, the student's help need, the auditor's help need and the third adjustment-related data, using the third edge weight calculation formula to obtain the third edge weight between each node to be processed, and using the Dijkstra algorithm according to the third edge weight to obtain the robot's third path planning, and controlling the robot according to the third path planning.

[0021] Preferably, the third edge weight calculation formula is used to obtain the third edge weight between each node to be processed based on the teacher's help need, the student's help need, the auditor's help need and the third adjustment related data, including: wherein the third edge weight calculation formula is:

[0022]

[0023] Where, is the third edge weight; is the help demand of the teacher at node i; is the help demand of the teacher at node j; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j;

[0024] γ1, γ2, γ3, and γ4 are weight adjustment coefficients of the third edge weight calculation formula; σ, ρ1, ρ2, and ρ3 are adjustment parameters; It is the third auxiliary adjustment value.

[0025] Preferably, a humanoid robot control system based on AI intelligent teaching further includes:

[0026] A data acquisition and processing module is used to acquire scheduling-related data in the classroom and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data;

[0027] A robot path planning module is used to obtain the personnel spacing value based on the pre-processed scheduling-related data, and to plan and control the robot's path based on the personnel spacing value;

[0028] A first edge weight calculation module, configured to obtain a first edge weight between each to-be-processed node using a first edge weight calculation formula according to the student's help need and the first adjustment-related data;

[0029] A first path planning submodule, configured to plan and control a first path of the robot when the personnel distance value is greater than or equal to a first preset distance;

[0030] A second edge weight calculation module is configured to obtain the second edge weight between each to-be-processed node using a second edge weight calculation formula according to the student's help need, the auditor's help need, and the second adjustment-related data;

[0031] A second path planning submodule, when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performs a second path planning and control of the robot;

[0032] A third edge weight calculation module is configured to obtain the third edge weights between the nodes to be processed using a third edge weight calculation formula based on the teacher's help need, the student's help need, the auditor's help need, and the third adjustment-related data;

[0033] The third path planning submodule performs third path planning and control of the robot when the personnel distance value is less than the second preset distance.

[0034] The present invention has the following advantages:

[0035] 1. By setting up a switching mechanism among independent evaluation, observation and feedback, and close collaboration stages, the robot can dynamically adjust its service strategy based on the real-time distance between people and the degree of help needed. This not only ensures students' independent thinking space, but also provides effective assistance to auditors and teachers at appropriate times, meeting the teaching needs of multiple levels and multiple roles.

[0036] 2. Utilizing a multi-dimensional weighting model based on the help needs of students, auditors, and teachers, as well as physical distance, static obstacle complexity, and crowd density, combined with the Dijkstra shortest path algorithm, we achieve refined and intelligent path planning, preventing robots from colliding or interfering in complex classroom environments and ensuring safe and smooth teaching.

[0037] 3. In the first stage, the nodes to be processed are identified based on the students' need for help. In the second stage, the needs of auditors are considered and teaching content is pushed, so that the robot can provide personalized assistance to different objects in a targeted manner, effectively improving the learning efficiency and satisfaction of students and auditors. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a flow chart of a humanoid robot control method based on AI intelligent teaching in the present invention;

[0039] Figure 2 This is a schematic diagram of the structure of a humanoid robot control system based on AI intelligent teaching in the present invention. DETAILED DESCRIPTION

[0040] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. 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 making creative efforts are within the scope of protection of the present invention.

[0041] Example 1: A humanoid robot control method based on AI intelligent teaching, such as Figure 1 As shown, the following steps are included:

[0042] S1: Acquire scheduling-related data in the classroom, and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data;

[0043] Obtain scheduling-related data in the classroom, the scheduling-related data including personnel type data and personnel location data, perform normalization and data cleaning preprocessing on the scheduling-related data, and obtain the preprocessed scheduling-related data, the personnel type data including teacher personnel, student personnel and auditor personnel; the personnel location data including teacher personnel location data and auditor personnel location data.

[0044] It needs to be further explained that the following two types of raw data are collected synchronously through the camera and positioning sensor arranged on the top of the classroom, as well as the RFID tag transceiver held by the teacher: personnel type data: identifying and marking the three identities of all teachers, students and auditors currently in the classroom; personnel position data: recording the real-time position of each teacher and each auditor in the classroom plane coordinate system, in centimeters, and mapping the collected type identification and position coordinates of teachers, students and auditors to a unified data format. The packet loss, misidentification or jumping noise points in the continuous frame images or positioning data are cleaned up, and the coordinate sequence of each participant is smoothed using median filtering; the positioning data that is missing more than the threshold is supplemented by linear interpolation or forward filling; after deleting the false detection data points whose residence time is less than the preset threshold, the pre-processed scheduling related data is finally obtained.

[0045] S2: Obtaining a personnel spacing value based on the pre-processed scheduling-related data, and performing path planning and control of the robot based on the personnel spacing value;

[0046] The personnel distance value is obtained based on the position data of the teacher and the position data of the auditor. When the personnel distance value is greater than or equal to the first preset distance, the robot's first path planning and control are performed; when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, the robot's second path planning and control are performed; when the personnel distance value is less than the second preset distance, the robot's third path planning and control are performed, wherein the first preset distance is greater than the second preset distance.

[0047] It needs to be further explained that all teachers and auditors are paired up, their Euclidean distance on the normalized plane is calculated, and the personnel distance value is obtained. When the personnel distance value is greater than or equal to the first preset distance, the robot's first path planning and control are performed; when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, the robot's second path planning and control are performed; when the personnel distance value is less than the second preset distance, the robot's third path planning and control are performed, where the first preset distance is greater than the second preset distance.

[0048] S3: When the personnel distance value is greater than or equal to a first preset distance, performing first path planning and control of the robot;

[0049] When the interpersonal distance value is greater than or equal to a first preset distance, the robot is in the independent evaluation stage. The help needs of students in the classroom are obtained, and students whose help needs are greater than the first preset need threshold are selected as nodes to be processed. First adjustment-related data of the nodes to be processed are obtained, and the first adjustment-related data include the closest distance between the robot and the teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed, and the crowd density between each node to be processed. Based on the help needs of the students and the first adjustment-related data, a first edge weight calculation formula is used to obtain the first edge weight between each node to be processed. The Dijkstra algorithm is used to obtain the first path planning of the robot based on the first edge weight. The robot is controlled according to the first path planning, and the duration and number of eye contact interactions of the auditor in the independent evaluation stage are detected. When the duration of the auditor in the independent evaluation stage is greater than the first preset duration and the number of eye contact interactions is less than a preset number, a stage-by-stage teaching briefing is automatically generated. The number of eye contact interactions is the number of eye contact interactions between the auditor and the teacher.

[0050] What needs to be further explained is that 1. Real-time behavior monitoring method: facial expression recognition, through the high-definition camera on the top of the classroom and the emotion analysis model, the students' facial expressions (such as frowning, opening the mouth and pursing the lips) are classified and mapped to the predefined emotional states of "confusion", "doubt" and "contemplation", and the confidence of several emotional states is synthesized by weight to form a preliminary help demand score; gesture / hand-raising detection: using the skeleton key point detection algorithm to determine whether the student raises his hand or makes a help gesture, and accumulates it into a discrete help request signal, and each signal is recorded as a help demand unit; 2. Interactive system feedback method: intelligent answering terminal, if the student is equipped with a tablet terminal or a classroom answerer, the number of consecutive errors, average answering time and number of times of giving up are counted, and mapped to the help demand score; teacher marking: the teacher marks the "high demand" of the students he is tutoring on his own teaching control panel "Medium demand" and "low demand" are manually labeled and converted into corresponding numerical values in real time; 3. Historical data prediction method, learning trajectory model: based on the students' classroom performance and homework grade fluctuations in the past period of time, the time series prediction model is trained to generate a predicted help demand value for the current moment; finally, the multi-dimensional signals from the above sources are normalized and weighted fused to obtain the students' help demand; the closest distance between the robot and the teacher or auditor is obtained by calculating the Euclidean distance; the physical distance between each node to be processed is calculated based on the normalized student coordinates to calculate the planar distance between the nodes; the static obstacle complexity between each node to be processed is calculated by using the classroom map and the obstacle detection deep learning model to count the number of obstacles and the complexity of their shapes; the crowd density between each node to be processed is calculated by counting the number of people around the nodes i and j with a radius of r meters in the area, and calculating the density value.

[0051] The calculation formula for the first edge weight is:

[0052]

[0053] Where, is the first edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; α1, α2, α3, α4, α5 are the weight adjustment coefficients of the first edge weight calculation formula; ε1 and ε2 are adjustment parameters; It is the first auxiliary adjustment value.

[0054] It needs to be further explained that the first adjustment-related data is the data when the student is the node to be processed; ε1 and ε2 take small constants to ensure that the denominator is not zero and does not significantly affect the weight; after normalizing the student's help demand and the first adjustment-related data, the first edge weight calculation formula is used to obtain the first edge weight between each node to be processed; the first edge weight (independent evaluation stage): with students as the subject, keep the robot's control route away from teachers and auditors.

[0055] S4: When the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performing second path planning and control of the robot;

[0056] When the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, it is in the observation and feedback stage, and the help needs of students and auditors in the classroom are obtained. Students whose help needs are greater than the first preset need threshold and auditors whose help needs are greater than the second preset need threshold are taken as nodes to be processed, and second adjustment-related data of the nodes to be processed are obtained. The second adjustment-related data include the closest distance between the robot and the teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed, and the crowd density between each node to be processed. According to the help needs of the students, the help needs of the auditors and the second adjustment-related data, the second edge weight calculation formula is used to obtain the second edge weight between each node to be processed, and the Dijkstra algorithm is used according to the second edge weight to obtain the second path planning of the robot. The robot is controlled according to the second path planning, and the robot pushes teaching content to the auditors.

[0057] The second edge weight calculation formula is:

[0058]

[0059] Where, is the second edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the help demand of the observer at node i; is the help demand of the observer at node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; β1, β2, β3, β4, β5 are the weight adjustment coefficients of the second edge weight calculation formula; δ1, δ2, θ are adjustment parameters; is the second auxiliary adjustment value.

[0060] What needs to be further explained is that the help needs of the auditors are obtained after adjustment through the above-mentioned real-time behavior monitoring method, interactive system feedback method and historical data prediction method; and the students whose help needs are greater than the first preset need threshold and the auditors whose help needs are greater than the second preset need threshold are taken as nodes to be processed; similarly, the second adjustment related data with students and auditors as nodes to be processed are obtained, and after normalization, the second edge weight between each node to be processed is obtained using the second edge weight calculation formula; relationship interpretation: the auditor is in an independent observation position, which can clearly see the overall situation and intervene at any time. It is the most common evaluation or learning posture; the robot responds accordingly, that is, directional information push, and silently pushes additional information visible only to the auditor through the AR glasses, handheld Pad or classroom side screen worn by the auditor; second edge weight (observation feedback stage): considering the importance of the auditors, let the robot choose to approach teachers and auditors.

[0061] S5: When the personnel distance value is less than the second preset distance, the third path planning and control of the robot is performed.

[0062] When the personnel distance value is less than the second preset distance, it is in the close collaboration stage, and the help needs of teachers, students and auditors are obtained. Students whose help needs are greater than the first preset need threshold, auditors whose help needs are greater than the second preset need threshold, and teachers whose help needs are greater than the third preset need threshold are taken as nodes to be processed, and third adjustment-related data of the nodes to be processed are obtained. The third adjustment-related data include the physical distance between each node to be processed, the static obstacle complexity between each node to be processed and the crowd density between each node to be processed. According to the help needs of teachers, students, auditors and the third adjustment-related data, the third edge weight calculation formula is used to obtain the third edge weight between each node to be processed, and the Dijkstra algorithm is used according to the third edge weight to obtain the third path planning of the robot, and the robot is controlled according to the third path planning.

[0063] The third edge weight calculation formula is:

[0064]

[0065] Where, is the third edge weight; is the help demand of the teacher at node i; is the help demand of the teacher at node j; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j;

[0066] γ1, γ2, γ3, and γ4 are weight adjustment coefficients of the third edge weight calculation formula; σ, ρ1, ρ2, and ρ3 are adjustment parameters; It is the third auxiliary adjustment value.

[0067] It needs to be further explained that when the distance between the personnel is less than the second preset distance, it is in the close collaboration stage. The help demand of the teachers is obtained after adjustment through the above-mentioned real-time behavior monitoring method, interactive system feedback method and historical data prediction method. At the same time, the help demand of the students and the help demand of the auditors are obtained. The students whose help demand is greater than the first preset demand threshold, the auditors whose help demand is greater than the second preset demand threshold, and the teachers whose help demand is greater than the third preset demand threshold are taken as the nodes to be processed. Similarly, the third adjustment related data with students, auditors and teachers as the nodes to be processed are obtained, and after normalization, the third edge weight calculation formula is used to obtain the third edge weight between each node to be processed; the control module drives the robot to move along the third path planning result, and conducts in-depth interaction and collaborative guidance with teachers, students and auditors in turn; when all the nodes to be processed are visited or the close collaboration time reaches the preset time, exit this stage and switch back to the "observation feedback" or "independent evaluation" stage to continue dynamic adjustment; third edge weight (close collaboration stage): pay attention to the help demand of teachers and auditors.

[0068] Example 2: Based on Example 1, a humanoid robot control system based on AI intelligent teaching, such as Figure 2 As shown, it also includes:

[0069] A data acquisition and processing module is used to acquire scheduling-related data in the classroom and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data;

[0070] A robot path planning module is used to obtain the personnel spacing value based on the pre-processed scheduling-related data, and to plan and control the robot's path based on the personnel spacing value;

[0071] A first edge weight calculation module, configured to obtain a first edge weight between each to-be-processed node using a first edge weight calculation formula according to the student's help need and the first adjustment-related data;

[0072] A first path planning submodule, configured to plan and control a first path of the robot when the personnel distance value is greater than or equal to a first preset distance;

[0073] A second edge weight calculation module is configured to obtain the second edge weight between each to-be-processed node using a second edge weight calculation formula according to the student's help need, the auditor's help need, and the second adjustment-related data;

[0074] A second path planning submodule, when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performs a second path planning and control of the robot;

[0075] A third edge weight calculation module is configured to obtain the third edge weights between the nodes to be processed using a third edge weight calculation formula based on the teacher's help need, the student's help need, the auditor's help need, and the third adjustment-related data;

[0076] The third path planning submodule performs third path planning and control of the robot when the personnel distance value is less than the second preset distance.

[0077] Finally, it should be noted that the above is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art can still modify the technical solutions described in the aforementioned embodiments or make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A humanoid robot control method based on AI intelligent teaching, characterized in that: The following steps are involved: S1: Acquire scheduling-related data in the classroom, and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data; S2: Obtaining a personnel spacing value based on the pre-processed scheduling-related data, and performing path planning and control of the robot based on the personnel spacing value; S3: When the personnel distance value is greater than or equal to a first preset distance, performing first path planning and control of the robot; S4: When the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performing second path planning and control of the robot; S5: When the personnel distance value is less than the second preset distance, the third path planning and control of the robot is performed.

2. The humanoid robot control method based on AI intelligent teaching according to claim 1, characterized in that: The obtaining of scheduling-related data in the classroom and performing data preprocessing on the scheduling-related data to obtain the preprocessed scheduling-related data includes: obtaining scheduling-related data in the classroom, the scheduling-related data including personnel type data and personnel position data, performing normalization and data cleaning preprocessing on the scheduling-related data to obtain the preprocessed scheduling-related data, the personnel type data including teachers, students and auditors; the personnel position data including teachers' position data and auditors' position data.

3. The humanoid robot control method based on AI intelligent teaching according to claim 2, characterized in that: The method of obtaining the personnel spacing value based on the preprocessed scheduling-related data and performing the robot's path planning and control based on the personnel spacing value includes: obtaining the personnel spacing value based on the teacher's position data and the auditor's position data, and when the personnel spacing value is greater than or equal to a first preset distance, performing the robot's first path planning and control; when the personnel spacing value is less than the first preset distance and greater than or equal to a second preset distance, performing the robot's second path planning and control; when the personnel spacing value is less than the second preset distance, performing the robot's third path planning and control, wherein the first preset distance is greater than the second preset distance.

4. The humanoid robot control method based on AI intelligent teaching according to claim 1, characterized in that: When the interpersonal distance value is greater than or equal to a first preset distance, performing first path planning and control of the robot includes: when the interpersonal distance value is greater than or equal to the first preset distance, in an independent evaluation phase, obtaining the help needs of students in the classroom, identifying students whose help needs are greater than a first preset need threshold as nodes to be processed, obtaining first adjustment-related data for the nodes to be processed, the first adjustment-related data including the closest distance between the robot and a teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed, and the crowd density between each node to be processed; obtaining first edge weights between each node to be processed using a first edge weight calculation formula based on the student's help needs and the first adjustment-related data; obtaining a first path plan for the robot using a Dijkstra algorithm based on the first edge weights; controlling the robot according to the first path plan; and simultaneously detecting the duration and number of eye contact interactions of the auditor in the independent evaluation phase. When the duration of the auditor in the independent evaluation phase is greater than the first preset duration and the number of eye contact interactions is less than a preset number, automatically generating a phased teaching briefing, the number of eye contact interactions being the number of eye contact interactions between the auditor and the teacher.

5. The method for controlling a humanoid robot based on AI intelligent teaching according to claim 4, characterized in that: The first edge weight calculation formula is used to obtain the first edge weight between each node to be processed according to the student's help need and the first adjustment-related data, including: wherein the first edge weight calculation formula is: Where, is the first edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; α1, α2, α3, α4, α5 are the weight adjustment coefficients of the first edge weight calculation formula; ε1 and ε2 are adjustment parameters; is the first auxiliary adjustment value.

6. The method for controlling a humanoid robot based on AI intelligent teaching according to claim 1, wherein: When the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, the robot is subjected to second path planning and control, including: when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, in the observation and feedback stage, obtaining the help needs of students and the help needs of auditors in the classroom, taking students whose help needs are greater than the first preset need threshold and auditors whose help needs are greater than the second preset need threshold as nodes to be processed, obtaining second adjustment-related data of the nodes to be processed, the second adjustment-related data including the closest distance between the robot and the teacher or auditor, the physical distance between each node to be processed, the static obstacle complexity between each node to be processed, and the crowd density between each node to be processed; according to the help needs of the students, the help needs of the auditors, and the second adjustment-related data, using a second edge weight calculation formula to obtain the second edge weight between each node to be processed; using the Dijkstra algorithm according to the second edge weight to obtain the second path planning of the robot; controlling the robot according to the second path planning, and at the same time, the robot pushes teaching content to the auditors.

7. The method for controlling a humanoid robot based on AI intelligent teaching according to claim 6, characterized in that: The second edge weight calculation formula is used to obtain the second edge weight between each node to be processed based on the student's help need, the auditor's help need, and the second adjustment-related data, including: wherein the second edge weight calculation formula is: Where, is the second edge weight; is the student help demand degree of node i; is the student help demand degree of node j; is the help demand of the observer at node i; is the help demand of the observer at node j; is the distance from the robot to the nearest teacher or auditor; is the distance from the robot to the nearest teacher or auditor; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; β1, β2, β3, β4, β5 are the weight adjustment coefficients of the second edge weight calculation formula; δ1, δ2, θ are adjustment parameters; is the second auxiliary adjustment value.

8. The humanoid robot control method based on AI intelligent teaching according to claim 1, characterized in that: When the personnel distance value is less than the second preset distance, the robot's third path planning and control are performed, including: when the personnel distance value is less than the second preset distance, it is in the close collaboration stage, obtaining the help needs of teachers, students and auditors, and taking students whose help needs are greater than the first preset need threshold, the auditors whose help needs are greater than the second preset need threshold, and the teachers whose help needs are greater than the third preset need threshold as nodes to be processed, obtaining third adjustment-related data of the nodes to be processed, the third adjustment-related data including the physical distance between each node to be processed, the static obstacle complexity between each node to be processed and the crowd density between each node to be processed, according to the help needs of teachers, students, auditors and the third adjustment-related data, using the third edge weight calculation formula to obtain the third edge weight between each node to be processed, and using the Dijkstra algorithm according to the third edge weight to obtain the third path planning of the robot, and controlling the robot according to the third path planning.

9. The method for controlling a humanoid robot based on AI intelligent teaching according to claim 8, characterized in that: The third edge weight calculation formula is used to obtain the third edge weight between each node to be processed based on the teacher's help need, the student's help need, the auditor's help need and the third adjustment related data, including: wherein the third edge weight calculation formula is: Where, is the third edge weight; is the help demand of the teacher at node i; is the help demand of the teacher at node j; is the physical distance between node i and node j; is the static barrier complexity between node i and node j; is the crowd density between node i and node j; γ1, γ2, γ3, and γ4 are weight adjustment coefficients of the third edge weight calculation formula; σ, ρ1, ρ2, and ρ3 are adjustment parameters; It is the third auxiliary adjustment value.

10. A humanoid robot control system based on AI intelligent teaching, a humanoid robot control method based on AI intelligent teaching according to any one of claims 1 to 9, characterized in that: Also includes: A data acquisition and processing module is used to acquire scheduling-related data in the classroom and perform data preprocessing on the scheduling-related data to obtain preprocessed scheduling-related data; A robot path planning module is used to obtain the personnel spacing value based on the pre-processed scheduling-related data, and to plan and control the robot's path based on the personnel spacing value; A first edge weight calculation module, configured to obtain a first edge weight between each to-be-processed node using a first edge weight calculation formula according to the student's help need and the first adjustment-related data; A first path planning submodule, configured to plan and control a first path of the robot when the personnel distance value is greater than or equal to a first preset distance; The second edge weight calculation module is used to obtain the weights between each node to be processed using the second edge weight calculation formula according to the student's help need, the auditor's help need and the second adjustment related data. Second edge weight; A second path planning submodule, when the personnel distance value is less than the first preset distance and greater than or equal to the second preset distance, performs a second path planning and control of the robot; A third edge weight calculation module is configured to obtain the third edge weights between the nodes to be processed using a third edge weight calculation formula based on the teacher's help need, the student's help need, the auditor's help need, and the third adjustment-related data; The third path planning submodule performs third path planning and control of the robot when the personnel distance value is less than the second preset distance.

Images