AI intelligent agent-based astronomical science popularization telescope realization method and AI intelligent agent-based astronomical science popularization telescope realization system

Abstract

The invention provides an astronomical science popularization telescope implementation method and system based on an AI intelligent agent, and belongs to the field of science popularization education. The implementation method comprises the following steps: converting a physical operation signal of a telescope into a digital signal to synchronously drive a digital star map picture to carry out view angle and zoom adjustment, and calling a star map application module to obtain a celestial body identifier, so as to realize real-time linkage observation of the physical pointing of the telescope and the display content of the digital star map. One-way knowledge infusion is converted into two-way real-time conversation; according to the real-time state of a physical operation signal, stable hovering detection is carried out to judge the intention of a user, a target celestial body identifier is screened out, an AI interaction module with a context awareness capability is constructed, personalized science popularization content is generated through retrieval enhancement generation and a user portrait recognition technology, and the user experience is improved. The technical problem that existing popular science content cannot be personalized and situational is solved, closed-loop interactive experience of operation-exploration and observation-learning is achieved, and the threshold of scientific exploration is lowered.

Description

Technical Field

[0001] This invention belongs to the field of popular science education, specifically relating to a method and system for implementing an astronomical popular science telescope based on an AI intelligent agent. Background Technology

[0002] As a fundamental natural science, astronomy plays an irreplaceable role in cultivating the scientific literacy, logical thinking, and exploratory spirit of the public, especially young people. Therefore, the research and application of astronomical science popularization equipment and tools has always been an important topic in the field of educational technology. However, current astronomical science popularization programs still have significant limitations and fail to fully meet the demands of modern science education for interactivity, engagement, and personalization.

[0003] Traditional optical telescopes rely entirely on manual operation, requiring users to possess extensive knowledge of star charts and star-finding skills to locate celestial objects. This high barrier to entry makes them unsuitable for beginners and teenagers to use independently. To address the difficulty of star finding, numerous electric telescopes have emerged on the market. For example, CN105896359A discloses an astronomical telescope device for automatic high-altitude star finding observation by UAVs, utilizing a motor-gear-mechanical telescopic mechanism for automatic telescope extension and retrieval. However, this technology suffers from three main drawbacks: First, the interaction method is rigid and unintuitive. Interaction still relies on button-based, menu-driven controllers, resulting in a poor user experience. Second, information presentation is limited. The controller screen typically only displays basic parameters such as the name and coordinates of celestial objects, failing to provide rich background knowledge, scientific principles, or related cultural stories, creating a disconnect between knowledge acquisition and observation. Third, there is a lack of proactive learning guidance. The device passively executes commands and cannot provide intelligent guidance or heuristic questioning based on the user's observation behavior, resulting in extremely weak interactivity.

[0004] With the development of mobile internet technology, solutions combining telescopes with smartphones or tablets have emerged. For example, some commercially available smart telescopes use built-in cameras to photograph the night sky, automatically identifying the currently pointed celestial region using "field-of-view calculation" technology, and transmitting the images to a mobile app in real time. Users control and view the telescope through the app. However, this technology has several drawbacks: First, the lack of physical interaction: Although the app interface is more user-friendly, the user's focus shifts entirely to the touchscreen, losing the realism and immersive experience of personally operating and rotating a physical telescope for exploration. The telescope itself becomes a passive "gimbal camera." Second, the passivity of information interaction: The information provided by the app is usually still presented in a database format, requiring users to actively click to view. It cannot achieve a dynamic, real-time, and conversational knowledge interaction like "watching and talking." Users cannot ask questions such as "What is this?" or "Why is it like this?" about the observed scenes, as they would when communicating with a real instructor. Third, insufficient content adaptability: The provided science content is usually "one-size-fits-all," unable to be personalized or differentiated based on the user's age, knowledge background, etc.

[0005] Other technological solutions utilize AR / VR technology to provide astronomical science popularization experiences. For example, AR star map apps running on smartphones can overlay virtual constellation patterns and celestial information onto real night sky images. VR applications, on the other hand, can create a fully immersive virtual universe. For instance, CN222087916U discloses an AR smart sightseeing telescope system that integrates camera, gyroscope positioning, cloud control, and AR display technology, enabling interactive viewing and entertainment experiences. The drawback of this type of technology is that it is entirely digital and detached from physical devices, lacking interaction with a real, tangible physical device (telescope), and thus failing to provide the scientific practical experience of "exploring the unknown through instruments." The learning process lacks a hands-on component.

[0006] In recent years, AI educational robots or software based on large language models have begun to emerge. For example, CN120523919A discloses a dynamic text response method and system for an educational question-and-answer robot, which can answer questions in specific subject areas and dynamically adjust the guidance content based on the student's specific learning situation or response feedback. However, the drawback of this technology is that it lacks context awareness; the AI ​​system is disconnected from the user's physical world. Therefore, it cannot bind knowledge explanation with the user's real-time observation behavior, and the learning experience is non-contextualized and abstract.

[0007] In summary, a significant technological gap exists in existing technologies: no solution effectively integrates "intuitive physical device operation," "real-time digital content feedback," and "context-aware AI intelligent dialogue." Specifically, the current market and patent literature generally lack a device that allows users (especially teenagers) to explore the digital night sky through simple physical actions (such as rotating or adjusting knobs), and that can proactively identify the user's observation target, triggering an AI agent capable of multi-turn dialogue, telling scientific and humanistic stories, and adjusting content based on the user's age. Therefore, developing a new type of astronomical science telescope that is easy to operate, integrates physical exploration and intelligent interaction, and can provide personalized and narrative science popularization content to address the pain points of existing technologies has significant practical importance and application value. Summary of the Invention

[0008] To address the problems of high operational barriers, weak human-computer interaction, and limited and personalized popular science content in existing astronomical popular science equipment, this invention provides a method and system for implementing an astronomical popular science telescope based on an AI agent. By deeply integrating intuitive physical telescope operation, real-time linked digital star maps, and an AI agent with context awareness capabilities, the astronomical telescope is transformed from a simple optical observation instrument into an intelligent popular science terminal that can understand user intentions, link with the digital world in real time, and proactively provide personalized and interactive popular science services.

[0009] To achieve the above-mentioned objectives, this invention provides a method for implementing an astronomical science telescope based on an AI agent, comprising the following steps: Step 1: Collect physical operation signals from the user's use of the telescope in real time and convert the physical operation signals into digital signals; Step 2: Analyze and optimize the digital signal to synchronously drive the digital star map image to adjust the viewing angle and zoom. At the same time, call the star map AI application module to obtain the celestial object identifiers corresponding to the center area of ​​the digital star map image, so as to realize the real-time linkage between the physical pointing of the telescope and the update of the digital star map image. Step 3: While the digital star map is being updated synchronously, stable hovering detection is performed based on the real-time status of the physical operation signal to determine the user's intention. Based on the user's intention, the target celestial body identifier is selected from the celestial body identifiers, and the AI ​​interaction module is triggered. Step 4: Based on the user profile information and target celestial body identifier identified at the time of triggering, the AI ​​interaction module retrieves astronomical knowledge from the knowledge base and dynamically constructs prompt words to generate personalized science popularization content. Then, the personalized science popularization content is converted into voice output and subsequent voice questions from the user are received to form an interactive closed loop and complete the science popularization explanation.

[0010] Preferably, step 1 specifically includes: Azimuth adjustment signals are acquired by sensors installed on the azimuth and pitch axes of the telescope, and zoom control signals are acquired by a knob encoder installed on the telescope barrel. The azimuth adjustment signals and zoom control signals serve as physical operation signals. The acquired azimuth adjustment signal and zoom control signal are converted into standardized azimuth angle, pitch angle and zoom increment value; The standardized azimuth, elevation, and zoom increment values ​​are packaged into a formatted string as a digital signal and sent to the central processing unit via serial communication.

[0011] More preferably, when acquiring the azimuth adjustment signal through the sensor, the analog or digital values ​​of the sensors on the azimuth and pitch axes are read cyclically using millisecond-level high-frequency readings; wherein, the sensor includes an angle sensor and an inertial measurement unit; When the angle sensor is a potentiometer, the change in voltage corresponding to the change in resistance is used to map the change in attitude rotation angle; when the sensor is an encoder, the AB phase pulse signals are counted, with the pulse count value corresponding to the rotation angle and the pulse transition direction corresponding to the rotation direction. The inertial measurement unit is used to synchronously acquire triaxial acceleration and triaxial angular velocity, and to perform fusion processing on the triaxial acceleration and triaxial angular velocity acquired by the inertial measurement unit.

[0012] More preferably, the fusion process includes: Zero-point calibration is performed on the gyroscope of the inertial measurement unit by collecting multiple sets of data while the telescope is stationary and taking the average value as the gyroscope data to eliminate zero drift error. The three-axis acceleration, three-axis angular velocity and gyroscope data are fused by complementary filtering algorithm to output standardized azimuth and pitch angles, thus completing the digital conversion of attitude change.

[0013] More preferably, the scaling control signal is acquired via a rotary encoder, including: The rotation direction is identified by detecting the level transition sequence of the AB phase pulse signals of the rotary encoder to match the user's observation and operation habits; and an anti-vibration threshold is set to filter mechanical vibration noise. The rotation speed is determined by the time interval between two pulses to achieve variable speed scaling, and the scaling increment step is dynamically set accordingly to achieve variable speed scaling control to adjust the user's observation magnification.

[0014] Preferably, step 2 specifically includes: Digital signals from the telescope are received at millisecond-level frequency via serial communication, and the horizontal and vertical angle displacements and image scaling increments are analyzed and extracted. Smoothing filtering is applied to the horizontal and vertical angle displacements to eliminate numerical jumps caused by physical jitter in the telescope. The azimuth and elevation angle parameters after smoothing and filtering are called into the view alignment interface of the star map AI application module to synchronize the physical pointing of the telescope with the celestial object markings updated in the digital star map.

[0015] More preferably, scaling adjustment is achieved by optimizing the image scaling increment, including: Based on the analyzed image scaling increment ,pass Calculate the new field of view ,in, The original field of view. With a fixed scaling factor, when the image scaling increment is positive, the field of view decreases, and the corresponding digital star map image is enlarged for observing celestial details; when the image scaling increment is negative, the field of view increases, and the corresponding digital star map image is reduced for observing a wide area of ​​the night sky. The image scaling increment is processed in stages, and boundary constraints are set for the field of view to match the actual magnification range of astronomical observations, so as to achieve scaling adjustment. The staged processing includes: when the knob is rotated slowly, the image scaling increment corresponds to fine adjustment of the field of view; when rotated quickly, it corresponds to rapid adjustment of the field of view. The adjustment range of the boundary constraints matches the actual magnification range from wide-angle observation to telephoto observation of the astronomical popular telescope.

[0016] Preferably, in step 3, the step of performing stable hovering detection based on the real-time state of the physical operation signal to determine the user's intent includes: Continuously monitor the real-time status of physical operation signals, including the displacement and zoom increment of the azimuth and pitch axes; When both the displacement and the scaling increment are zero, the static duration timer value of the telescope is incremented; when either the displacement or the scaling increment is not zero, the static duration timer value is reset. When the cumulative static duration value exceeds the preset duration threshold, the user is determined to be in a stable observation state.

[0017] More preferably, the preset duration threshold is 1.5 to 3.0 seconds.

[0018] The threshold is set based on the actual observation habits of users of astronomical science telescopes: When astronomy enthusiasts are using the telescope, if they are consciously and specifically observing a celestial object, they will actively maintain the physical attitude of the telescope stable, and the stable time is usually no less than 1.5 seconds; if they are randomly scanning, the attitude of the telescope will continuously change, and the stable time will not exceed 1 second. This threshold range can avoid false triggering caused by rapid scanning, while capturing the user's true observation intention, balancing sensitivity and accuracy. When the cumulative static duration exceeds this preset threshold, it will be determined that the user has consciously and stably pointed the astronomical science telescope at the celestial object, and has a clear intention to observe and acquire scientific knowledge.

[0019] Preferably, in step 4, the user profile information is obtained in the following way: By identifying users' language expression, questioning style, and habits, at least two user profiles are automatically distinguished: the first is the young children, whose expression characteristics include the use of reduplicated words, short sentences, and concrete and interesting questions; the second is the teenagers, whose expression characteristics include logical rationality, standardized word usage, and focus on theoretical knowledge.

[0020] Preferably, the dynamic construction of prompt words in step 4 includes: providing a template containing fixed basic framework fragments and dynamically filled content as prompt words to guide the large language model to generate personalized popular science content; The fixed basic framework fragment includes tone style, word limit, vocabulary selection rules and prohibited items that match the user profile; the dynamically filled content includes at least celestial body identifiers parsed from the current digital star map and astronomical knowledge retrieved from the knowledge base.

[0021] More preferably, the astronomical knowledge includes at least one of scientific attributes, humanistic stories, observation guidance, and related principles.

[0022] Preferably, the process of converting personalized science content into voice output includes: Personalized science content is transmitted to a text-to-speech engine, which selects corresponding speech synthesis parameters based on user profile information to generate audio with different intonations; among them, a lively and gentle intonation is used for children, while a calm and clear intonation is used for teenagers.

[0023] The present invention also provides a system for implementing an astronomical science telescope based on an AI agent, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the method for implementing the astronomical science telescope based on an AI agent.

[0024] Compared with the prior art, the beneficial effects of the present invention include at least the following: This invention provides a method and system for implementing an astronomical science telescope based on an AI intelligent agent. By converting the physical operation signals of the telescope into digital signals, it synchronously drives the digital star map to adjust the viewing angle and zoom. At the same time, it calls the star map AI application module to obtain celestial object identifiers, realizing real-time linkage observation between the physical pointing of the telescope and the content displayed on the digital star map, transforming one-way knowledge transmission into two-way real-time dialogue. Based on the real-time state of the physical operation signals, it performs stable hovering detection to determine the user's intention, filters out target celestial object identifiers, and constructs an AI interaction module with context awareness capabilities. Through retrieval enhancement generation and user profile recognition technology, it generates personalized science popularization content, solving the technical problem that existing science popularization content cannot be personalized and contextualized, realizing a closed-loop interactive experience of operation as exploration and observation as learning, and lowering the threshold for scientific exploration. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0026] Figure 1 A flowchart illustrating the implementation method of the AI-based astronomical science telescope provided by this invention.

[0027] Figure 2 This invention provides a schematic diagram of the structure of an astronomical science telescope based on an AI intelligent agent.

[0028] Figure 3 This is a schematic diagram of the interface of an astronomical science telescope based on an AI agent when a user asks a question, as provided in the example.

[0029] Figure 4 This is a schematic diagram of the interface of an AI-based astronomical science telescope receiving user questions, provided as an example. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of 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. The components of the embodiments of the present invention described and given in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0031] like Figure 1As shown in the embodiment, an implementation method for an astronomical science telescope based on an AI agent includes the following steps: S1. Real-time acquisition of physical operation signals from the user's use of the telescope, and conversion of the physical operation signals into digital signals.

[0032] S1-1 hardware configuration: like Figure 2 As shown, high-precision potentiometers or rotary encoders are installed on the azimuth and elevation axes of the telescope body to sense changes in the user's attitude when operating the telescope. These high-precision potentiometers or rotary encoders are rigidly fixed to the azimuth and elevation axes of the telescope, and the rotational movement of the telescope shaft can be synchronously transmitted to the sensors, ensuring accurate attitude sensing. A stepless rotary encoder is installed on the side of the telescope tube or in a user-friendly grip area to control the zoom of the image. This encoder is a purely mechanical, stepless structure, suitable for the core requirement of fine zoom in astronomical observations.

[0033] All the aforementioned sensors are connected to the input / output (I / O) pins of the core microcontroller (in this embodiment, an Arduino UNO development board). The microcontroller is internally mounted within the telescope, providing core control for signal acquisition. In this embodiment, an additional MPU6050 six-axis attitude sensor (integrating a three-axis accelerometer and a three-axis gyroscope) is rigidly mounted on the inner wall of the telescope tube, forming a dual attitude sensing structure with the sensors on the azimuth and pitch axes. This further enhances the attitude sensing accuracy required for astronomical observation. This six-axis sensor is connected to the microcontroller via a communication bus.

[0034] S1-2 Signal Acquisition and Processing: (1) Attitude signal: The firmware program on the Arduino UNO development board reads the analog or digital values ​​of the sensors on the azimuth and pitch axes at a high frequency of 5 milliseconds (i.e., 200 times per second). This high-frequency acquisition can accurately capture subtle changes in the telescope's attitude, meeting the needs of rapid target tracking in astronomical observation. If it is a potentiometer, the voltage value corresponding to its resistance change is read. This voltage value is linearly related to the rotation angle of the telescope axis, directly mapping the attitude change. If it is an encoder, the AB phase pulse signal is counted. The pulse count value corresponds to the rotation angle, and the pulse jump direction corresponds to the rotation direction, realizing the digital conversion of attitude change.

[0035] Simultaneously, the MPU6050 six-axis attitude sensor synchronously acquires raw data of three-axis acceleration and three-axis angular velocity. The firmware first performs zero-point calibration on the gyroscope—by continuously acquiring data from 100 times under the telescope's static state and averaging the results, eliminating hardware zero-drift errors. Then, a complementary filtering algorithm fuses the accelerometer and gyroscope data: the accelerometer calculates the telescope's static absolute attitude angle, and the gyroscope calculates the dynamic angle change. The fusion of these two data yields noise-free and drift-free precise pitch angles (corresponding to pitch axis movements) and azimuth angles (corresponding to azimuth axis movements), ensuring the accuracy and stability of attitude sensing during astronomical observations. In summary, all attitude sensor data is uniquely mapped to the telescope's physical attitude, achieving precise conversion of "telescope attitude changes into digital signals."

[0036] (2) Screen scaling signal: The firmware reads the signal from the continuously variable rotary encoder at a high frequency of milliseconds, the same as the attitude signal. The rotation direction is identified by detecting the level transition sequence of the encoder's AB phase pulses: clockwise rotation corresponds to screen magnification, and counterclockwise rotation corresponds to screen shrinkage, perfectly matching the user's observation and operation habits. The firmware has a built-in 50-millisecond anti-jitter threshold to filter out false signals caused by mechanical jitter and avoid invalid scaling.

[0037] Simultaneously, the rotation speed is determined based on the time interval between two pulses, enabling variable-speed scaling: a scaling step of 1 for slow rotation, suitable for fine-tuning of target details; a scaling step of 2 for medium-speed rotation; and a scaling step of 3 for fast rotation, suitable for the need to quickly adjust the observation magnification. The encoder's installation location and operating logic are perfectly suited for handheld telescope observation scenarios, allowing users to perform scaling operations with one hand without affecting the telescope's attitude. Ultimately, the user's physical scaling operation is converted into a precise digital signal with directional and velocity attributes, adapting to the scaling requirements of astronomical observations.

[0038] S1-3 Data Reception: The script running on the processing unit continuously monitors serial port data from the telescope's built-in microcontroller with a millisecond-level response frequency. The serial communication baud rate is 115200 bps, ensuring stable data transmission.

[0039] S2. Analyze and optimize digital signals to synchronously drive the digital star map image to adjust the viewing angle and zoom. At the same time, call the star map AI application module to obtain the celestial object identifiers corresponding to the center area of ​​the digital star map image, so as to realize the real-time linkage between the physical pointing of the telescope and the update of the digital star map image.

[0040] This step, acting as a bridge between hardware and software, is responsible for accurately mapping the received physical operation signals into real-time changes in the digital star map image, achieving real-time synchronization between the physical pointing of the astronomical science telescope and the observation perspective of the digital star map. All signal processing logic is adapted to the accuracy and smoothness requirements of astronomical observation and is deeply coupled with the usage scenarios of the astronomical science telescope.

[0041] S2-1 Data Analysis: Upon receiving a standardized, complete data command string ending with a newline character, the script immediately performs structured parsing, filters out invalid commands, and accurately extracts three core valid values: horizontal viewing angle displacement, vertical viewing angle displacement, and image zoom increment. Simultaneously, it verifies the timestamp information in the commands to ensure that the parsed values ​​are true mappings of the telescope's real-time physical operations. Specifically, the horizontal and vertical viewing angle displacements are digital representations of the telescope's azimuth and pitch axis attitude changes, while the image zoom increment is a digital representation of the telescope's stepless knob rotation operation. All the parsed values ​​correspond one-to-one with the user's physical operations on the telescope.

[0042] S2-2 API Command Conversion and Call: This step performs targeted optimization processing on the parsed digital signal and then calls the dedicated interface of the star map AI application module adapted for astronomical science telescopes. This interface matches the astronomical observation needs of the telescope. The specific processing and calling logic is as follows: (1) Viewpoint synchronization: The script processes the horizontal and vertical viewpoint displacements obtained by parsing using a smoothing filtering algorithm. A fixed smoothing coefficient is introduced to perform weighted fusion calculations on continuous displacements, eliminating numerical jumps caused by minor shaking during telescope hand operation and avoiding significant tremors in the digital star map image. The filtered azimuth and elevation angle parameters are used as core input parameters to call the adapted "precise viewpoint alignment interface". The star map AI application module is instructed to align the center point of the rendered star map image with the coordinates of the sky area pointed to by the user's physical telescope, realizing the linkage of "which star sky the telescope is physically pointing to and which star sky the digital star map is synchronously displaying". The azimuth deviation between the two is controlled within a very small range, meeting the precise aiming requirements of astronomical science popularization observation.

[0043] (2) Scaling Synchronization: The script adjusts the magnification of the digital star map precisely based on the parsed scaling increment of the image with directional attributes using a general field-of-view conversion formula. This formula is suitable for various astronomical science observation scenarios. The specific general calculation method is as follows: ;in, The adjusted field of view of the digital star map. This is the real-time field of view before adjustment. A fixed scaling factor adapted for astronomical observations. The zoom increment is the result of the image analysis. When the zoom increment is positive, the field of view decreases, the corresponding digital star map image is magnified, and celestial details can be observed. When the zoom increment is negative, the field of view increases, the corresponding digital star map image is shrunk, and a wider area of ​​the starry sky can be observed.

[0044] Meanwhile, this step performs tiered adaptation processing on the image scaling increment: when the telescope's stepless knob rotates slowly, the image scaling increment is small, and the field of view is finely adjusted to suit detailed observation of astronomical targets; when the knob rotates rapidly, the image scaling increment is large, and the field of view is quickly adjusted to suit the need for scanning a large area of ​​the starry sky; and reasonable boundary constraints are set for the field of view, and its adjustment range matches the actual magnification range of wide-angle to telephoto observation of astronomical popular science telescopes, avoiding ineffective scaling operations.

[0045] (3) Real-time feedback: The entire process of digital signal analysis, filtering optimization, formula conversion and interface call is completed in a short time, which is a millisecond-level response. When the user rotates the astronomical popular science telescope to adjust the observation position, raises or lowers the telescope tube to adjust the observation elevation angle, and rotates the stepless knob to adjust the observation magnification, the digital star map on the screen will complete the viewpoint movement and image scaling synchronously with low latency and no obvious lag, following the changes in the physical state of the telescope, and finally achieving an immersive and seamless linkage experience where what you point to is what you see. This is in line with the core usage requirements of the astronomical popular science telescope - the user does not need to pay attention to the physical equipment and digital image, and can focus on star observation and popular science knowledge acquisition.

[0046] S3. While the digital star map is updated synchronously, stable hovering detection is performed based on the real-time status of the physical operation signal to determine the user's intention. Based on the user's intention, the target celestial object is selected from the celestial object identifiers, and the AI ​​interaction module is triggered. This step is the core of realizing the context-aware intelligent interaction function of the astronomical science telescope. Unlike traditional passive science popularization equipment, this step uses intelligent algorithms to determine the user's astronomical observation intention and proactively initiates science popularization interaction at the appropriate time. All algorithm logic is designed around the usage characteristics of the astronomical science telescope and is deeply bound to the telescope's observation behavior, as detailed below: S3-1 Target Recognition: While the digital star map is updated synchronously, the script continuously monitors the digital star map, which is precisely synchronized with the physical pointing of the astronomical science telescope, in real time. The viewing angle and zoom of the digital star map are adjusted synchronously with the changes in the physical attitude of the astronomical science telescope. The range of the sky presented by the digital star map is completely consistent with the starry sky area pointed to by the physical astronomical science telescope, forming a real-time and precise correspondence between the physical observation angle of the telescope and the digital star map. The script uses the synchronized star map as the core visual basis to complete the subsequent star content recognition and interactive triggering.

[0047] S3-2 User Intent Determination: To avoid frequent triggering of invalid science popularization interactions when users rapidly rotate the telescope to scan the night sky, thus affecting the observation experience, a stable hovering judgment logic adapted to the observation habits of astronomical science popularization telescopes was designed. This logic is based on the real-time status determination of the telescope's hardware operation signals, specifically as follows: The system continuously monitors the real-time status of physical operation signals, including the displacement and zoom increments of the azimuth and pitch axes. Upon receiving each physical operation signal from the telescope, the real-time status is determined: if both displacement and zoom increment are zero, it indicates the user has not performed any physical operation on the astronomical telescope, and the telescope remains stationary, aligned with a specific celestial region; in this case, the telescope's stationary duration timer is accumulated. If the displacement or zoom increment is not zero, it indicates the user is rotating the telescope to adjust the observation azimuth, raising or lowering the telescope tube, or adjusting the zoom knob; in this case, the telescope's stationary duration timer is immediately reset to zero, and the timing process restarts. The real-time status of the physical operation signals records the real-time displacement of the telescope's azimuth and pitch axes and the zoom increment of the stepless knob; the telescope's stationary duration timer is used to accumulate the stable duration of the telescope without physical operation.

[0048] In this embodiment, the trigger threshold for the telescope's static duration timer is adjustable within a range of 1.5 to 3.0 seconds. This threshold is set based on the actual observation habits of users of astronomical science telescopes: when astronomy enthusiasts are consciously and specifically observing a celestial body, they will actively maintain the telescope's physical attitude stable, with a stabilization time typically not less than 1.5 seconds; if they are randomly scanning, the telescope's attitude will continuously change, and the stabilization time will not exceed 1 second. This threshold range can both avoid false triggers caused by rapid scanning and capture the user's true observation intention, balancing sensitivity and accuracy. When the telescope's static duration timer exceeds this preset threshold, it is determined that the user has consciously and stably aimed the astronomical science telescope at the celestial body, demonstrating a clear intention to observe and acquire scientific knowledge. In this embodiment, the stable telescope static duration timer is set to 2.0 seconds.

[0049] Once the stable hovering trigger conditions are met, the script immediately calls the preset AI interaction module and transmits the confirmed target celestial object identification information and the information from the digital star map to the AI ​​interaction module. The AI ​​interaction module then generates explanation content adapted to the astronomical science popularization scenario based on the target celestial object identification information. At the same time, a "single-time celestial object lock state" is set. The specific meaning of this lock state is: before the user actively moves the astronomical science popularization telescope and causes the celestial object identification to change, the popularization explanation interaction will not be triggered repeatedly for the same locked celestial object to avoid repeated broadcasting of the same content. This lock state is deeply linked to the physical operation signal of the telescope. That is, when the user rotates the telescope to change the observation target, the lock state will be automatically released, and the stable hovering detection process will be re-entered. A new interaction will be triggered only after the new user's observation intention is confirmed.

[0050] S4, the AI ​​interaction module retrieves astronomical knowledge from the knowledge base based on the user profile information and target celestial body identifiers identified when the trigger is activated, and dynamically constructs prompt words to generate personalized science popularization content. Then, the personalized science popularization content is converted into voice output, and subsequent voice questions from the user are received to form an interactive closed loop and complete the science popularization explanation.

[0051] Following the stable trigger signal in step S3, the AI ​​interaction module is activated. Using the observation images from the astronomical telescope and user interaction information as core inputs, it completes the entire process from information integration to content output through four progressively layered processing logics, ultimately generating and outputting highly relevant and personalized astronomical science content. The steps in this process follow a coherent logical relationship of "preliminary constraints → material retrieval → fusion generation → terminal output." The output results of the upstream steps serve as the core input basis for the downstream steps, with no independently separated processing steps, as detailed below: S4-1 User Profile Recognition: The AI ​​interaction module prioritizes the automatic determination of user profiles. This determination process requires no manual configuration and is based on the user's voice interaction characteristics. Specifically, by recognizing the user's language expression, questioning style, and expression habits, it automatically distinguishes two core user profiles: one is for young children, characterized by the use of reduplicated words, interjections, and short sentences, with questions leaning towards concrete and engaging associations; the other is for teenagers, characterized by logical and rational expression, questions leaning towards astronomical principles and related subject knowledge, and standardized and complete vocabulary. The user profile results output in this stage will serve as the core adaptation constraint for all subsequent stages, directly determining the overall adaptation standard for subsequent astronomical knowledge selection, language expression style, and content knowledge depth. It is the prerequisite foundation for the entire popular science content generation process.

[0052] S4-2 Enhanced Search and Integration of Multi-Dimensional Astronomical Information: The core input for this step is the complete digital star map image collected in step S3, which is precisely synchronized with the physical pointing of the astronomical science telescope, as well as the user's voice interaction requirements. After the AI ​​interaction module completes visual content analysis of the digital star map image, it uses the analyzed star field features and celestial object visual attributes as the search criteria to complete a precise matching query in a pre-built structured astronomical knowledge base. The structured astronomical knowledge base pre-organizes multi-dimensional standardized information on various astronomical targets, including: scientific attributes (celestial object type, distance, internal structure), humanistic stories (Chinese and foreign astronomical anecdotes, myths and legends, and the origins of ancient and modern naming), observation guidance (brightness magnitude, best observation time), and related principles (corresponding basic physical concepts and astronomical laws). After the query is completed, the telescope extracts the complete astronomical knowledge points that match the current star map image, and based on the user profile information determined in the first step, it performs targeted screening and simplification of the retrieved knowledge points: when targeting children, it selects interesting and concrete content and simplifies complex principles; when targeting teenagers, it retains complete scientific principles and subject-related knowledge to ensure the rigor and knowledge of the content. This step outputs adapted, multi-dimensional astronomical knowledge materials, providing authentic and accurate core knowledge support for subsequent content generation.

[0053] S4-3 Dynamically Constructing Prompt Words to Generate Personalized Science Popularization Content: This stage is a crucial fusion step, connecting the preceding and following sections. The core inputs are the user profile information from the first step, the enhanced astronomical knowledge materials from the second step, and the visual features of the star chart. The AI ​​interaction module organically integrates these three core information types, dynamically constructing exclusive customized prompt words according to preset science popularization expression rules. The core considerations in the construction process are: matching the language style and knowledge acceptance level of the user profile, ensuring the objectivity and rigor of astronomical knowledge, and controlling the conciseness of the science popularization content. The prompt words explicitly limit the output specifications: explanations for children are limited to 38 characters, primarily using metaphors and associations; explanations for teenagers are limited to 48 characters, primarily using principles and connections. After receiving the customized prompt words, the large language model generates customized astronomical science popularization content that fits the current observation scenario and is adapted to the user profile.

[0054] In the embodiment, the following example prompt words are given: Example 1: Prompt word construction adapted for children aged 8-12 (Orion Nebula observation scenario) (1) Fixed basic framework segment: You are a starry sky science popularization partner, with a lively and relaxed tone, share with fairy-tale-like short sentences, ≤38 words, use words such as starry sky or star clusters, prohibit professional terms, and naturally interweave interesting associations; (2) Dynamically filled content: The current observation is of the Orion Nebula, which is a star-forming region. The ancients called it part of the constellation Shenxiu, which looks like a glowing cotton candy. (3) Complete customized prompts: You are a star science popularization partner. Use a lively and relaxed tone, share with fairy-tale-like short sentences, ≤38 words, use words such as starry sky or star clusters, prohibit professional terms, and naturally interweave interesting associations; The current observation is the Orion Nebula, which is a star-forming region. The ancients called it part of the constellation Shen, which looks like a ball of glowing cotton candy.

[0055] The children's version of the science text generated based on the prompt: This nebula is Orion's starlight cotton candy, and new little stars are slowly being born inside it~.

[0056] Example 2: Constructing prompts tailored to teenagers (1) Fixed basic framework segment: You are a astronomical science popularization partner, with a rational and calm tone, using standardized expressions to explain astronomical knowledge, ≤48 words, closely related to the discipline, ensuring the content is rigorous, and naturally interspersed with ancient and modern knowledge; (2) Dynamically filled content: The current observation is the Orion Nebula, which is an emission nebula, 1344 light-years away from Earth. Its core is a star-forming region, involving the physical principle of gas ionization. (3) Complete customized prompts: You are a astronomical science popularization partner. Your tone is rational and calm. Use standardized expressions to explain astronomical knowledge. ≤48 words. Closely related to the discipline. Ensure the content is rigorous. Naturally interweave ancient and modern knowledge. The current observation is the Orion Nebula, which is an emission nebula. It is 1344 light-years away from Earth. Its core is a star-forming region. It involves the physical principle of gas ionization.

[0057] Based on the prompt, a science popularization text for teenagers is generated: The Orion Nebula is a typical emission nebula, located 1,344 light-years from Earth. The ionization of gas in its core region is a key condition for star formation.

[0058] S4-4 Science Popularization Content Terminal Output and Interaction Closed Loop: This is the final output stage, with the core input being the customized astronomical science popularization content generated in S4-3. The telescope sends this content to the text-to-speech engine, converting it into natural and fluent audio that matches the user profile: a lively and gentle tone for children, and a calm and clear tone for teenagers. After the audio file is generated, it is played through the speaker, providing automatic science popularization explanations of the currently observed celestial bodies. After the audio playback is complete, the telescope will automatically return to real-time monitoring mode, waiting for the user's next round of telescope physical operation or voice interaction commands. At this point, a complete interactive process of telescope physical operation - user intent triggering - AI science popularization explanation officially ends.

[0059] Based on the same inventive concept, the embodiment also provides a system for implementing an astronomical science telescope based on an AI agent, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the method for implementing the astronomical science telescope based on an AI agent.

[0060] like Figure 3 and 4 As shown, if a user wants to know "scientific knowledge about Altair and Vega," this invention, based on the user's intent, triggers an AI interaction module to provide suitable scientific content. This content is then sent to a text-to-speech engine, converting it into a natural and fluent audio file. The telescope plays this audio through a speaker, achieving automatic explanation and thus fulfilling the user's scientific knowledge request.

[0061] Finally, it should be noted that the above-described embodiments are merely specific implementations of the present invention, used to illustrate the technical solutions of the present invention, and not to limit it. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that any person skilled in the art can still modify or easily conceive of changes to the technical solutions described in the foregoing embodiments within the technical scope disclosed in the present invention, or make equivalent substitutions for some of the technical features; and these modifications, changes, or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for implementing an astronomical science telescope based on an AI intelligent agent, characterized in that, Includes the following steps: Step 1: Collect physical operation signals from the user's use of the telescope in real time and convert the physical operation signals into digital signals; Step 2: Analyze and optimize the digital signal to synchronously drive the digital star map image to adjust the viewing angle and zoom. At the same time, call the star map AI application module to obtain the celestial object identifiers corresponding to the center area of ​​the digital star map image, so as to realize the real-time linkage between the physical pointing of the telescope and the update of the digital star map image. Step 3: While the digital star map is being updated synchronously, stable hovering detection is performed based on the real-time status of the physical operation signal to determine the user's intention. Based on the user's intention, the target celestial body identifier is selected from the celestial body identifiers, and the AI ​​interaction module is triggered. Step 4: Based on the user profile information and target celestial body identifier identified at the time of triggering, the AI ​​interaction module retrieves astronomical knowledge from the knowledge base and dynamically constructs prompt words to generate personalized science popularization content. Then, the personalized science popularization content is converted into voice output and subsequent voice questions from the user are received to form an interactive closed loop and complete the science popularization explanation.

2. The method for implementing an astronomical science telescope based on an AI agent according to claim 1, characterized in that, Step 1 specifically includes: Azimuth adjustment signals are acquired by sensors installed on the azimuth and pitch axes of the telescope, and zoom control signals are acquired by a knob encoder installed on the telescope barrel. The azimuth adjustment signals and zoom control signals serve as physical operation signals. The acquired azimuth adjustment signal and zoom control signal are converted into standardized azimuth angle, pitch angle and zoom increment value; The standardized azimuth, elevation, and zoom increment values ​​are packaged into a formatted string as a digital signal and sent to the central processing unit via serial communication.

3. The method for implementing an astronomical science telescope based on an AI agent according to claim 2, characterized in that, When acquiring azimuth adjustment signals through sensors, the analog or digital values ​​of the sensors on the azimuth and pitch axes are read cyclically at millisecond-level high frequency; wherein, the sensors include angle sensors and inertial measurement units; When the angle sensor is a potentiometer, the change in voltage corresponding to the change in resistance is used to map the change in attitude rotation angle; when the sensor is an encoder, the AB phase pulse signals are counted, with the pulse count value corresponding to the rotation angle and the pulse transition direction corresponding to the rotation direction. The inertial measurement unit is used to synchronously acquire triaxial acceleration and triaxial angular velocity, and to perform fusion processing on the triaxial acceleration and triaxial angular velocity acquired by the inertial measurement unit.

4. The method for implementing an astronomical science telescope based on an AI agent according to claim 3, characterized in that, The fusion process includes: Zero-point calibration is performed on the gyroscope of the inertial measurement unit by collecting multiple sets of data while the telescope is stationary and taking the average value as the gyroscope data to eliminate zero drift error. The three-axis acceleration, three-axis angular velocity and gyroscope data are fused by complementary filtering algorithm to output standardized azimuth and pitch angles, thus completing the digital conversion of attitude change.

5. The method for implementing an astronomical science telescope based on an AI intelligent agent according to claim 1, characterized in that, Scaling control signals are acquired via a rotary encoder, including: The rotation direction is identified by detecting the level transition sequence of the AB phase pulse signals of the rotary encoder to match the user's observation and operation habits; and an anti-vibration threshold is set to filter mechanical vibration noise. The rotation speed is determined by the time interval between two pulses to achieve variable speed scaling, and the scaling increment step is dynamically set accordingly to achieve variable speed scaling control to adjust the user's observation magnification.

6. The method for implementing an astronomical science telescope based on an AI agent according to claim 3, characterized in that, Step 2 specifically includes: Digital signals from the telescope are received at millisecond-level frequency via serial communication, and the horizontal and vertical angle displacements and image scaling increments are analyzed and extracted. Smoothing filtering is applied to the horizontal and vertical angle displacements to eliminate numerical jumps caused by physical jitter in the telescope. The azimuth and elevation angle parameters after smoothing and filtering are called into the view alignment interface of the star map AI application module to synchronize the physical pointing of the telescope with the celestial object markings updated in the digital star map.

7. The method for implementing an astronomical science telescope based on an AI agent according to claim 1, characterized in that, Step 3, which involves performing stable hovering detection based on the real-time state of the physical operation signal to determine the user's intent, includes: Continuously monitor the real-time status of physical operation signals, including the displacement and zoom increment of the azimuth and pitch axes; When both the displacement and the scaling increment are zero, the static duration timer value of the telescope is incremented; when either the displacement or the scaling increment is not zero, the static duration timer value is reset. When the cumulative static duration value exceeds the preset duration threshold, the user is determined to be in a stable observation state.

8. The method for implementing an astronomical science telescope based on an AI intelligent agent according to claim 1, characterized in that, User profile information is obtained through the following methods: By identifying users' language expression, questioning style, and habits, at least two user profiles are automatically distinguished: the first is the young children, whose expression characteristics include the use of reduplicated words, short sentences, and concrete and interesting questions; the second is the teenagers, whose expression characteristics include logical rationality, standardized word usage, and focus on theoretical knowledge.

9. The method for implementing an astronomical science telescope based on an AI intelligent agent according to claim 1, characterized in that, The dynamic construction of prompt words mentioned in step 4 includes: providing a template containing fixed basic framework fragments and dynamically filled content as prompt words to guide the large language model to generate personalized popular science content; The fixed basic framework fragment includes tone style, word limit, vocabulary selection rules and prohibited items that match the user profile; the dynamically filled content includes at least celestial body identifiers parsed from the current digital star map and astronomical knowledge retrieved from the knowledge base.

10. A system for implementing an astronomical science telescope based on an AI intelligent agent, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method for implementing an AI-based astronomical science telescope as described in any one of claims 1-9.

Images