Intelligent measurement and AI analysis system for middle school electrical experiment
By integrating voltage, current, and temperature acquisition modules into middle school electrical experiments and setting up online large-scale models and offline local analysis modes on a host computer platform, the problems of insufficient data analysis capabilities and network dependence in middle school electrical experiments are solved, and stability and teaching efficiency are improved under different network environments.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-02-10
- Publication Date
- 2026-04-14
AI Technical Summary
There are problems in middle school electrical experiments, such as the single dimension of measurement data (lack of correlation monitoring between temperature and electrical properties), the fact that digital equipment only has recording functions but lacks in-depth intelligent analysis capabilities, and the fact that existing intelligent teaching systems rely too much on the network environment, making it impossible to carry out effective data processing and analysis under conditions of no network or weak network.
An intelligent measurement and AI analysis system for electrical experiments in middle schools was designed, including a hardware measurement terminal and a host computer processing platform. The hardware terminal integrates voltage and current acquisition modules, temperature acquisition modules, and communication modules. The microcontroller main control module performs local calculations, and the host computer platform is set up with online large model and offline local analysis modes to realize intelligent data analysis and practice diagnosis.
It achieves improved stability and adaptability in different network environments, can simultaneously monitor electrical and thermal characteristics, provides intuitive data analysis and personalized learning suggestions, reduces dependence on the network, and improves teaching efficiency and classroom interaction experience.
Smart Images
Figure CN121861983A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of educational informatization technology, specifically to an intelligent measurement and AI analysis system for electrical experiments in middle schools. Background Technology
[0002] Physics experiments in middle school are an important part of cultivating students' scientific inquiry ability and logical thinking. Among them, electrical experiments (such as exploring Ohm's law and measuring the volt-ampere characteristic curve of a light bulb) account for a large proportion. In traditional electrical experiment teaching, pointer-type voltmeters and ammeters are usually used in combination with sliding rheostats and the load to be measured to form an experimental circuit. However, this traditional experimental mode has many limitations in actual teaching.
[0003] On the one hand, pointer-type instruments not only suffer from parallax and estimation errors in reading, but also have complicated wiring ports. Students often spend a lot of time on circuit connections and correcting misconnections, making it difficult to focus on exploring physical laws. On the other hand, traditional measurement methods can only obtain voltage and current data and cannot simultaneously monitor the temperature changes of the load. As a result, when students analyze nonlinear physical phenomena such as the increase in the resistance of a light bulb filament with increasing temperature, they lack intuitive temperature data support and find it difficult to establish a quantitative relationship between electrical characteristics and thermal effects.
[0004] With the development of educational informatization technology, Digital Information Systems (DIS) experimental equipment has gradually entered the classroom. Although existing digital experimental equipment can automatically collect data through sensors and display waveforms on computer screens, it essentially only functions as a data recorder and lacks in-depth intelligent analysis capabilities. When students are faced with the collected discrete data points or curves, the system often cannot provide immediate interpretation of physical laws or error analysis, causing students to mechanically record data without understanding the physical meaning behind it. In addition, although some existing intelligent teaching aids attempt to introduce artificial intelligence technology to comment on experimental results, these systems usually rely heavily on a stable Internet connection to access cloud computing power. Once the laboratory network environment is poor or the server response is delayed, the entire teaching process will be interrupted. At the same time, relying solely on the analysis of large cloud models is a waste of resources when faced with simple linear fitting tasks, while local systems lacking cloud support are often too limited in function to handle the interpretation of complex physical phenomena. This strong dependence on the network environment and the disconnect between local and cloud analysis capabilities limit the popularization and application of intelligent experimental systems in actual middle school classrooms.
[0005] Furthermore, existing intelligent homework / question bank platforms mostly focus on automatic question judging and question recommendation based on online question banks, or on generating learning paths based on knowledge graphs / knowledge point statistics. However, they are usually independent of physics experiment measurement terminals, and cannot unify experimental data analysis and practice diagnosis in the same teaching loop. Moreover, the description of the sequential correspondence between "each question / blank - knowledge point - standard answer" is insufficient, making it difficult to form a traceable and interpretable fine-grained answer record set, thus affecting the reliability and consistency of personalized diagnosis. Therefore, this invention designs an intelligent measurement and AI analysis system for middle school electrical experiments based on the above-mentioned problems. Summary of the Invention
[0006] To address the shortcomings of existing technologies, this invention provides an intelligent measurement and AI analysis system for middle school electrical experiments. It solves the problems existing in current middle school electrical experiment teaching, such as the single dimension of measurement data (lack of correlation monitoring between temperature and electrical characteristics), the fact that digital equipment only has recording functions but lacks in-depth intelligent analysis capabilities, and the fact that existing intelligent teaching systems rely too much on the network environment, resulting in the inability to perform effective data processing and analysis under conditions of no network or weak network.
[0007] To achieve the above objectives, the present invention is implemented through the following technical solution: an intelligent measurement and AI analysis system for middle school electrical experiments, including a hardware measurement terminal and a host computer processing platform; The hardware measurement terminal includes: The power supply and interface module is used to provide the operating voltage for the system. The microcontroller main control module serves as the control core of the system. The voltage and current acquisition module is connected to the load under test and is used to acquire the voltage and current data of the load. The temperature acquisition module is installed on the surface of the component being measured and is used to acquire temperature data. And a communication module, used to establish a data transmission link between the hardware measurement terminal and the host computer processing platform; The microcontroller main control module is connected to the voltage and current acquisition module, the temperature acquisition module and the communication module respectively. It is used to receive the acquired data and perform local calculations, and send the data to the host computer processing platform through the communication module. The host computer processing platform runs host computer software, which includes: The data communication and storage module is used to receive and store experimental data from the hardware measurement terminal; The data visualization module is used to plot physical property curves based on experimental data; And an AI analysis module for intelligent analysis of experimental data; The AI analysis module is configured with two working modes: Online large model analysis mode: When a normal network connection is detected and a valid API key is available, the remote large model interface is invoked to perform physical law analysis on the experimental data based on preset prompts. Offline local analysis mode: When the remote large model interface cannot be called, the local offline analysis algorithm is called to perform statistical calculations and linear fitting analysis on the experimental data.
[0008] Preferably, the local computing logic of the microcontroller main control module is as follows: The microcontroller main control module periodically acquires voltage data. and current data Calculate the resistance value according to Ohm's law Calculate the power value according to the electric power formula. The hardware measurement terminal also includes a display module, which is connected to the microcontroller main control module via a communication interface. The microcontroller main control module is configured to display the real-time acquired voltage data. Current data Temperature data and the calculated resistance value and power value The text is refreshed and displayed on the display module as multi-line text.
[0009] Preferably, the specific circuit structure of the hardware measurement terminal is as follows: The voltage and current acquisition module uses a sensor module, which is connected to the experimental circuit in series or parallel, and is connected via a serial port or interface. The temperature acquisition module uses a single-bus digital temperature sensor, with its probe in close contact with the outer surface of the load being measured. The communication module uses a WiFi wireless communication module or a Bluetooth wireless communication module, which connects to the microcontroller main control module via a serial port and sends the measurement dataset containing timestamps using a text protocol.
[0010] Preferably, the specific execution steps of the offline local analysis mode of the host computer processing platform include: Calculate the average resistance and standard deviation of each measurement point in the selected experimental dataset. Based on the standard deviation, determine the resistance stability by performing a linear fit between the resistance value and the temperature value to determine the trend of resistance change with temperature. Calculate the ratio of voltage to current to determine whether the data conforms to Ohm's law and detect whether the power exceeds the preset rated power threshold. If the threshold is exceeded, generate an overload risk warning.
[0011] Preferably, the specific execution steps of the online large model analysis mode of the host computer processing platform include: Construct prompts that include experimental data, physical experimental background, and analysis requirements; The prompt word is sent to the remote large model server via an HTTP request; Receive natural language text returned by the server, the text containing a qualitative analysis of the resistance change trend, power relationship and possible sources of experimental error; The natural language text is displayed in the analysis results area of the host computer software.
[0012] Preferably, the host computer software further includes a practice diagnostic module; The practice diagnostic module runs in an interface area independent of the experimental data visualization module and the experimental data AI analysis module. The practice questions are pre-compiled by the teacher and loaded locally. The practice diagnosis module is configured to: automatically judge students' answers based on the standard answers, and associate each question or sub-question with at least one knowledge point identifier to form a set of answer records; The practice diagnosis module is further configured to: without generating questions or modifying standard answers, summarize the answer record set by knowledge points to construct prompt words, and call the remote large model interface or offline analysis strategy to output personalized learning suggestions.
[0013] Preferably, the host computer software further includes a practice question display module; The exercise display module runs in an interface area independent of the AI analysis module; The exercise display module loads a pre-set static question bank, which contains pre-defined question text, options, and standard answers; The system is configured such that the content displayed by the exercise display module does not change with changes in experimental data or the output results of the AI analysis module, and does not call the large model interface to generate questions.
[0014] Preferably, the specific functions of the data visualization module include: Read the stored experimental data and plot the IU characteristic curve with voltage as the x-axis and current as the y-axis; Alternatively, an RT characteristic curve can be plotted with temperature on the x-axis and resistance on the y-axis. The curve is dynamically refreshed as experimental data is updated, or it is displayed statically after the experiment ends.
[0015] Preferably, the microcontroller main control module is an STM32 series microcontroller; The display module uses an OLED display screen; The communication module used is the ESP8266 WiFi module.
[0016] Preferably, the system also includes a mobile display subsystem; The mobile terminal display subsystem runs on a mobile device and connects to the communication module or the host computer processing platform either as a TCP client or via a local area network. The mobile display subsystem is configured to receive experimental data in real time and display voltage, current, resistance, power, and temperature values on the mobile device screen, but it does not have AI analysis capabilities.
[0017] Preferably, the power supply and interface module further includes a protection circuit; The protection circuit includes a voltage regulator chip connected to the power input terminal and a resettable fuse or current-limiting resistor connected to the external experimental circuit interface to prevent short circuits in the experimental circuit from damaging the system.
[0018] This invention provides an intelligent measurement and AI analysis system for electrical experiments in middle schools. It offers the following advantages: 1. This invention sets up two complementary analysis modes on the host computer processing platform: an online large model and an offline local algorithm. It can automatically switch the processing logic according to the network environment and API key status, effectively solving the problem of insufficient stability caused by the excessive reliance of existing intelligent teaching equipment on cloud resources, and significantly improving the system's adaptability and robustness in different teaching environments.
[0019] 2. This invention integrates voltage and current acquisition modules and temperature acquisition modules synchronously in the hardware measurement terminal, and establishes a time-series correlation between temperature data and electrical data in the data processing flow. This allows for a direct demonstration and quantitative analysis of the impact of temperature changes on resistance characteristics, overcoming the limitation of traditional middle school electrical experiment instruments that only focus on electrical parameters and ignore the Joule heating effect. This helps students build a more complete physics knowledge system.
[0020] 3. This invention logically isolates the practice diagnosis module from the experimental data acquisition / visualization / AI analysis module, and adopts a non-generative assessment strategy of "teacher-preset question bank + automatic program judgment" to ensure the authority and consistency of the question content and standard answers. At the same time, without generating questions or modifying standard answers, it introduces large models or offline analysis to diagnose the answer records and knowledge mastery and provide personalized learning suggestions. This avoids the illusion of large models interfering with the assessment content, significantly reduces the cost of teacher grading and commenting, and improves the efficiency of the classroom practice and assessment closed loop and the safety of teaching.
[0021] 4. This invention performs edge computing (local real-time calculation of resistance R and power P) through the microcontroller main control module and combines it with the OLED display module to refresh multiple lines of text, enabling the hardware measurement terminal to have independent data processing and display capabilities. Even without connecting to a host computer or mobile device, it can still be used as a high-precision, multi-functional smart digital meter, which not only meets the portability requirements for rapid measurement, but also reduces the computational burden on the communication link and the host computer, and realizes zero-delay direct reading of experimental data.
[0022] 5. This invention, through the collaborative work of the data visualization module of the host computer software and the mobile terminal display subsystem, can transform abstract and discrete experimental measurement data into intuitive IU characteristic curves or RT change curves in real time. At the same time, it supports the distribution of data streams to mobile terminals via WiFi local area network, which makes it convenient for teachers to monitor the measurement status of each experimental group in real time when they move around the classroom. This reduces the cognitive threshold for students to understand the physical function relationships and improves the classroom efficiency and interactive experience of experimental teaching. Attached Figure Description
[0023] Figure 1 This is one of the system flow diagrams of the present invention; Figure 2 This is the second schematic diagram of the system flow of the present invention; Figure 3 This is the third system flow diagram of the present invention; Figure 4 This is the fourth system flow diagram of the present invention; Figure 5 This is the fifth system flow diagram of the present invention; Figure 6 This is the sixth schematic diagram of the system flow of the present invention; Figure 7 This is a schematic diagram of the hardware principle of the present invention; Figure 8 This is a schematic diagram of the computer-side host computer software interface of the present invention; Figure 9 This is a schematic diagram of the virtual needle meter reading of the present invention; Figure 10 This is one of the schematic diagrams of the preset exercise question window interface of the present invention; Figure 11 This is a schematic diagram of the mobile application interface of the present invention; Figure 12 This is a schematic diagram of the PCB component layout of the main control board of the present invention; Figure 13 This is the second schematic diagram of the preset practice question window interface of the present invention. Detailed Implementation
[0024] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0025] Please see the appendix Figure 1 -Appendix Figure 13 This invention provides an intelligent measurement and AI analysis system for middle school electrical experiments, including a hardware measurement terminal and a host computer processing platform. The specific circuit structure of the hardware measurement terminal is as follows: the voltage and current acquisition module adopts a sensor module, which is connected to the experimental circuit in series or parallel, and is connected via a serial port or interface. The temperature acquisition module uses a single-bus digital temperature sensor, with its probe in close contact with the outer surface of the load being measured. The communication module uses either a WiFi or Bluetooth wireless communication module, connected to the microcontroller main control module via a serial port, and sends measurement datasets containing timestamps using a text protocol. The microcontroller main control module is an STM32 series microcontroller. The display module uses an OLED display screen. The communication module uses an ESP8266 WiFi module. The hardware measurement terminal includes: The power supply and interface module is used to provide the operating voltage for the system. The power supply and interface module also includes a protection circuit. The protection circuit includes a voltage regulator chip connected to the power input terminal and a resettable fuse or current limiting resistor connected to the external experimental circuit interface to prevent short circuits in the experimental circuit from damaging the system. The microcontroller main control module, as the control core of the system, has the following local calculation logic: the microcontroller main control module periodically acquires voltage data. and current data Calculate the resistance value according to Ohm's law Calculate the power value according to the electric power formula. The hardware measurement terminal also includes a display module, which is connected to the microcontroller main control module via a communication interface. The microcontroller main control module is configured to display the real-time acquired voltage data. Current data Temperature data and the calculated resistance value and power value The text is refreshed and displayed on the display module in the form of multi-line text. Specifically, this module can perform edge computing on the system and complete the processing of basic physical quantities without the involvement of a host computer.
[0026] Data acquisition timing: The microcontroller sets a timer interrupt (e.g., triggering every 200ms). In the interrupt service routine, data read commands are triggered sequentially, sending read commands to the voltage and current modules to obtain the raw voltage data across the load and the raw current data flowing through the load. Reset the single bus and send temperature conversion and read commands to the DS18B20 to obtain the ambient or device temperature. ; Physical quantity calculation: After filtering the raw data (such as by moving average filtering), the microcontroller performs the calculation based on the physical formula: Resistance calculation: ,like If it is 0, then mark It is infinite; Power calculation At this point, a complete data frame is formed in the system memory. ; Local real-time display: A 0.96-inch OLED display is driven via an I²C interface; the microcontroller calculates... The numerical value is converted into a string, and the display driver library is called to refresh the screen in real time as multi-line text. This allows students to use it as a high-precision digital meter without connecting it to a computer.
[0027] The voltage and current acquisition module is connected to the load under test and is used to acquire the voltage and current data of the load. The temperature acquisition module is installed on the surface of the component being measured and is used to acquire temperature data. And a communication module, used to establish a data transmission link between the hardware measurement terminal and the host computer processing platform; The microcontroller main control module is connected to the voltage and current acquisition module, the temperature acquisition module and the communication module respectively. It is used to receive the acquired data and perform local calculations, and send the data to the host computer processing platform through the communication module. Specifically, the hardware of the system mainly consists of hardware measurement terminals, and its core is to build a stable and secure experimental data acquisition platform. Power and Interface Module Configuration: The system connects to a standard 5V power supply via a Type-C or USB interface. The input terminal is equipped with a power management circuit that uses a low-dropout linear regulator (LDO, such as AMS1117-3.3) to convert the 5V voltage to 3.3V to power the microcontroller and peripheral sensors. To ensure experimental safety, a self-resetting fuse is connected in series at the power input terminal, which automatically cuts off the circuit in case of a short circuit or excessive current in the external circuit to protect the internal chip. At the same time, the system provides external wiring terminals (such as screw-type terminals) for connecting external loads under test (such as light bulbs, fixed resistors, sliding rheostats, etc.) and experimental power supplies.
[0028] Microcontroller main control module startup: The STM32F103C8T6 microcontroller is used as the main control core. After the system is powered on, the main control module first performs clock tree configuration, GPIO port initialization, interrupt priority setting and communication bus (USART, I²C) reset.
[0029] Sensor and communication module connection: Voltage and current acquisition: SUI-201 or similar voltage and current sensor modules are used. The measuring end is connected to the external wiring terminal, and the data communication end is connected to the microcontroller through UART or I²C interface. Temperature acquisition: The DS18B20 digital temperature sensor is used and is connected to the GPIO pin of the microcontroller via a single-wire protocol. The sensor probe is designed to be movable or surface-mount, which is convenient for close contact with the component being measured (such as the surface of a resistor or the glass shell of a bulb) to obtain real-time temperature. Communication module: The ESP8266 WiFi module is used, which is connected to the microcontroller through the USART3 interface and configured in Station mode to connect to the lab router or hotspot.
[0030] The host computer processing platform runs host computer software, which includes: The data communication and storage module is used to receive and store experimental data from the hardware measurement terminal; The data visualization module is used to plot physical characteristic curves based on experimental data. Specific functions of the data visualization module include: reading stored experimental data and plotting an IU characteristic curve with voltage as the x-axis and current as the y-axis; or plotting an RT characteristic curve with temperature as the x-axis and resistance as the y-axis; the curves are dynamically refreshed as experimental data is updated or statically displayed after the experiment; and an AI analysis module is used for intelligent analysis of the experimental data; the AI analysis module is configured to have two working modes: Online large model analysis mode: When a normal network connection is detected and a valid API key is available, the remote large model interface is invoked. Based on preset prompts, the experimental data is analyzed for physical laws. The specific execution steps of the online large model analysis mode of the host computer processing platform include: constructing prompts containing experimental data, physical experimental background, and analysis requirements; sending the prompts to the remote large model server via an HTTP request; receiving natural language text returned by the server, which contains qualitative analysis of resistance change trends, power relationships, and possible sources of experimental error; and displaying the natural language text in the analysis results area of the host computer software. Offline Local Analysis Mode: When the remote large model interface cannot be called, a local offline analysis algorithm is invoked to perform statistical calculations and linear fitting analysis on the experimental data. The specific execution steps of the offline local analysis mode of the host computer processing platform include: calculating the average resistance and standard deviation of each measurement point in the selected experimental dataset; judging the resistance stability based on the standard deviation; performing linear fitting between the resistance value and the temperature value; judging the trend of resistance change with temperature; calculating the ratio of voltage to current; judging whether the data conforms to Ohm's law; and detecting whether the power exceeds the preset rated power threshold, and generating an overload risk warning when the threshold is exceeded. The host computer software also includes a practice diagnosis module; the practice diagnosis module runs in an interface area independent of the AI analysis module; the practice diagnosis module loads a preset static question bank, which contains pre-set question text, options, and standard answers; the system is configured such that the content displayed by the practice diagnosis module does not change with the changes in experimental data or the output results of the AI analysis module, and does not call the large model interface to generate questions or modify standard answers. Specifically, the practice and diagnostic module is used to consolidate experimental knowledge and form a closed loop of "teacher pre-made exercises - student answers - program judgment and recording - AI personalized suggestions". Its feature is that content isolation and diagnostic enhancement coexist, that is, the questions and standard answers are not affected by AI, while AI only learns, diagnoses and generates suggestions based on the answer records. To ensure the objectivity and consistency of the grading results, automatic grading is completed by rule comparison and error tolerance judgment strategy, and the student's answer, standard answer, right or wrong result and corresponding knowledge point identification for each sub-item are written into the answer record set; the AI analysis module only performs learning diagnosis and suggestion generation on the answer record set, does not participate in right or wrong judgment, and does not write back or overwrite the grading results.
[0031] In terms of question bank management, question bank files and question illustrations are preferably stored in a local directory and maintained offline by teachers. The host computer software can still complete question bank loading, answer collection, automatic question judging, and knowledge point mastery statistics even without a network connection. A network connection is only used to obtain natural language diagnostic suggestions when the online large model analysis mode is enabled.
[0032] Furthermore, the AI analysis module can simultaneously receive experimental data analysis results and practice diagnostic results, generating a comprehensive learning report. This comprehensive learning report includes conclusions about the patterns in the experimental data, a list of weak knowledge points, typical error examples, and suggestions for consolidating weak knowledge points. It also supports exporting the report as a structured file for teacher tracking and analysis.
[0033] Question bank organization: Teachers compile the question bank based on Word documents or structured files (such as JSON / XML). Each question or sub-question should contain at least the following fields: TYPE, KNOWLEDGE POINTS, QUESTION, OPTIONS (for multiple-choice questions), and ANSWER. Furthermore, the KNOWLEDGE POINTS field can contain multiple knowledge point identifiers, arranged in the order of "per question / per blank"; the ANSWER field can contain multiple sub-answers with the same number of knowledge point identifiers, and the sub-answers are organized sequentially using commas, semicolons, vertical bars (|), or line breaks; student-submitted answers are also parsed into multiple student sub-answers according to the same separation rules. The host computer software maps the k-th knowledge point identifier to the k-th standard sub-answer and the k-th student sub-answer, generating a sub-item-level answer record (question number, sub-item number, corresponding knowledge point, sub-answer correct / incorrect, student sub-answer, standard sub-answer, timestamp), thereby realizing fine-grained diagnosis and personalized suggestion generation for knowledge points.
[0034] Image association: Static images related to the question are stored in the local data directory. The image files are bound to the question number according to the preset naming rules. The host computer software automatically matches and displays them when loading the question, thus achieving a consistent presentation of "question stem + illustration". Answer Collection and Grading: Students input their answers on the practice interface (selection letters for multiple-choice questions, text or numerical values for fill-in-the-blank / calculation / short-answer questions). The system automatically grades the answers based on the standard answer. When a sub-answer can be parsed into a numerical value, a numerical comparison strategy is preferred, and it is judged as correct if it meets the preset absolute error threshold or relative error threshold (e.g., relative error ≤ 5% or absolute error ≤ 0.05). When a sub-answer is a formula, unit, or text description, the text is preferably normalized (removing spaces, unifying capitalization, and replacing equivalent symbols) before matching and judging. The system writes the grading result of each sub-item and the corresponding knowledge point identifier into the answer record set, generating an answer record set containing the question number / small question mark (or sub-item number), correct / incorrect result, student answer, standard answer, and corresponding knowledge point identifier, and supports switching between previous / next questions and showing / hiding answers.
[0035] Knowledge point statistics and prompt word construction: The system summarizes the answer record set according to the knowledge point dimension, forms a structured summary of "knowledge point - number of correct / incorrect times - typical errors", and fills the summary into the preset prompt word template; Mastery calculation and sorting: The system calculates the mastery of knowledge points based on the number of correct and incorrect answers in the answer record set (e.g., mastery = number of correct answers / (number of correct answers + number of incorrect answers) or its weighted function), and generates a list of weak knowledge points by sorting them from low to high mastery. Error collection and re-practice mechanism: The system selects incorrect questions from the answer record set to form an error collection, and can retrieve corresponding questions from the question bank according to the weak knowledge points to generate a re-practice set; after the student completes the re-practice, the answer record set and mastery level are updated until the mastery level reaches the threshold or the maximum number of re-practice rounds are reached, thus forming a traceable closed-loop consolidation process; Privacy and security constraints: When using the online large model analysis mode to generate personalized suggestions, the system only uploads structured summaries and statistical results, and does not upload student identity information; the prompt word template includes constraints such as not generating new questions, not directly outputting the original question's standard answer, not modifying the question bank and standard answers, word limit, and language switching, in order to improve the security and traceability of classroom assessments.
[0036] Personalized suggestion output: After clicking "AI Analysis", the system calls the remote large model interface or offline analysis strategy to process the prompt words and outputs personalized learning suggestions including the overall accuracy, a list of weak knowledge points and training and error correction suggestions for the weak knowledge points, which are displayed on the practice interface. It should be noted that the practice diagnostic module is configured not to use large models to generate questions or modify standard answers, thereby ensuring the rigor and traceability of the assessment.
[0037] The system also includes a mobile display subsystem; The mobile terminal display subsystem runs on a mobile device and connects to the communication module or the host computer processing platform either as a TCP client or via a local area network. The mobile display subsystem is configured to receive experimental data in real time and display voltage, current, resistance, power and temperature values on the mobile device screen, but it does not have AI analysis capabilities. Specifically, this module realizes the transformation from single-point measurement to continuous data stream and supports collaborative work between PC and mobile devices; Data Packaging and Transmission: The microcontroller packages a set of data {U,I,R,P,T} at the current moment according to a specific text protocol (such as JSON format or comma-separated CSV format) and sends it to the ESP8266 module via serial port. The ESP8266 uses TCP / UDP protocol to pass the data packet through to the specified port in the local area network. Host computer processing platform (PC): Python host computer software running on the teacher's or laboratory's computer listens on a specified port; Parsing and Storage: Upon receiving a data frame, parse each physical quantity, add a high-precision system timestamp, and store it in a memory list or local database (e.g., ...). document); Dynamic plotting: The data visualization module calls plotting libraries (such as...) With time t or voltage U as the horizontal axis, the I-U characteristic curve, PU curve or RT change curve are plotted in real time. The curve is dynamically refreshed as new data arrives, intuitively displaying the physical laws. Mobile monitoring (phone): The mobile application (based on frameworks such as Kivy) acts as a TCP client to connect to the system. It is only responsible for receiving data streams and updating the numerical display list on the interface, which makes it convenient for teachers to check the experimental status of students in each group when they are patrolling the classroom. However, it does not perform complex curve fitting or AI analysis in order to keep the software lightweight.
[0038] The user selects a segment of collected experimental data (e.g., a set of data from the "light bulb volt-ampere characteristic curve") in the host computer software, clicks the intelligent analysis button, and the system automatically switches between the following two modes based on the current network environment and configuration: Mode 1: Online Large Model Analysis Mode This mode is executed when the system detects that the internet connection is normal and a valid large model APIKey is configured; Prompt word construction: The host computer software integrates a prompt word template specifically for physics experiments. The program fills the selected experimental data (such as a JSON format sequence of voltage U, current I, temperature T, resistance R, and power P) into the template along with the experimental background and analysis task to form prompt words; for example: "Based on the following light bulb experimental data, please analyze the reasons for the nonlinearity of the IU curve, discuss the trend of resistance change in combination with temperature changes, and point out possible sources of error." API calls: Calling remote large model interfaces (such as GPT, Wenxin Yiyan, etc.) via HTTP POST requests; Result analysis: Receive the natural language text returned by the large model. This text usually contains a qualitative description of the nonlinear features (such as as the voltage increases, the filament temperature increases, which leads to an increase in resistivity, so the IU curve shows a downward bending trend). Display: The analysis report will be displayed in the software's "AI Teaching Assistant" window.
[0039] Mode 2: Offline Local Analysis Mode
[0040] When the network is disconnected or the APIKey is not configured, the system automatically falls back to the local algorithm module, which does not depend on any external services, ensuring the continuity of teaching. Statistical analysis: Calculating the average resistance of the data points and standard deviation ,like If the value is less than a preset threshold, it is determined to be a fixed resistor; otherwise, it is determined to be a variable resistor. Linear / nonlinear fitting: for Perform linear regression on the data and calculate the correlation coefficient. ,like If the value is close to 1, the output will conform to Ohm's law. To determine the trend of the data, if the slope The output shows "resistance increases with increasing temperature". Threshold discrimination: Examining the power of any data point Does it exceed the preset rated power? If the limit is exceeded, a "risk of overload" warning will be generated. Error evaluation: Calculate the jump amplitude between adjacent data points. If there is a sudden change, output the evaluation "Possible poor contact or unstable reading". Result Synthesis: The above calculation results are filled into a preset sentence template to generate standardized analysis text.
[0041] Through the above implementation methods, this invention digitizes and upgrades traditional middle school electrical experiments. On the hardware side, the combination of STM32 and high-precision sensors enables simultaneous measurement and display of five dimensions: U, I, R, P, and T. On the software side, a unique "online + offline" dual-mode analysis engine provides in-depth physics teaching assistance by utilizing the natural language understanding capabilities of large models, while ensuring basic teaching functions in offline environments through local algorithms. Meanwhile, the independent exercise module forms a complete "test-practice-evaluation" teaching loop.
[0042] In summary, this invention provides an intelligent measurement and AI analysis system for middle school electrical experiments. By setting up two complementary analysis modes on the host computer processing platform—an online large model and an offline local algorithm—it can automatically switch processing logic according to the network environment and API key status. When the network is good, it can use the large model to provide in-depth, natural language-based analysis of physical laws and errors; while in laboratory environments with abnormal networks or no external network access, it can ensure the availability of basic statistical and fitting functions through local algorithms. This effectively solves the problem of insufficient stability caused by the excessive reliance on cloud resources in existing intelligent teaching equipment, significantly improving the system's adaptability and robustness in different teaching environments. Furthermore, by synchronously integrating voltage, current, and temperature acquisition modules in the hardware measurement terminal and establishing a temporal correlation between temperature data and electrical data in the data processing flow, the system can simultaneously acquire the electrical characteristics (U, I, R, P) and thermal characteristics (T) of the load. This allows for a direct demonstration and quantitative analysis of the impact of temperature changes on resistance characteristics (e.g., verifying the nonlinear law of the increase in the resistance of a light bulb filament with increasing temperature). This overcomes the limitation of traditional middle school electrical experiment instruments that only focus on electrical parameters and ignore the Joule heating effect, helping students build a more complete physics knowledge system.
[0043] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.
Claims
1. An intelligent measurement and AI analysis system for middle school electrical experiments, characterized in that, This includes hardware measurement terminals and a host computer processing platform; The hardware measurement terminal includes: a power supply and interface module, a microcontroller main control module, a voltage and current acquisition module, a temperature acquisition module, and a communication module; The microcontroller main control module is connected to the voltage and current acquisition module, the temperature acquisition module and the communication module respectively. It is used to receive the acquired data and perform local calculations, and send the data to the host computer processing platform through the communication module. The host computer processing platform runs host computer software, which includes a data communication and storage module, a data visualization module, and an AI analysis module. The AI analysis module is configured with two working modes: Online large model analysis mode: When a normal network connection is detected and a valid API key is available, the remote large model interface is invoked to perform physical law analysis on the experimental data based on preset prompts. Offline local analysis mode: When the remote large model interface cannot be called, the local offline analysis algorithm is called to perform statistical calculations and linear fitting analysis on the experimental data.
2. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The local computing logic of the microcontroller main control module is as follows: The microcontroller's main control module periodically acquires voltage and current data, calculates the resistance value according to Ohm's law, and calculates the power value according to the power formula. The hardware measurement terminal also includes a display module, which is connected to the microcontroller main control module through a communication interface. The microcontroller main control module is configured to refresh and display the real-time collected voltage data, current data, temperature data, and calculated resistance and power values on the display module in the form of multi-line text.
3. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The specific circuit structure of the hardware measurement terminal is as follows: The voltage and current acquisition module adopts a sensor module, which is connected to the experimental circuit in series or parallel, and transmits data to the microcontroller main control module through a serial port or interface. The temperature acquisition module uses a single-bus digital temperature sensor, whose probe is in close contact with the outer surface of the load being measured. The communication module uses at least one of serial communication or wireless communication to send a measurement dataset containing timestamps using a text protocol.
4. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The specific execution steps of the offline local analysis mode include: Calculate the average resistance and standard deviation of each measurement point in the selected experimental dataset, and determine the resistance stability based on the standard deviation; Linear fitting is performed on the resistance value and temperature value to determine the trend of resistance change with temperature; Calculate the ratio of voltage to current and determine whether the data conforms to Ohm's law; It detects whether the electrical power exceeds the preset rated power threshold and generates an overload risk warning when the threshold is exceeded.
5. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The specific execution steps of the online large model analysis mode include: Construct prompts that include experimental data, physical experimental background, and analysis requirements; The prompt word is sent to the remote large model server via an HTTP request; Receive natural language text returned by the server, the text containing a qualitative analysis of the resistance change trend, power relationship and possible sources of experimental error; The natural language text is displayed in the analysis results area of the host computer software.
6. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The host computer software also includes a practice diagnostic module; The practice diagnosis module includes a question bank management submodule, a response collection submodule, an automatic question judging submodule, a knowledge point mapping submodule, and a personalized suggestion generation submodule; The question bank management submodule is used to load local question bank files pre-compiled by teachers. The question bank files are Word documents or structured files, in which each question or sub-question is configured with question type fields, knowledge point fields and standard answer fields, and allows the association of static images; The automatic question-judging submodule is used to compare students' answers with standard answers and generate a set of answer records that includes question number / small question mark (or sub-item number), correct or incorrect result and related knowledge points; The personalized suggestion generation submodule is used to summarize the answer record set by knowledge point and construct prompt words, call the remote large model interface or offline analysis strategy to generate personalized learning suggestions and display them; wherein the automatic question judgment result is generated by rule comparison and error tolerance judgment strategy, and AI is only used to diagnose and generate suggestions for the answer record, and does not participate in the right and wrong judgment, generate questions, or modify the question bank and standard answers. The question bank files and question illustrations are stored in a local directory and maintained offline by the teacher. The host computer software can still complete the loading of the question bank, collection of answers, automatic question judging, and statistics on the mastery of knowledge points even without a network connection; a network connection is only used when the online large model analysis mode is enabled.
7. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 1, characterized in that, The specific functions of the data visualization module include: Read the stored experimental data and plot the IU characteristic curve with voltage as the abscissa and current as the ordinate; or plot the RT characteristic curve with temperature as the abscissa and resistance as the ordinate; the curve is dynamically refreshed as the experimental data is updated or statically displayed after the experiment is completed.
8. A method for practice diagnosis and personalized suggestion generation, characterized in that, The intelligent measurement and AI analysis system for middle school electrical experiments according to any one of claims 1-7 includes: Teachers pre-set standard answers and configure corresponding knowledge point identifiers for each question or sub-question in the question bank file; Students input their answers on the practice interface of the host computer software; The host computer software automatically judges the questions based on the standard answers, forming a set of answer records that includes the question number / small question mark (or sub-item number), the correct or incorrect result, and the knowledge point identifier; The answer record set is statistically summarized according to the knowledge point dimension, prompt words are constructed, and remote large model interface or offline analysis strategy is called to generate personalized learning suggestions and output them on the practice interface.
9. The method for practice diagnosis and personalized suggestion generation according to claim 8, characterized in that, The question bank file is a Word document containing the fields TYPE (question type), KNOWLEDGE POINTS (knowledge point), QUESTION (question stem), OPTIONS (options), and ANSWER (answer). Static images associated with the questions are stored in the local directory according to preset naming rules, and the host computer software automatically matches and loads them for display based on the question number.
10. The method for practice diagnosis and personalized suggestion generation according to claim 8, characterized in that, The personalized learning recommendations include at least the overall accuracy rate, a list of weak knowledge points, and training suggestions for those weak knowledge points.
11. The method for practice diagnosis and personalized suggestion generation according to claim 8, characterized in that, The answer record set further includes student answers, standard answers, question type identifiers, grading timestamps, and exercise numbers, and is stored locally as a structured data file, supporting export as JSON, CSV, or tabular files.
12. The intelligent measurement and AI analysis system for middle school electrical experiments according to claim 6, characterized in that, The knowledge point mapping submodule supports associating multiple knowledge point identifiers with the same sub-question and calculating the knowledge point mastery level according to preset weights. The knowledge point mastery level is the ratio of the correct number of times the corresponding sub-question is answered to the total number of times or its weighted function.
13. The method for practice diagnosis and personalized suggestion generation according to claim 8, characterized in that, In the teacher-prepared practice question template, the KNOWLEDGE POINTS field contains multiple knowledge point identifiers arranged in the order of sub-questions / blanks, and the ANSWER field contains multiple standard answers corresponding to the order of these knowledge points. When grading questions, the host computer software breaks down student answers into multiple sub-answers and compares them with corresponding standard answers. It generates sub-item-level answer records that include question number, sub-item number, corresponding knowledge point identifier, whether the sub-answer is correct or incorrect, student sub-answer, and standard sub-answer. Based on the sub-item-level answer records, it performs statistical summarization by knowledge point dimension to generate the personalized learning suggestions.
14. The method for practice diagnosis and personalized suggestion generation according to claim 13, characterized in that, Both the standard answer field ANSWER and the student answer field are split into multiple sub-answers using preset separators. The preset separators include at least commas, semicolons, vertical bars (|), or line breaks. When the number of sub-answers is inconsistent, the missing sub-answer is treated as an empty string or matched by extending the last item, and the exception is recorded in the answer record set for diagnosis.
15. The method for practice diagnosis and personalized suggestion generation according to claim 14, characterized in that, For sub-answers that can be parsed into numerical values, an error tolerance judgment strategy is adopted: when the absolute error between the student's sub-answer and the standard sub-answer does not exceed a preset absolute threshold or the relative error does not exceed a preset relative threshold, it is judged as correct; The threshold can be configured according to the question type or knowledge point; Furthermore, after generating the correct / incorrect judgment result, the automatic judgment submodule writes the correct / incorrect judgment result into the answer record set in a read-only manner. The AI analysis module only reads the answer record set to generate learning suggestions, without writing back or overwriting the correct / incorrect judgment result.