Optical phenomenon multi-scale observation teaching device and intelligent guidance method thereof

Abstract

The invention relates to the technical field of teaching equipment, in particular to an optical phenomenon multi-scale observation teaching device and an intelligent guidance method thereof, which comprise five modules of optical parameter extraction, operation intention recognition, intelligent guidance, multi-channel feedback and learning evaluation, and aims to convert passive observation into active guidance, remarkably improve learning efficiency and operation accuracy, and improve teaching quality. The optical parameter extraction module obtains image data from microscopic equipment and extracts parameters such as light intensity distribution, the operation intention recognition module collects gesture and eye movement data of students, recognizes current intentions in combination with historical records and predicts errors, and the intelligent guiding module constructs a student cognitive state model and generates personalized guiding strategies. The multi-channel feedback module outputs feedback signals based on strategy activation visual, auditory and tactile channels, and the learning evaluation module receives student response data, operation behavior data and optical parameter change data, updates a cognitive state model and generates a multi-dimensional learning evaluation report.

Description

Technical Field

[0001] The present invention relates to the technical field of teaching equipment, and in particular to a multi-scale observation teaching device for optical phenomena and an intelligent guidance method thereof, which are used to realize intelligent observation and personalized guidance in optical experiment teaching. Background Art

[0002] In current teaching practices, optical microscope imaging experiments are an important part of helping students understand and consolidate optical theory and cultivate practical skills. Traditional optical microscopes consist of an eyepiece and an objective lens, which enable the clear visualization of tiny objects through dual magnification. However, as described in Chinese patent CN113311576A, traditional optical experiments have significant shortcomings:

[0003] On the one hand, due to the small field of view of the microscope, the observer needs to keep his eyes close to the eyepiece and stare at the strong light source for a long time, which can easily lead to visual fatigue and eye damage; on the other hand, the narrow field of view limits the display effect of the teacher's experimental results, and makes it difficult for teachers to supervise and guide each student's experimental operations in real time, resulting in irregular operations and reduced teaching quality.

[0004] Although CN113311576A proposes an experimental auxiliary teaching system that realizes the external display of microscope imaging through image acquisition module, image storage module and image display module, expands the observation field range, reduces visual fatigue, and supports one-to-many display, the system still has the following shortcomings:

[0005] 1. It only provides passive display functions and lacks intelligent guidance mechanisms for student operations;

[0006] 2. Unable to identify students' operational intentions and potential errors, making it difficult to provide targeted guidance;

[0007] 3. Limited image processing capabilities, not supporting real-time analysis and measurement of multi-dimensional optical parameters;

[0008] 4. Failure to intuitively relate observed phenomena to optical theory, resulting in a disconnect between theory and practice;

[0009] 5. Lack of learning assessment and feedback mechanism, which makes it impossible to effectively evaluate students' understanding of optical principles.

[0010] Therefore, there is an urgent need for a new optical teaching device that can provide intelligent guidance and personalized learning support. Summary of the Invention

[0011] The purpose of the present invention is to provide a multi-scale observation teaching device for optical phenomena and an intelligent guidance method thereof, so as to solve the problems existing in the prior art such as passive display, lack of intelligent guidance, and disconnection between theory and practice.

[0012] The present invention proposes a multi-scale observation teaching device for optical phenomena, comprising:

[0013] An optical parameter extraction module is used to acquire optical image data from an optical microscope and extract optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index;

[0014] An operation intention recognition module is in communication with the optical parameter extraction module and is used to collect student gesture and eye movement data, identify current operation intentions based on the gesture and eye movement data combined with historical operation records, and predict potential operation errors;

[0015] an intelligent guidance module, communicatively connected to the operation intention recognition module and the optical parameter extraction module, for receiving the operation intention, optical parameters, and prediction errors, constructing a student cognitive state model, and generating a multi-level personalized guidance strategy based on the cognitive state model;

[0016] a multi-channel feedback module, communicatively connected to the intelligent guidance module, configured to receive the guidance strategy and collaboratively activate the visual, auditory, and tactile channels to output feedback signals based on the guidance strategy;

[0017] A learning evaluation module is respectively communicated with the multi-channel feedback module, the operation intention recognition module and the optical parameter extraction module, and is used to receive the student's response data to the feedback signal, the operation behavior data and the optical parameter change data, update the cognitive state model based on the response data, the operation behavior data and the optical parameter change data, generate a multi-dimensional learning evaluation report, and send the optimized guidance parameters to the intelligent guidance module.

[0018] Preferably, the optical parameter extraction module includes:

[0019] An image acquisition unit, used to acquire image data of an optical microscope device through a high-speed CMOS sensor;

[0020] a parameter extraction unit, communicatively connected to the image acquisition unit, for performing multi-layer convolutional neural network feature extraction, morphological analysis, and spatial frequency domain analysis on the image data;

[0021] a parameter characterization unit, in communication with the parameter extraction unit, for constructing the extracted features into a five-dimensional parameter space representation, including light intensity, position, time, and polarization state;

[0022] The parameter verification unit is in communication with the parameter characterization unit and is used to calculate the confidence index of the parameter and transmit the high-confidence parameter to other modules.

[0023] Preferably, the operation intention recognition module includes:

[0024] A gesture capture unit, which tracks the positions of 22 key points of the hand in real time using a binocular infrared camera array;

[0025] Eye tracking unit, used to collect eye movement data and generate gaze heat maps;

[0026] a multimodal fusion unit, communicatively connected to the gesture capture unit and the eye tracking unit, for integrating gesture, eye movement and historical operation data through a Bayesian network;

[0027] The intention prediction unit is in communication with the multimodal fusion unit and is used to identify the current operation intention based on the intention-behavior mapping matrix and predict possible operation errors by comparing with standard operation procedures.

[0028] Preferably, the intelligent guidance module includes:

[0029] Cognitive modeling unit, used to construct optical knowledge maps and students' cognitive states based on graph structures;

[0030] a guidance strategy unit, communicating with the cognitive modeling unit, for generating hierarchical guidance strategies including direct demonstration, step decomposition, guidance prompts, error feedback, and autonomous exploration;

[0031] a strategy optimization unit, communicatively connected to the guidance strategy unit, for dynamically optimizing the guidance strategy based on a reinforcement learning algorithm;

[0032] The content generation unit is in communication with the policy optimization unit and is configured to generate context-aware multimedia guidance content according to the selected policy.

[0033] Preferably, the multi-channel feedback module includes:

[0034] A visual feedback unit, used to overlay parameter data, operation trajectories, and visual prompts through augmented reality display technology;

[0035] Auditory feedback unit, used to convey parameter change information through 3D spatial audio and voice prompts;

[0036] a tactile feedback unit for providing tactile guidance through a precision force feedback device;

[0037] A feedback coordination unit is communicatively connected with the visual feedback unit, the auditory feedback unit and the tactile feedback unit, and is used to collaboratively control the three-channel feedback intensity and timing according to the guidance strategy.

[0038] Preferably, the learning assessment module includes:

[0039] Behavior capture unit, used to record students' operation sequences, eye tracking and time allocation patterns;

[0040] A multi-dimensional evaluation unit, in communication with the behavior capture unit, for evaluating learning outcomes from three dimensions: operational skills, conceptual understanding, and comprehensive quality;

[0041] a model updating unit, communicatively connected to the multidimensional evaluation unit, for updating the cognitive state model in real time based on the evaluation result;

[0042] A feedback optimization unit is communicatively connected to the model updating unit and is used to generate optimization parameters for the intelligent guidance module.

[0043] Preferably, the cognitive state model includes:

[0044] Knowledge point mastery vector, including the mastery level of 187 optical knowledge points;

[0045] Learning bottleneck identification matrix, used to identify key obstacles in current learning;

[0046] Misconception Mapping Sheet, used to record conceptual confusion and misunderstandings;

[0047] A knowledge memory decay predictor for estimating knowledge retention curves and optimizing review timing.

[0048] As an advantage, it also includes:

[0049] Experiment configuration module, used to set the experiment difficulty coefficient, evaluation criteria and intervention threshold;

[0050] Data management module, used to store and manage students' historical learning data, operation records and evaluation reports;

[0051] The teacher interface module is used to provide teachers with student learning status visualization and teaching strategy adjustment functions;

[0052] System adaptive module, used to continuously optimize algorithm parameters and model structure based on usage data.

[0053] Preferably, the multi-level personalized guidance strategy adaptively adjusts the intervention degree according to the learning stage, including:

[0054] In the beginner stage, direct demonstration and detailed step-by-step breakdown are provided;

[0055] During the practice phase, directional prompts and error feedback are provided;

[0056] In the advancement phase, create exploratory challenges and subtle guidance;

[0057] During the mastery phase, minimal intervention is provided only in cases of severe deviation.

[0058] The intelligent guidance method for teaching multi-scale observation of optical phenomena includes the following steps:

[0059] Acquire image data from optical microscopes and extract optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index;

[0060] Collect students' gesture and eye movement data to identify current operation intentions and predict potential operation errors;

[0061] Generate multi-level personalized guidance strategies based on students' cognitive state models, operational intentions, and optical parameters;

[0062] Coordinately activate visual, auditory, and tactile channels to output feedback signals according to the guidance strategy;

[0063] Collect students' response data to feedback signals, operation behavior data, and optical parameter change data;

[0064] updating the cognitive state model based on the response data, operational behavior data, and optical parameter change data, generating a multi-dimensional learning evaluation report, and optimizing guidance parameters;

[0065] The above steps are executed cyclically to form a closed-loop intelligent guidance process.

[0066] The beneficial effects of the present invention include:

[0067] 1. Transform passive observation into active guidance, significantly improving learning efficiency and operational accuracy;

[0068] 2. Enhance perceptual experience and deepen conceptual understanding and skill retention through a multi-channel feedback mechanism;

[0069] 3. Establish a refined cognitive state model to achieve personalized teaching and intelligent guidance;

[0070] 4. Optimize the allocation of teaching resources, improve teacher guidance efficiency and equipment utilization;

[0071] 5. Improve learning experience, increase student participation and willingness to explore, and reduce experimental anxiety. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] Figure 1 This is a system architecture diagram of the multi-scale observation teaching device for optical phenomena of the present invention;

[0073] Figure 2 This is a structural block diagram of the optical parameter extraction module of the present invention;

[0074] Figure 3 This is a structural block diagram of the operation intention recognition module of the present invention;

[0075] Figure 4This is a structural block diagram of the intelligent guidance module of the present invention;

[0076] Figure 5 This is a structural block diagram of the multi-channel feedback module of the present invention;

[0077] Figure 6 This is a structural diagram of the learning assessment module of the present invention;

[0078] Figure 7 This is a diagram of the structure of the cognitive state model of the present invention;

[0079] Figure 8 A hierarchical diagram of the multi-level personalized guidance strategy of the present invention;

[0080] Figure 9 This is a flow chart of the intelligent guidance method for teaching multi-scale observation of optical phenomena according to the present invention;

[0081] Figure 10 This is a schematic diagram of the application of the present invention in optical interference experiment teaching. DETAILED DESCRIPTION

[0082] Please refer to the attached Figure 1-10 The preferred embodiments of the present invention are described in detail below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and are not used to limit the present invention.

[0083] like Figure 1 As shown, the multi-scale observation teaching device for optical phenomena provided by the present invention includes an optical parameter extraction module 1, an operation intention recognition module 2, an intelligent guidance module 3, a multi-channel feedback module 4 and a learning evaluation module 5.

[0084] Optical Parameter Extraction Module 1 acquires optical image data from optical microscopes and extracts optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index. This module utilizes a high-resolution image sensor and advanced image processing algorithms to capture submicron-level details of optical phenomena, providing the data foundation for subsequent intelligent analysis.

[0085] The Operation Intention Recognition Module 2 is in communication with the Optical Parameter Extraction Module 1 and is used to collect student gesture and eye movement data. Based on this data and historical operation records, it identifies the current operation intention and predicts potential operation errors. This module uses multimodal sensors to capture students' explicit operation behaviors and implicit attention focus, achieving a precise understanding of students' experimental operation intentions.

[0086] The intelligent guidance module 3 is connected to the operation intention recognition module 2 and the optical parameter extraction module 1. It receives the operation intention, optical parameters, and prediction errors, constructs a model of the student's cognitive state, and generates a multi-level personalized guidance strategy based on this cognitive state model. This module is the core decision-making unit of the system and can adaptively adjust the guidance method based on the student's cognitive level and operation characteristics.

[0087] Multi-channel feedback module 4 is in communication with intelligent guidance module 3 and is configured to receive guidance strategies and, based on these strategies, collaboratively activate visual, auditory, and tactile channels to output feedback signals. This module provides students with an immersive learning experience through multi-sensory collaborative feedback, enhancing their intuitive understanding of abstract optical concepts.

[0088] The learning assessment module 5 is communicatively connected to the multi-channel feedback module 4, the operation intention recognition module 2, and the optical parameter extraction module 1. It receives the student's response data to the feedback signal, the operation behavior data, and the optical parameter change data. Based on this data, the cognitive state model is updated, a multi-dimensional learning assessment report is generated, and the optimized guidance parameters are sent to the intelligent guidance module 3. This module closes the feedback loop of the intelligent teaching system, realizing a continuously optimized teaching process.

[0089] The system workflow is as follows: first, the optical parameter extraction module 1 acquires images from the optical microscope and extracts key optical parameters; at the same time, the operation intention recognition module 2 captures the student's gesture and eye movement data to identify the operation intention; then, the intelligent guidance module 3 generates a personalized guidance strategy based on the student's cognitive state, operation intention and optical parameters; the multi-channel feedback module 4 outputs multi-sensory feedback according to the guidance strategy; finally, the learning evaluation module 5 collects response data, updates the cognitive model and optimizes the guidance parameters to form a closed-loop teaching system.

[0090] Each module communicates through a standardized data interface, ensuring real-time and accurate information transmission. The system adopts a modular design, and each functional unit can be flexibly configured according to teaching needs, supporting optical experimental teaching scenarios of varying complexity.

[0091] like Figure 2 As shown, the optical parameter extraction module 1 includes an image acquisition unit 11 , a parameter extraction unit 12 , a parameter characterization unit 13 and a parameter verification unit 14 .

[0092] The image acquisition unit 11 is used to acquire image data from the optical microscope using a high-speed CMOS sensor. This unit utilizes a 12-megapixel, 120-fps high-speed CMOS sensor, coupled with a fine-focus mechanism, capable of capturing submicron-level optical details. The sensor's dynamic range preferably reaches 70 dB, ensuring high-quality images under varying light intensities. The image acquisition unit 11 connects to the optical microscope using a dedicated interface, ensuring stable and real-time image transmission.

[0093] The parameter extraction unit 12 is communicatively connected to the image acquisition unit 11 and is used to perform multi-layer convolutional neural network feature extraction, morphological analysis, and spatial frequency domain analysis on the image data. This unit adopts a multi-level image processing strategy. First, key optical features in the image are extracted using a trained convolutional neural network. Morphological analysis is then performed to identify the structural characteristics of typical optical phenomena such as light spots and stripes. Finally, spatial frequency domain analysis is performed to obtain the frequency characteristics of periodic structures such as interference fringes and diffraction patterns. Preferably, the convolutional neural network adopts a 6-layer structure, including 3 convolutional layers and 3 fully connected layers. It is specially trained for common optical phenomena and has a recognition accuracy rate of over 95%.

[0094] The parameter characterization unit 13 is in communication with the parameter extraction unit 12 and is used to construct the extracted features into a five-dimensional parameter space representation, including light intensity, position, time, and polarization state. This unit performs parameterized characterization on the extracted original features and establishes a standardized optical parameter model. The optical parameter space can be expressed as:

[0095] ,

[0096] in: For spatial location In time Light intensity distribution; and are the polarization angle and phase, respectively. Through this parameterized representation, the system can accurately and quantitatively describe complex optical phenomena.

[0097] The parameter verification unit 14 is in communication with the parameter characterization unit 13 and is used to calculate the confidence index of the parameters and transmit the high-confidence parameters to other modules. This unit evaluates the accuracy and reliability of the extracted parameters by comparing the theoretical model with historical data. The confidence calculation adopts a weighted scoring mechanism:

[0098] ,

[0099] in: For parameters Confidence score of Scoring for the degree of fit with the theoretical model; Score for consistency with historical data; score for temporal stability; 、 、 is the weight coefficient and satisfies Preferably, The confidence threshold is set to 0.8. Parameters above this value are considered high-confidence parameters and transferred to other functional modules; parameters below this value enter the re-verification process.

[0100] The workflow of the optical parameter extraction module 1 is as follows: first, the image acquisition unit 11 acquires the optical microscopic image; then, the parameter extraction unit 12 performs multi-level analysis and processing on the image; then, the parameter characterization unit 13 constructs the extracted features into a standardized parameter representation; finally, the parameter verification unit 14 calculates the parameter confidence and filters low-confidence data to ensure that high-quality optical parameter information is transmitted to other modules of the system.

[0101] like Figure 3 As shown, the operation intention recognition module 2 includes a gesture capture unit 21 , an eye tracking unit 22 , a multimodal fusion unit 23 and an intention prediction unit 24 .

[0102] The gesture capture unit 21 is used to track the positions of 22 key hand points in real time using a binocular infrared camera array. This unit utilizes a binocular infrared camera with a resolution of 1280×960 and a frame rate of 90fps, coupled with a dedicated gesture recognition algorithm. This unit can accurately locate key hand points in 3D space with an accuracy of ±1mm. The system has pre-set 42 gesture commands commonly used in optical experiments, including knob adjustments, lens replacement, and optical path alignment. The recognition accuracy can reach 96% in sufficient ambient light.

[0103] The eye tracking unit 22 is used to collect eye movement data and generate gaze heat maps. This unit uses a binocular eye tracker with a sampling rate of 250Hz, which can accurately capture the student's gaze focus and attention distribution. The system can generate gaze heat maps in real time, showing the intensity distribution of the student's attention area. Preferably, the system sets the attention shift detection threshold to 300ms. That is, when a student gazes at a certain area for more than 300ms, it is judged as intentional attention; rapid glances below this threshold are filtered out to reduce false positives.

[0104] The multimodal fusion unit 23 is connected to the gesture capture unit 21 and the eye tracking unit 22 and is used to integrate gesture, eye movement, and historical operation data through a Bayesian network. This unit is the core processing unit for operation intention recognition and uses a Bayesian network model to achieve probabilistic fusion of multi-source information to form a unified operation intention representation. The model can be expressed as:

[0105] ,

[0106] in: For operational intention; is gesture data; For eye movement data; For historical operation records; is the posterior probability of intention given the observed data; is the likelihood probability of the observed data under the given intention condition; is the prior probability of intention; is the normalization factor. Through this probabilistic fusion mechanism, the system can effectively handle the uncertainty of sensor data and improve the robustness of intent recognition.

[0107] The intention prediction unit 24 is in communication with the multimodal fusion unit 23, and is used to identify the current operation intention based on the intention-behavior mapping matrix, and predict possible operation errors by comparing with the standard operation process. The unit maintains a 42×15 intention-behavior mapping matrix, covering all common operation and parameter combinations, and realizes the mapping from observation data to specific operation intentions. At the same time, by comparing the deviation between the current operation sequence and the standard operation process, the system can predict possible operation errors 0.8-1.2 seconds in advance, providing the possibility for timely intervention. Preferably, the system divides potential errors into three levels according to the severity of the error: slight deviation, moderate error and severe error, corresponding to different intervention strategies.

[0108] The workflow of the Operation Intention Recognition Module 2 is as follows: First, the Gesture Capture Unit 21 and the Eye Tracking Unit 22 synchronously collect student operation behavior data; then, the Multimodal Fusion Unit 23 integrates multi-source data using a Bayesian network; finally, the Intention Prediction Unit 24 identifies the operation intention and predicts potential errors, transmitting the results to the Intelligent Guidance Module. The system utilizes an incremental learning mechanism that automatically adjusts recognition parameters based on the student's historical operation characteristics, improving personalized recognition results.

[0109] like Figure 4 As shown, the intelligent guidance module 3 includes a cognitive modeling unit 31 , a guidance strategy unit 32 , a strategy optimization unit 33 and a content generation unit 34 .

[0110] The cognitive modeling unit 31 is used to construct an optical knowledge map and student cognitive status based on a graph structure. The unit maintains an optical knowledge graph containing 187 knowledge points and 426 associations, covering major fields such as geometric optics, wave optics, and quantum optics. The system uses the Bayesian knowledge tracking algorithm to model the student's cognitive state in real time, generating a 187-dimensional knowledge point mastery vector, with each dimension taking a value of 0-1, indicating the degree of mastery of the corresponding knowledge point. Preferably, the system automatically identifies learning bottlenecks through mastery gradient analysis, that is, areas where the mastery is significantly lower than that of adjacent knowledge points, providing key focus areas for subsequent guidance strategy generation.

[0111] The guidance strategy unit 32 is in communication with the cognitive modeling unit 31 and is used to generate a hierarchical guidance strategy, including direct demonstration, step breakdown, guidance prompts, error feedback, and autonomous exploration. This unit selects the appropriate guidance level and intervention intensity based on the student's cognitive state and the experimental task requirements. The guidance strategy is divided into five levels: direct demonstration (the system automatically completes the operation), step breakdown (complex operations are broken down into detailed steps), guidance prompts (providing directional suggestions), error feedback (providing feedback only when errors are made), and autonomous exploration (minimal intervention). Preferably, the system automatically adjusts the experiment difficulty coefficient based on the student's mastery level, ranging from 0.6 to 1.4, where 1.0 represents standard difficulty.

[0112] The policy optimization unit 33 is in communication with the guidance strategy unit 32 and is used to dynamically optimize the guidance strategy based on a reinforcement learning algorithm. This unit uses a policy gradient-based reinforcement learning method to model the guidance process as a Markov decision process and optimize the guidance strategy by maximizing the long-term cumulative reward. The reward function design comprehensively considers learning effect and student experience:

[0113] ,

[0114] in: is the total reward value; The degree of improvement in knowledge status; for student engagement; The time required to complete the task; 、 、 is the weight coefficient. Preferably, , , focusing on learning outcomes. Strategy optimization uses online learning methods, which can continuously adjust guidance strategies based on students' real-time responses to achieve personalized adaptation. The average value of the difference between the knowledge mastery vector at the end of the current session and the vector at the beginning of the session is calculated. Each parameter is normalized and the dimensions are unified: E / 100 and T / , ensuring that all parameters are in the range [0,1] to avoid participation being dominated by large values.

[0115] The content generation unit 34 is in communication with the strategy optimization unit 33 and is responsible for generating context-aware multimedia guidance content based on the selected strategy. This unit includes a dynamic guidance content generation engine that supports mixed multimedia output such as text, images, animation, and voice. The system includes built-in parameterized simulation models for 64 typical optical phenomena, capable of generating corresponding demonstration content based on the current experimental scenario. Preferably, the guidance content automatically adjusts the density of professional terminology and the level of explanation detail based on students' cognitive characteristics to ensure easy understanding and targeted content.

[0116] The intelligent guidance module 3 operates as follows: First, the cognitive modeling unit 31 constructs a cognitive state model based on the student's historical performance and current actions; then, the guidance strategy unit 32 selects an appropriate guidance strategy based on the cognitive state; then, the strategy optimization unit 33 optimizes the strategy parameters through reinforcement learning; and finally, the content generation unit 34 generates specific guidance content and transmits it to the multi-channel feedback module for presentation. This entire process forms a closed-loop optimization system, continuously improving guidance effectiveness.

[0117] like Figure 5 As shown, the multi-channel feedback module 4 includes a visual feedback unit 41 , an auditory feedback unit 42 , a tactile feedback unit 43 and a feedback coordination unit 44 .

[0118] Visual feedback unit 41 is used to overlay parameter data, operation trajectories, and visual cues using augmented reality display technology. This unit utilizes a micro-projection augmented reality display system with a resolution of 1920×1080 and a refresh rate of 120Hz. With a field of view of 100°, it seamlessly overlays virtual images with actual optical phenomena within the student's field of view. The system supports real-time overlay of experimental parameters, visualization of operation trajectories (a hybrid representation of heat maps and vector diagrams), and eight types of visual cues, with brightness, color, and shape dynamically adjusted based on importance. Preferably, visual information is arranged hierarchically, with important information located in the center of the visual field and less important information distributed peripherally to reduce visual clutter.

[0119] The auditory feedback unit 42 is used to convey parameter change information through 3D spatial audio and voice prompts. This unit includes a 3D spatial audio engine, a parameter change sound mapping system, and a speech synthesis system that supports emotional modulation. By mapping abstract parameters into intuitive sounds, such as light intensity changes into volume changes and frequency changes into pitch changes, the system can create an intuitive connection between parameters and sounds. Preferably, the system dynamically adjusts sound characteristics based on the importance of the prompt, including volume (40-75dB), timbre (soft to clear), and urgency (calm to tense), creating a multi-layered auditory experience.

[0120] The tactile feedback unit 43 is used to provide tactile guidance through a precision force feedback device. The unit adopts a precision tactile feedback system with 16-bit resolution and a force feedback range of 0.1-5N. It has 32 built-in standard tactile modes and can simulate the tactile experience corresponding to different physical phenomena. The system converts abstract optical parameters into specific tactile experiences through a parameter-tactile mapping engine, such as mapping light intensity gradients to force field strength and mapping interference fringes to rhythmic pulses. Preferably, the tactile feedback adopts a progressive design, with a slight initial tactile prompt. As the adjustment deviation increases, the tactile intensity gradually increases, forming an intuitive directional guidance.

[0121] The feedback coordination unit 44 is in communication with the visual feedback unit 41, the auditory feedback unit 42, and the tactile feedback unit 43, and is used to collaboratively control the three-channel feedback intensity and timing according to the guidance strategy. This unit is the central controller of multi-channel feedback, responsible for coordinating the operating mode, intensity level, and timing arrangement of each feedback channel according to the strategy configuration provided by the intelligent guidance module. The system adopts a hierarchical control strategy, with the high-level controller responsible for channel selection and overall intensity, and the low-level controller responsible for fine-tuning the internal parameters of each channel. Preferably, the system dynamically adjusts the feedback intensity based on student response and environmental conditions to ensure that the feedback is both clear and perceptible without causing excessive interference.

[0122] The multi-channel feedback module 4 operates as follows: First, the feedback coordination unit 44 receives the guidance strategy from the intelligent guidance module; then, based on the strategy's content and priority, it coordinates and configures the parameters of each feedback channel; finally, the visual, auditory, and tactile feedback units output feedback signals synchronously or in a specific sequence. The system supports real-time monitoring of feedback effectiveness and dynamically adjusts feedback methods based on student responses, forming an adaptive feedback mechanism.

[0123] like Figure 6 As shown, the learning evaluation module 5 includes a behavior capturing unit 51 , a multi-dimensional evaluation unit 52 , a model updating unit 53 and a feedback optimization unit 54 .

[0124] The behavior capture unit 51 is used to record students' operation sequences, eye gaze trajectories, and time allocation patterns. By integrating multi-source sensor data, this unit constructs a comprehensive record of students' learning behavior. The system captures multidimensional data, including operation sequence and rhythm, eye gaze movement paths, attention allocation ratios, and trial-and-error behavior characteristics, providing an objective basis for subsequent evaluation. Preferably, the system performs feature extraction and pattern recognition on the captured behavioral data to identify typical learning styles and problem-solving strategies, such as systematic exploration, trial-and-error experimentation, and intuitive operation, providing support for personalized assessment.

[0125] The multidimensional assessment unit 52 is in communication with the behavior capture unit 51 and is used to evaluate learning outcomes across three dimensions: operational skills, conceptual understanding, and overall quality. This unit utilizes a multi-level assessment framework to comprehensively assess learning outcomes from different perspectives. The operational skills dimension includes four indicators: operational accuracy (with standard deviation), operational fluency (based on speed stability), adaptability (ability to cope with abnormal situations), and time efficiency; the conceptual understanding dimension includes four indicators: parameter association understanding, phenomenon-theory mapping accuracy, transfer application ability, and problem-solving efficiency; and the overall quality dimension includes three indicators: exploration and innovation index, systems thinking score, and autonomous learning ability. Preferably, each indicator uses a standardized score range of 0-100, and the weight configuration is dynamically adjusted according to the experiment type to ensure the scientific and targeted nature of the evaluation results.

[0126] The model updating unit 53 is in communication with the multidimensional evaluation unit 52 and is used to update the cognitive state model in real time based on the evaluation results. This unit uses a Bayesian knowledge network real-time update mechanism to dynamically adjust the knowledge point mastery assessment based on student performance. The system considers the knowledge memory decay factor, simulates the human memory law, predicts the knowledge retention curve, and reasonably arranges the review nodes. The knowledge state update formula is:

[0127] ,

[0128] in: For time the state of knowledge; Rate your current performance; is the learning gain coefficient; is the attenuation coefficient; is the last review time. Dynamically adjusted according to students' learning styles, ranging from 0.1-0.4; Set according to the complexity of the knowledge point, ranging from 0.001 to 0.01.

[0129] The feedback optimization unit 54 is in communication with the model update unit 53 and is used to generate optimization parameters for the intelligent guidance module. Based on the updated cognitive model and learning assessment results, this unit generates guidance strategy optimization recommendations, including intervention level adjustment, guidance content customization, and feedback channel selection. The system adopts a closed-loop optimization design, directly feeding learning assessment results back to the guidance system to achieve continuous improvement in the teaching process. Preferably, the system maintains individual student learning profiles, recording learning trajectories and preferences, supporting long-term learning planning and personalized guidance strategy customization.

[0130] The learning assessment module 5 operates as follows: First, the behavior capture unit 51 comprehensively records student learning behavior data; then, the multidimensional assessment unit 52 evaluates learning outcomes from multiple dimensions; then, the model update unit 53 updates the cognitive state model based on the assessment results; finally, the feedback optimization unit 54 generates guidance strategy optimization parameters and sends them to the intelligent guidance module, completing the teaching loop. The system supports a combination of real-time evaluation and periodic summaries to meet diverse teaching needs.

[0131] like Figure 7 As shown, the cognitive state model includes a knowledge point mastery vector, a learning bottleneck identification matrix, an error cognition mapping table, and a knowledge memory decay predictor.

[0132] The knowledge point mastery vector contains the mastery level of 187 optical knowledge points. This vector is a core component of the cognitive state model. Each element ranges from 0 to 1, indicating the mastery level of the corresponding knowledge point. The system uses fine-grained knowledge representation to decompose the optical knowledge system into knowledge points at multiple levels, including basic concepts, theoretical models, experimental principles, and practical skills. Preferably, a pre-relationship network is established between knowledge points, forming a directed acyclic graph structure to support knowledge dependency reasoning and assist in learning path planning.

[0133] The learning bottleneck identification matrix is used to identify key obstacles in current learning. By analyzing the gradient distribution of knowledge point mastery vectors, the matrix identifies areas with abnormally low mastery or significant differences from related knowledge points, thereby determining learning bottlenecks. The system uses multiple bottleneck identification algorithms, including local minimum detection, mastery gradient analysis, and knowledge association path blockage detection, to comprehensively determine the location and nature of learning obstacles. Preferably, the system categorizes bottlenecks into three types: conceptual obstacles, skill obstacles, and complex obstacles, and adopts different teaching intervention strategies for different types.

[0134] The Misperception Mapping Table records conceptual confusion and misunderstandings. This table maintains typical error patterns exhibited by students during the learning process, including conceptual confusion, misunderstanding of principles, and reversal of cause and effect. By analyzing operational behaviors and response patterns, the system infers potential misperceptions and constructs a personalized error model. The system optimally includes 60 common error patterns in optical learning and can dynamically expand personalized error models through learning behavior analysis, improving the accuracy and specificity of error diagnosis.

[0135] The knowledge memory decay predictor is used to estimate knowledge retention curves and optimize review timing. This predictor is based on the Ebbinghaus forgetting curve principle and combines individual student memory characteristics to simulate how knowledge memory strength changes over time. Using the memory decay model, the system predicts the memory retention curve for each knowledge point and rationally arranges knowledge review and reinforcement sessions. The memory strength prediction model uses an exponential decay function:

[0136] ,

[0137] in: For time Memory strength; For learning time points; is the personalized memory decay coefficient, and the constraint condition t > , ensuring that the exponential term is negative and that the memory strength value is monotonically decreasing in the range of 0-1. Preferably, the system dynamically adjusts the memory strength based on the student's historical performance. The value range is 0.01-0.1, which realizes the personalized memory model.

[0138] The cognitive state model works as follows: the system continuously updates the knowledge point mastery vector using multi-source data; identifies learning bottlenecks based on mastery distribution; detects cognitive errors through behavioral analysis; and predicts the knowledge retention curve based on memory decay patterns. These four components work together to construct a comprehensive and dynamic representation of a student's cognitive state, providing a precise basis for the generation of intelligent guidance strategies.

[0139] The optical phenomenon multi-scale observation teaching device provided in this embodiment further includes an experiment configuration module, a data management module, a teacher interface module and a system adaptation module.

[0140] The experiment configuration module is used to set the experiment difficulty coefficient, evaluation criteria, and guided intervention threshold. This module provides an experimental parameter configuration interface, allowing teachers to adjust the system operating parameters according to teaching objectives and student levels. The experiment difficulty coefficient ranges from 0.6 to 1.4, where 1.0 represents standard difficulty; the evaluation criteria include loose, standard, and strict modes, corresponding to different accuracy requirements; the guided intervention threshold determines the timing of system intervention and ranges from 0.1 to 0.9, with lower values indicating more active intervention. Preferably, the system supports automatic configuration mode, which automatically adjusts configuration parameters based on students' historical performance and learning goals, reducing the workload of teachers.

[0141] The data management module is used to store and manage students' historical learning data, operation records, and assessment reports. This module utilizes a layered data architecture, including a real-time data cache layer, a session data storage layer, and a long-term data archiving layer. The system provides structured organization for learning data, supporting multi-dimensional query and analysis. Optimally, the system utilizes incremental backup and differential storage strategies to effectively reduce storage overhead while ensuring data integrity and traceability. Data storage adheres to privacy protection principles, employing encrypted storage and access control mechanisms to safeguard student privacy.

[0142] The Teacher Interface Module provides teachers with visualization of student learning status and the ability to adjust teaching strategies. This module offers an intuitive teacher control interface, including real-time monitoring, learning trajectory analysis, and statistical report generation. The system supports multi-level data visualization, from macro-statistics to micro-behavioral details, to meet diverse teaching analysis needs. Ideally, the system provides a teaching intervention interface, allowing teachers to override the system's automated decisions and implement manual intervention when necessary. The intervention process and results are recorded as a reference for system optimization.

[0143] The system's adaptive module continuously optimizes algorithm parameters and model structure based on usage data. This module enables the adaptive evolution of the system, continuously optimizing the operating parameters and algorithm models of each functional module by analyzing extensive usage data. The system employs a hybrid optimization strategy, combining supervised learning (based on teacher intervention records) and unsupervised learning (based on learning effect analysis), to achieve multi-level optimization. Optimally, the system performs weekly parameter fine-tuning, monthly model updates, and semesterly architecture evaluations to ensure continuous improvement in system performance.

[0144] These four auxiliary modules work in conjunction with the core functional module to form a complete teaching support ecosystem. The experiment configuration module ensures the system's adaptability to diverse teaching scenarios; the data management module provides reliable data support; the teacher interface module enhances human-computer collaboration; and the system adaptation module enables continuous optimization. Through the synergy of these modules, the system provides comprehensive, flexible, and efficient teaching support services.

[0145] like Figure 8 As shown, the multi-level personalized guidance strategy of the present invention adaptively adjusts the intervention degree according to the learning stage, including four levels: beginner stage, practice stage, improvement stage and mastery stage.

[0146] During the initial learning phase, the system provides both direct demonstrations and detailed step-by-step instructions. For students new to specific optical experiments or concepts, the system employs a highly structured approach. Direct demonstrations simultaneously demonstrate standard operating procedures through visual, auditory, and tactile channels, helping students establish a preliminary understanding. Detailed step-by-step instructions break down complex experimental procedures into simple, manageable steps, guiding students through them step by step. Optimally, the system adopts a 90% intervention ratio during this phase, providing a near-fully guided learning experience and focusing on developing basic operational skills and safety awareness.

[0147] During the practice phase, the system provides directional prompts and error feedback. Once students have mastered basic operations, the system transitions to a moderate-intensity guidance mode. Directional prompts no longer provide detailed instructions for each step, but instead provide overall direction and key points. Error feedback provides timely corrections when students' operations deviate from the expected path. Optimally, the system reduces the intervention ratio to 50-60% during this phase, encouraging students to think and solve problems independently while maintaining necessary safety supervision to cultivate practical skills and problem-solving awareness.

[0148] During the advanced learning phase, the system creates exploratory challenges and subtle guidance. For students with a solid foundation, the system shifts to a less intensive guidance model. Exploratory challenges set open-ended tasks, encouraging students to go beyond standard experimental procedures and explore innovative paths. Subtle guidance provides implicit support through subtle feedback without interrupting the learning process. Optimally, the system reduces intervention to 20-30% during this phase, focusing primarily on deepening concepts and cultivating innovative thinking, guiding students to establish a deep connection between theory and practice.

[0149] During the mastery phase, the system provides minimal intervention only when significant deviations occur. For students nearing mastery, the system adopts a near-non-intervention support model. The system primarily serves as a safety net, providing minimal intervention only when operations significantly deviate from safe ranges or core principles. Optimally, the system reduces intervention to 5-10% during this phase, prioritizing support for independent exploration and innovative practice, fostering professional confidence and academic innovation.

[0150] The core mechanism of this multi-level personalized guidance strategy is dynamic adaptability. The system smoothly transitions between different levels based on students' real-time performance, ensuring just the right level of support. Furthermore, the system independently assesses learning stages for different knowledge points and skills, enabling fine-grained personalized guidance. Even within the same experiment, different levels of guidance strategies can be applied to different operational steps, maximizing learning efficiency and experience.

[0151] like Figure 9 As shown, the intelligent guidance method for multi-scale observation teaching of optical phenomena provided by the present invention includes the following steps:

[0152] Acquire image data from an optical microscope and extract optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index. This step uses high-precision image acquisition equipment to capture high-quality images of optical phenomena and employs advanced image processing algorithms to extract key optical parameters. The system can identify and quantify characteristic parameters of optical phenomena, such as interference fringes, diffraction patterns, and polarization state changes, providing foundational data for subsequent analysis.

[0153] The system collects student gesture and eye movement data to identify current operational intentions and predict potential operational errors. This step uses multimodal sensors to capture students' operational behaviors and attentional focus, combining historical operational records and behavioral patterns to accurately identify current operational intentions. The system uses a Bayesian network to fuse multi-source data to reliably infer operational intentions. By comparing with standard operating procedures, the system can predict potential operational errors in advance, creating conditions for timely intervention.

[0154] A multi-level personalized guidance strategy is generated based on the student's cognitive state model, operational intention, and optical parameters. This step is the core decision-making process of the system, integrating two key processes: cognitive modeling and strategy optimization. The system first constructs an accurate representation of the cognitive state by maintaining a 187-dimensional knowledge mastery vector. Then, based on the cognitive state, current task, and operational intention, the system selects the most appropriate guidance strategy from five levels, ranging from direct demonstration to minimal intervention. Finally, a reinforcement learning algorithm is used to optimize the strategy parameters to achieve personalized guidance.

[0155] Based on the guidance strategy, the system collaboratively activates visual, auditory, and tactile channels to output feedback signals. This step transforms the abstract guidance strategy into concrete multi-sensory feedback. Through the coordinated operation of augmented reality display, 3D spatial audio, and precise tactile feedback, the system creates an immersive learning experience for students. The system dynamically adjusts the feedback parameters of each channel based on the guidance content and importance, ensuring effective and non-intrusive information delivery.

[0156] Collect data on students' responses to feedback signals, their operational behavior, and changes in optical parameters. This step comprehensively captures students' learning behaviors and response patterns, building a rich evaluation dataset. The system records the time and manner of students' responses to feedback, subsequent changes in operational behavior, and changes in optical parameters caused by operational adjustments, providing a multi-dimensional basis for learning assessment.

[0157] Based on response data, operational behavior data, and optical parameter change data, the cognitive state model is updated, a multidimensional learning assessment report is generated, and guidance parameters are optimized. This step closes the feedback loop of intelligent teaching, achieving continuous optimization. The system uses a Bayesian update mechanism to adjust knowledge mastery assessments; a multidimensional assessment framework is used to generate a comprehensive assessment covering operational skills, conceptual understanding, and overall quality; and guidance parameters are optimized based on the assessment results to improve subsequent teaching effectiveness.

[0158] The above steps are repeated repeatedly, forming a closed-loop intelligent guidance process. This step ensures that the system continuously and dynamically supports the learning process. As learning progresses, the system continuously accumulates personalized data, continuously optimizes cognitive models and guidance strategies, and achieves increasingly precise personalized support. Preferably, the system generates a learning summary report after each teaching session, helping students and teachers review their learning trajectory and plan subsequent learning paths.

[0159] This method achieves intelligent guidance and personalized support in optics teaching through a closed-loop process of "observation-analysis-guidance-feedback-evaluation", significantly improving teaching effectiveness and learning experience.

[0160] like Figure 10 As shown, the present invention can be applied to optical interference experiment teaching. Taking Young's double-slit interference experiment as an example, the system can provide full intelligent guidance support.

[0161] First, the optical parameter extraction module acquires interference fringe images in real time, accurately measuring parameters such as fringe spacing, light intensity distribution, and contrast. The system can identify fringe quality issues, such as insufficient contrast and stray light interference, providing a basis for subsequent adjustments.

[0162] At the same time, the operation intention recognition module captures students' intentions to adjust the light source position, adjust the distance between the two slits, and so on. The system can predict incorrect operations that may cause the interference fringes to disappear, such as when the light source deviates too much from the center axis or when the slit width is improperly adjusted.

[0163] Based on the student's cognitive state and operational intent, the intelligent guidance module generates personalized guidance strategies. For beginners, the system provides detailed step-by-step instructions, such as "First adjust the light source to the center position, then fine-tune the distance between the two slits." For advanced learners, the system provides principled guidance, such as "Consider the relationship between optical path difference and fringe spacing."

[0164] The multi-channel feedback module uses augmented reality technology to overlay light path diagrams, wave equations, and expected interference patterns in students' fields of view. It also uses spatial audio feedback to display changes in light intensity, and tactile feedback to guide fine-tuning adjustments. This multi-sensory integration significantly enhances intuitive understanding of abstract concepts.

[0165] The learning assessment module records the students' adjustment process from chaotic fringes to clear interference patterns, evaluates their understanding of concepts such as optical path difference, phase, wavelength, etc., generates personalized learning reports, and provides targeted suggestions for subsequent learning.

[0166] In a controlled trial involving 60 university students, the group using this system for optical interferometry experiments showed a 42.6% increase in experimental success rate, a 36.8% improvement in conceptual understanding, a 28.4% reduction in experimental completion time, and a 51.8% increase in learning satisfaction compared to the group using traditional instruction. These data strongly demonstrate the significant advantages of this invention in optics teaching.

[0167] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A multi-scale observation teaching device for optical phenomena, characterized in that: include: An optical parameter extraction module is used to acquire optical image data from an optical microscope and extract optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index; An operation intention recognition module is in communication with the optical parameter extraction module and is used to collect student gesture and eye movement data, identify current operation intentions based on the gesture and eye movement data combined with historical operation records, and predict potential operation errors; an intelligent guidance module, communicatively connected to the operation intention recognition module and the optical parameter extraction module, for receiving the operation intention, optical parameters, and prediction errors, constructing a student cognitive state model, and generating a multi-level personalized guidance strategy based on the cognitive state model; a multi-channel feedback module, communicatively connected to the intelligent guidance module, configured to receive the guidance strategy and collaboratively activate the visual, auditory, and tactile channels to output feedback signals based on the guidance strategy; A learning evaluation module is communicatively connected to the multi-channel feedback module, the operation intention recognition module and the optical parameter extraction module respectively, and is used to receive students' response data to feedback signals, operation behavior data and optical parameter change data, update the cognitive state model based on the response data, operation behavior data and optical parameter change data, generate a multi-dimensional learning evaluation report, and send optimized guidance parameters to the intelligent guidance module.

2. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The optical parameter extraction module includes: An image acquisition unit, used to acquire image data of an optical microscope device through a high-speed CMOS sensor; a parameter extraction unit, communicatively connected to the image acquisition unit, for performing multi-layer convolutional neural network feature extraction, morphological analysis, and spatial frequency domain analysis on the image data; a parameter characterization unit, in communication with the parameter extraction unit, for constructing the extracted features into a five-dimensional parameter space representation, including light intensity, position, time, and polarization state; The parameter verification unit is in communication with the parameter characterization unit and is used to calculate the confidence index of the parameter and transmit the high-confidence parameter to other modules.

3. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The operation intention recognition module includes: A gesture capture unit, which tracks the positions of 22 key points of the hand in real time using a binocular infrared camera array; Eye tracking unit, used to collect eye movement data and generate gaze heat maps; a multimodal fusion unit, communicatively connected to the gesture capture unit and the eye tracking unit, for integrating gesture, eye movement and historical operation data through a Bayesian network; The intention prediction unit is in communication with the multimodal fusion unit and is used to identify the current operation intention based on the intention-behavior mapping matrix and predict possible operation errors by comparing with standard operation procedures.

4. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The intelligent guidance module includes: Cognitive modeling unit, used to construct optical knowledge maps and students' cognitive states based on graph structures; a guidance strategy unit, communicating with the cognitive modeling unit, for generating hierarchical guidance strategies including direct demonstration, step decomposition, guidance prompts, error feedback, and autonomous exploration; a strategy optimization unit, communicatively connected to the guidance strategy unit, for dynamically optimizing the guidance strategy based on a reinforcement learning algorithm; The content generation unit is in communication with the policy optimization unit and is configured to generate context-aware multimedia guidance content according to the selected policy.

5. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The multi-channel feedback module includes: A visual feedback unit, used to overlay parameter data, operation trajectories, and visual prompts through augmented reality display technology; Auditory feedback unit, used to convey parameter change information through 3D spatial audio and voice prompts; a tactile feedback unit for providing tactile guidance through a precision force feedback device; A feedback coordination unit is communicatively connected with the visual feedback unit, the auditory feedback unit and the tactile feedback unit, and is used to collaboratively control the three-channel feedback intensity and timing according to the guidance strategy.

6. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The learning assessment module includes: Behavior capture unit, used to record students' operation sequences, eye tracking and time allocation patterns; A multi-dimensional evaluation unit, in communication with the behavior capture unit, for evaluating learning outcomes from three dimensions: operational skills, conceptual understanding, and comprehensive quality; a model updating unit, communicatively connected to the multidimensional evaluation unit, for updating the cognitive state model in real time based on the evaluation result; A feedback optimization unit is communicatively connected to the model updating unit and is used to generate optimization parameters for the intelligent guidance module.

7. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The cognitive state model includes: Knowledge point mastery vector, including the mastery level of 187 optical knowledge points; Learning bottleneck identification matrix, used to identify key obstacles in current learning; Misconception Mapping Sheet, used to record conceptual confusion and misunderstandings; A knowledge memory decay predictor for estimating knowledge retention curves and optimizing review timing.

8. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: Also includes: Experiment configuration module, used to set the experiment difficulty coefficient, evaluation criteria and intervention threshold; Data management module, used to store and manage students' historical learning data, operation records and evaluation reports; The teacher interface module is used to provide teachers with student learning status visualization and teaching strategy adjustment functions; System adaptive module, used to continuously optimize algorithm parameters and model structure based on usage data.

9. The optical phenomenon multi-scale observation teaching device according to claim 1, characterized in that: The multi-level personalized guidance strategy adaptively adjusts the degree of intervention according to the learning stage, including: In the beginner stage, direct demonstration and detailed step-by-step breakdown are provided; During the practice phase, directional prompts and error feedback are provided; In the advancement phase, create exploratory challenges and subtle guidance; During the mastery phase, minimal intervention is provided only in cases of severe deviation.

10. An intelligent guidance method for teaching multi-scale observation of optical phenomena, using the system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Acquire image data from optical microscopes and extract optical parameters including light intensity distribution, diffraction angle, interference fringe spacing, polarization state, and refractive index; Collect students' gesture and eye movement data to identify current operation intentions and predict potential operation errors; Generate multi-level personalized guidance strategies based on students' cognitive state models, operational intentions, and optical parameters; Coordinately activate visual, auditory, and tactile channels to output feedback signals according to the guidance strategy; Collect students' response data to feedback signals, operation behavior data, and optical parameter change data; updating the cognitive state model based on the response data, operational behavior data, and optical parameter change data, generating a multi-dimensional learning evaluation report, and optimizing guidance parameters; The above steps are executed cyclically to form a closed-loop intelligent guidance process.

Images