Facility horticulture intelligent production practical training system and method based on digital twinning

Abstract

This invention discloses a smart production training system and method for facility horticulture based on digital twins, belonging to the field of smart production training technology. It includes: a smart horticulture production training module, a virtual twin module, a smart control module, and a training and intelligent evaluation module. The invention uses the smart horticulture production training module to collect environmental, equipment, and crop growth data and execute control commands; the virtual twin module constructs a one-to-one mapped virtual three-dimensional scene, realizing real-time virtual-real mapping between physical entities and the virtual scene; the smart control module generates control strategies based on simulated data and training conditions and issues control commands; the training and intelligent evaluation module provides operation guidance, visual display, data recording, and automated scoring and comprehensive evaluation, improving the safety of smart production training in facility horticulture and solving the problems of low safety in existing technologies due to long cycles, high costs, slow rotation, and high risks.

Description

Technical Field

[0001] This invention relates to the field of intelligent production training technology, and in particular to an intelligent production training system and method for facility horticulture based on digital twins. Background Technology

[0002] Existing intelligent production training in facility horticulture mainly relies on seasonal field practice or static simulation software for teaching and training. The practical training is highly dependent on crop growth cycles and seasonal climate, resulting in lengthy training periods and a lack of flexibility and immediacy in teaching arrangements, making it difficult to meet the needs of high-frequency, large-scale training rotations. The training process requires significant investment in venues, facilities, agricultural inputs, and labor costs, and is easily affected by uncertainties such as the environment, pests and diseases, and equipment malfunctions, leading to poor training stability and high costs. Furthermore, traditional training methods and static simulation software cannot actively reproduce high temperatures, low temperatures, insufficient sunlight, water and fertilizer imbalances, and equipment malfunctions. The current technology often faces extreme working conditions and dangerous scenarios, making it difficult to conduct targeted emergency response training, resulting in insufficient coverage and realism in practical training. At the same time, existing technologies lack digital twin mapping, virtual-real interactive control, crop growth model extrapolation, and intelligent evaluation mechanisms. This makes it impossible to achieve visualized presentation, accurate prediction, and intelligent intervention of the entire process of facility horticulture production. It is difficult to effectively solve the prominent pain points of "difficulty in entering greenhouses, slow rotation, and high risk" in agricultural colleges and vocational training. It is also unable to provide low-cost, repeatable, and highly realistic industry-education integration practical training support for the cultivation of smart agriculture talents, and it is difficult to adapt to the intelligent production of modern facility horticulture. Summary of the Invention

[0003] To address the technical problems of low safety in intelligent production training for facility horticulture due to long cycles, high costs, slow turnover, and high risks in existing technologies, this invention provides an intelligent production training system and method for facility horticulture based on digital twins. The technical solution is as follows: On the one hand, a digital twin-based intelligent production training system for facility horticulture is provided. This system includes: an intelligent horticulture production training module, a virtual twin module, an intelligent control module, and a training and intelligent evaluation module. The intelligent horticulture production training module collects facility horticulture data from the target facility horticulture production site and executes control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data, and crop growth status data. The virtual twin module is used to construct a virtual 3D scene that maps one-to-one with the intelligent horticulture production training module based on digital twin technology, synchronizing environmental parameters of the target facility horticulture production site in real time. The system tracks equipment operation status and crop growth patterns to achieve a virtual-real mapping between physical entities and virtual scenes. The intelligent control module generates horticultural intelligent production training control strategies based on simulation data from the virtual twin module and preset training conditions, and issues control commands to the horticultural intelligent production training module to support condition simulation and virtual-real interactive control. The training and intelligent assessment module provides trainees with standardized operation guidance for the entire process of facility horticulture intelligent production, synchronous real-time visualization of virtual twin scenes, data collection and trajectory recording throughout the training operation process, and automated scoring, intelligent assessment, and comprehensive evaluation based on operational behavior and training results.

[0004] On the other hand, a training method for intelligent production in facility horticulture based on digital twins is provided. This method includes: collecting facility horticulture data from the target facility horticulture production site and executing control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data, and crop growth status data; constructing a virtual three-dimensional scene that maps one-to-one with the intelligent horticulture production training module based on digital twin technology, and synchronizing the environmental parameters, equipment operating status, and crop growth morphology of the target facility horticulture production site in real time to achieve a virtual-real mapping between physical entities and virtual scenes; generating intelligent horticulture production training control strategies based on the simulation data of the virtual twin module and preset training conditions, and issuing control commands to the intelligent horticulture production training module to support condition simulation and virtual-real interactive control; providing trainees with standardized operation guidance for the entire process of intelligent facility horticulture production, synchronous real-time visualization of the virtual twin scene, data collection and trajectory recording throughout the training operation process, and automated scoring, intelligent assessment, and comprehensive evaluation based on operational behavior and training results.

[0005] Beneficial effects The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: By constructing a training system that integrates data acquisition, digital twin mapping, intelligent control, and teaching assessment, the system can acquire real-time data on all dimensions of facility horticulture and accurately execute control commands, achieving high-fidelity real-time mapping and dynamic synchronization between physical and virtual scenes. Based on the adaptive generation of safe and feasible control strategies under the training conditions, the system improves the reliability and intelligence level of the simulation and virtual-real interactive control. At the same time, it provides trainees with standardized operation guidance, visualized training displays, full-process data traceability, and automated assessment, effectively improving the standardization of training, teaching efficiency, and objectivity of assessment, and significantly enhancing the authenticity, interactivity, and training effect of intelligent production training in facility horticulture. Attached Figure Description

[0006] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0007] Figure 1 A schematic diagram of the structure of a digital twin-based intelligent production training system for facility horticulture provided in an embodiment of this application; Figure 2 A flowchart illustrating the dynamic update process of a digital twin-based intelligent production training system for facility horticulture, as provided in this application embodiment. Figure 3 A flowchart of a digital twin-based smart production training method for facility horticulture provided in this application embodiment. Detailed Implementation

[0008] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0009] like Figure 1The diagram shown is a system structure diagram of a digital twin-based intelligent production training system for facility horticulture provided in this application embodiment. The system includes: an intelligent horticulture production training module, a virtual twin module, an intelligent control module, and a training and intelligent evaluation module. The intelligent horticulture production training module collects facility horticulture data from the target facility horticulture production site and executes control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data, and crop growth status data. The virtual twin module constructs a virtual three-dimensional scene that maps one-to-one with the intelligent horticulture production training module based on digital twin technology, synchronizing the target facility horticulture production site in real time. The system simulates environmental parameters, equipment operating status, and crop growth patterns to achieve a virtual-real mapping between physical entities and virtual scenes. The intelligent control module generates horticultural intelligent production training control strategies based on the simulation data and preset training conditions of the virtual twin module, and issues control commands to the horticultural intelligent production training module to support condition simulation and virtual-real interactive control. The training teaching and intelligent evaluation module provides trainees with standardized operation guidance for the entire process of facility horticulture intelligent production, synchronous real-time visualization of virtual twin scenes, data collection and trajectory recording of the entire training operation process, and automated scoring, intelligent assessment, and comprehensive evaluation based on operational behavior and training results.

[0010] In this embodiment, the horticultural intelligent production training module, virtual twin module, intelligent control module, and training and intelligent evaluation module are deployed sequentially and operate collaboratively. First, the horticultural intelligent production training module comprehensively collects environmental, equipment, and crop growth status data from the facility horticulture site and executes control commands, providing the system with a precise data source and ensuring the real-time and reliable execution of controls. Subsequently, the virtual twin module constructs a one-to-one virtual 3D scene based on digital twin technology and synchronizes physical scene parameters in real time, achieving high-fidelity virtual-real mapping and significantly improving the visualization and immersion of the training scene. Next, the intelligent control module generates adaptive control strategies based on virtual twin simulation data and preset training conditions and issues commands to realize multi-type working condition simulation and virtual-real interactive control, enhancing the intelligence and adaptive control capabilities of the training system. Finally, the training and intelligent evaluation module provides standardized operation guidance, real-time visual display, operation data collection and recording, and automated assessment and evaluation, realizing standardized, traceable, and objective evaluation of the entire training process, effectively improving the teaching quality, operational standardization, and training effect of facility horticulture intelligent production training.

[0011] Further environmental data include air temperature, relative humidity, soil temperature, relative humidity, light intensity, carbon dioxide concentration, nutrient solution EC value, and nutrient solution pH value. Equipment status data includes the start / stop status, running time, operating current, operating voltage, and fault codes of fans, water pumps, shade nets, window openers, and integrated water and fertilizer machines; Crop growth status data includes plant height, stem diameter, leaf area, number of leaves, growth vigor level, degree of disease and pest occurrence, and yield prediction parameters.

[0012] In this embodiment, multi-dimensional and comprehensive data collection is conducted on the environment, equipment, and crop growth status of the facility horticulture production site. Specifically, environmental data such as air temperature, relative humidity, soil temperature, relative humidity, light intensity, carbon dioxide concentration, nutrient solution EC value, and nutrient solution pH value are collected; equipment status data such as the start / stop status, running time, operating current, operating voltage, and fault codes of fans, water pumps, shade nets, window openers, and integrated water and fertilizer machines are collected; and crop growth status data such as plant height, stem diameter, leaf area, number of leaves, growth level, degree of pest and disease occurrence, and yield prediction parameters are collected. This enables comprehensive perception and quantitative representation of key information at the production site, providing real, complete, and high-precision data support for subsequent digital twin modeling, intelligent control strategy generation, and practical training evaluation. This effectively improves the accuracy of the system's virtual-real mapping, the scientific nature of control decisions, and the objectivity and comprehensiveness of practical training assessment.

[0013] Furthermore, the specific process for constructing a virtual 3D scene that maps one-to-one with the horticultural intelligent production training module is as follows: Geometric parameters of the greenhouse frame size, enclosure structure, and spatial zoning are collected; equipment models, installation coordinates, and pipeline routes are marked; and spatial modeling of cultivation ridges and planting layout is performed to complete the digital restoration of the entity, mesh modeling, texture mapping, and scene assembly, forming a high-fidelity three-dimensional twin model that is completely consistent with the entity of the target facility horticulture production site. Parametric modeling is performed based on the geometric dimensions of the target facility horticulture production site. Corresponding material attributes are assigned according to the material characteristics of the entity. The spatial position of the three-dimensional model is determined according to the physical installation coordinates. After calibration, accuracy verification and consistency matching based on the digital twin model, a digital three-dimensional model that is completely identical to the physical entity is formed.

[0014] In this embodiment, when constructing a virtual 3D scene that maps one-to-one with the horticultural intelligent production training module, firstly, a 3D laser scanning and visual measurement fusion algorithm is used to collect high-precision geometric parameters of the greenhouse frame size, enclosure structure, and spatial partitioning. Combined with an equipment coordinate calibration algorithm, the spatial point calibration of equipment models, installation coordinates, and pipeline routes of equipment such as fans, water pumps, and integrated water and fertilizer machines is completed. Based on the irregular triangular mesh (TIN) algorithm, the cultivation ridges and planting layout are refined into spatial models. The entity digital restoration, mesh modeling, texture mapping, and scene assembly are completed in sequence. Then, a basic model is constructed based on the physical entity's geometric dimensions using a parametric-driven modeling algorithm. According to the entity's material characteristics, corresponding physical rendering attributes are assigned through a material attribute matching algorithm. The model pose is determined based on the actual installation coordinates through a spatial coordinate transformation algorithm. Finally, the model is calibrated and verified through a twin model accuracy verification algorithm and a multi-scale consistency matching algorithm, ultimately forming a high-fidelity digital 3D twin model that is completely consistent with the physical entity's geometry, material, and spatial position.

[0015] It should be understood that, such as Figure 2 The diagram shows the dynamic update flowchart of the intelligent production training system for facility horticulture based on digital twins provided in this application embodiment. The specific process is as follows: First, obtain the virtual and real data synchronization verification interval, and compare it with the preset verification interval critical interval (composed of a critical lower limit and a critical upper limit). If the verification interval is within the critical interval, directly maintain the refresh time interval benchmark threshold. If the verification interval is greater than the critical upper limit, calculate the verification interval offset of the positive deviation, input it into the verification interval-refresh time interval threshold mapping relationship, output the interval threshold downward adjustment coefficient, and then combine the refresh time interval benchmark threshold with the downward adjustment coefficient. If the verification interval is less than the critical lower limit, calculate the verification interval deviation of the negative deviation, input it into the same mapping relationship, output the interval threshold upward adjustment coefficient, and then combine the benchmark threshold with the upward adjustment coefficient. Finally, all three scenarios converge to obtain the target refresh time interval threshold, completing the control process.

[0016] Furthermore, the virtual twin module receives environmental parameters, equipment operating status, and crop growth morphology data collected by the horticultural smart production training module in real time, and synchronously drives the virtual three-dimensional scene to be dynamically updated. The specific steps for performing dynamic updates are as follows: The virtual-real data synchronization verification interval represents the period during which the virtual twin module performs consistency comparisons between the physical data of the target facility horticulture production site and the virtual model data; A pre-constructed mapping relationship between the verification interval and the refresh time interval threshold is used to characterize the quantitative correspondence and linkage adjustment rules between the virtual and real data synchronization verification interval, the critical interval of the verification interval, the offset of the verification interval, the deviation of the verification interval, the adjustment coefficient of the interval threshold, the adjustment coefficient of the interval threshold, and the baseline threshold of the refresh time interval. If the virtual and real data synchronization verification interval is within the preset verification interval critical range, the refresh time interval baseline threshold is maintained. The preset verification interval critical range represents the closed interval formed by the preset verification interval critical lower limit and the preset verification interval critical upper limit.

[0017] Dynamic updates also include: If the virtual and real data synchronization verification interval is greater than the preset verification interval critical upper limit, the verification interval offset is input into the verification interval-refresh time interval threshold mapping relationship, and the interval threshold reduction coefficient is output. The refresh time interval base threshold and the interval threshold reduction coefficient are combined to obtain the target refresh time interval threshold. The verification interval offset represents the positive difference between the virtual and real data synchronization verification interval and the preset verification interval critical upper limit. If the virtual and real data synchronization verification interval is less than the preset verification interval critical lower limit, the verification interval deviation is input into the verification interval-refresh time interval threshold mapping relationship, and the interval threshold adjustment coefficient is output. The refresh time interval base threshold and the interval threshold adjustment coefficient are combined to obtain the target refresh time interval threshold. The verification interval deviation represents the negative difference between the virtual and real data synchronization verification interval and the preset verification interval critical lower limit.

[0018] In this embodiment, the virtual twin module receives environmental parameters (air temperature, relative humidity, etc.), equipment operating status (start / stop and parameters of fans, water pumps, etc.), and crop growth morphology data (plant height, leaf area, etc.) collected in real time by the horticultural smart production training module. A data preprocessing algorithm is used to denoise, normalize, and remove outliers from the received data, eliminating invalid data caused by environmental interference and sensor errors. This ensures the accuracy and completeness of the input data, providing high-quality data support for subsequent dynamic updates of the virtual 3D scene. This effectively avoids scene update deviations caused by data distortion and improves the consistency of virtual-real mapping. Subsequently, based on the preprocessed data, combined with... An adaptive PID control algorithm drives the virtual 3D scene to update dynamically. The specific steps are as follows: First, the core definition of the virtual-real data synchronization verification interval is clarified. It represents the fixed period of time during which the virtual twin module uses a periodic comparison algorithm to compare the consistency of the physical data of the target facility horticulture production site with the virtual model data. This interval can be initially set according to the complexity of the training scenario and the data update requirements, and it supports dynamic fine-tuning based on real-time data feedback. The rationality of its setting directly determines the balance between the real-time performance of the scene update and the consumption of system resources. It can avoid the virtual-real data from becoming disconnected due to an excessively long interval, and it can also prevent the system's computing power from being wasted due to an excessively short interval. Secondly, based on big data mining algorithms, a mapping relationship between the verification interval and the refresh interval threshold is constructed in advance. This mapping relationship is achieved through in-depth analysis of historical training data, virtual and real data deviation data, and system operating parameters. It accurately represents the quantitative correspondence and linkage adjustment rules between the virtual and real data synchronization verification interval, the verification interval critical interval, the verification interval offset, the verification interval deviation, the interval threshold adjustment coefficient, the interval threshold adjustment coefficient, and the refresh interval baseline threshold. The verification interval critical interval is a closed interval formed by the preset verification interval critical lower limit and the preset verification interval critical upper limit. During the preset process, the optimal critical value is determined through multiple iterative tests, taking into account the real-time requirements of the training scenario and the upper limit of the system computing power, to ensure the scientific and reasonable setting of the interval and provide a clear basis for subsequent threshold adjustments.During the dynamic scene update process, the virtual twin module uses an adaptive PID adjustment algorithm to collect the current virtual-real data synchronization verification interval in real time and compare it with the preset verification interval critical range to achieve dynamic adaptive adjustment of the refresh time interval threshold: If the virtual-real data synchronization verification interval is within the preset verification interval critical range, it indicates that the current virtual-real data consistency is good, and the system does not need to adjust the refresh frequency. The adaptive PID adjustment algorithm outputs a maintenance command to maintain the refresh time interval baseline threshold, ensuring a stable scene update frequency. This ensures both the real-time performance of the virtual-real mapping and avoids unnecessary system computing power consumption, improving system operating efficiency. If the virtual-real data synchronization verification interval is greater than the preset verification interval critical upper limit, it indicates that the current virtual-real data comparison cycle is too long and data disconnection is likely to occur. In this case, the verification interval offset (i.e., the positive difference between the virtual-real data synchronization verification interval and the preset verification interval critical upper limit) is calculated using a difference calculation algorithm. This offset is then input into the pre-built verification interval-refresh time interval threshold mapping relationship. The coefficient calculation model outputs the corresponding interval threshold downward adjustment coefficient. Then, a multiplication algorithm is used to combine the refresh interval baseline threshold and the interval threshold downward adjustment coefficient to obtain the target refresh interval threshold. This shortens the scene refresh interval, improves the data synchronization frequency and scene update real-time performance, quickly reduces the deviation between virtual and real data, and ensures that the virtual scene and physical entity remain consistent. If the virtual and real data synchronization verification interval is less than the preset verification interval critical lower limit, it indicates that the current virtual and real data comparison cycle is too short, which will cause a waste of system computing power. At this time, the difference calculation algorithm is used to calculate the verification interval deviation (i.e., the negative difference between the virtual and real data synchronization verification interval and the preset verification interval critical lower limit). This deviation is input into the verification interval-refresh interval threshold mapping relationship, and the corresponding interval threshold upward adjustment coefficient is output. A multiplication algorithm is used to combine the refresh interval baseline threshold and the interval threshold upward adjustment coefficient to obtain the target refresh interval threshold. This extends the scene refresh interval, reduces system computing power consumption, and ensures that the consistency of virtual and real data is within a reasonable range. Throughout the dynamic update process, the adaptive PID adjustment algorithm provides real-time feedback on the adjustment results and performs real-time calibration of the target refresh time interval threshold. Combined with the dynamic correction of the mapping relationship between the verification interval and the refresh time interval threshold, it ensures that each threshold adjustment accurately adapts to the current virtual-real data comparison status, realizing intelligent and adaptive adjustment of the dynamic update of the virtual 3D scene. This not only ensures the high-fidelity mapping between the virtual scene and the physical entity, but also achieves a reasonable allocation of system computing power, effectively improving the operational stability, real-time performance, and economy of the virtual twin module. It provides reliable scenario support for subsequent working condition simulation, virtual-real interactive control, and practical training evaluation.

[0019] Furthermore, the specific steps for generating intelligent horticultural production training and control strategies are as follows: By calling preset training conditions, including conventional production training conditions, extreme environment training conditions, equipment failure training conditions, and water and fertilizer imbalance training conditions, each training condition has preset corresponding target parameters, triggering conditions, and training assessment standards. The simulation data from the virtual twin module will be compared item by item with the target parameters of the currently enabled preset training conditions. The deviation values ​​between the simulation data and the target parameters of the currently enabled preset training conditions will be analyzed to clarify the type and degree of deviation. The types of deviation include environmental parameter deviation, equipment operating status deviation, and crop growth morphology deviation.

[0020] In this embodiment, the system first calls up multiple preset training scenarios, clarifying the core parameters and execution standards for each scenario. The routine production training scenario pre-sets environmental parameters, equipment operating parameters, and crop growth target parameters consistent with daily facility horticulture production, for basic training. The extreme environment training scenario pre-sets target parameters and triggering conditions under extreme conditions such as high temperature, low temperature, and low light, to simulate production scenarios in complex environments. The equipment failure training scenario pre-sets common failure types, triggering conditions, and corresponding handling standards for equipment such as fans and pumps, to cultivate trainees' troubleshooting abilities. The water and fertilizer imbalance training scenario pre-sets target parameters and assessment standards for scenarios such as insufficient or excessive nitrogen, phosphorus, and potassium supply, and abnormal water supply, to enhance water and fertilizer regulation operation capabilities. Each scenario clearly defines its corresponding target parameters, triggering conditions, and training assessment standards, providing a clear basis for the generation of subsequent regulation strategies, effectively improving the relevance and comprehensiveness of the training, and allowing trainees to experience the regulation needs under different production scenarios. Subsequently, the system extracts the simulated data output by the virtual twin module. This data is completely synchronized with the environmental conditions, equipment operation, and crop growth status of the physical training site, ensuring the authenticity and accuracy of the data. The simulated data is then precisely compared item by item with the target parameters corresponding to the currently activated training conditions. A difference calculation algorithm is used to analyze the deviation values ​​between the two, clarifying the specific numerical range of the deviation and accurately determining the type of deviation. If the deviation stems from differences between environmental parameters such as air temperature, humidity, and CO2 concentration and the target values ​​of the operating conditions, it is determined to be an environmental parameter deviation. If the deviation stems from the operating status of equipment such as fans and water pumps not meeting the requirements of the operating conditions, it is determined to be an equipment operating status deviation. If the deviation stems from crop growth indicators such as plant height and leaf area not meeting the target values ​​of the operating conditions, it is determined to be a crop growth morphology deviation. The degree of deviation is also quantified, distinguishing between slight, moderate, and severe deviations, providing a precise basis for subsequent targeted adjustments. This step ensures that the control strategy is precisely matched with the training conditions, aligning with actual production scenarios. It allows trainees to grasp the control logic under different conditions and provides scientific guidance for subsequent environmental adjustments, equipment operation, and water and fertilizer management through precise deviation analysis. This effectively enhances the practicality and professionalism of the training, achieving a deep integration of theoretical teaching and practical operation. It helps trainees quickly master the core control skills of intelligent horticultural production while ensuring the scientific validity and feasibility of the control strategy. This avoids equipment damage or abnormal crop growth caused by blind operation, further strengthening the training effect and ensuring that the training process is highly consistent with actual production scenarios.

[0021] Furthermore, the generation of intelligent horticultural production training and control strategies also includes: Based on the type and degree of deviation, corresponding control schemes are generated. The control schemes include environmental control schemes, equipment operation control schemes, and water and fertilizer supply control schemes. The environmental control schemes correspond to the adjustment of parameters such as air temperature and humidity and light intensity. The equipment operation control plan corresponds to the start-up, shutdown, and parameter adjustment of equipment such as fans and water pumps; The water and fertilizer supply control scheme corresponds to the adjustment of fertilizer and water supply of the integrated water and fertilizer machine; The corresponding control schemes generated are tested for feasibility to ensure that they will not cause damage to physical equipment, abnormal crop growth, or safety risks to the training. If the test is passed, the various control schemes are integrated to form a complete horticultural smart production training control strategy. If the test fails, the control parameters are iteratively optimized and readjusted until the test is passed.

[0022] In this embodiment, an environmental control scheme is generated to address deviations in environmental parameters, focusing on adjusting air temperature and humidity, light intensity, carbon dioxide concentration, nutrient solution EC value, and pH value. By quantifying the adjustment targets and execution ranges, the scheme ensures that the parameters quickly return to the target operating range, creating stable and suitable environmental conditions for crop growth. To address deviations in equipment operation status, an equipment operation control scheme is generated, covering start-stop control, speed adjustment, opening calibration, and operating parameter optimization for fans, water pumps, shade nets, window openers, and integrated water and fertilizer machines, ensuring a high degree of coordination between equipment actions and actual training conditions. To address deviations in crop growth morphology, a water and fertilizer supply control scheme is generated, with the integrated water and fertilizer machine as the execution terminal. This scheme precisely adjusts the fertilizer supply, water supply, and nutrient ratio to correct deviations in growth indicators such as plant height, stem diameter, and leaf area. The parallel generation of these three control schemes significantly improves the strategy response speed and the targeting of the control. The system then conducts multi-dimensional feasibility verification of each control scheme, sequentially checking the equipment operation safety boundaries, crop growth tolerance thresholds, and operational risk points in the training. This ensures that control commands will not cause equipment overload, crop stress damage, or training safety hazards. Pre-verification effectively avoids operational risks and improves strategy reliability. If the verification fails, an iterative optimization mechanism is initiated to correct and adjust parameters exceeding limits, illegal commands, or unreasonable logic. The verification process is then re-executed until all indicators are qualified. Finally, the verified environmental control, equipment operation control, and water and fertilizer supply control schemes are integrated into a unified whole, clarifying the execution sequence, linkage relationships, and priorities. This forms a complete, compliant, and implementable intelligent horticultural production training control strategy, ensuring that the control effect is highly consistent with the training objectives and that the training process is safe, standardized, and traceable. This significantly enhances the scientific nature of intelligent control and the practical value of training.

[0023] Furthermore, the specific process for verifying the feasibility of the generated corresponding control plan is as follows: If the deviation type is environmental parameter deviation, an environmental control plan is generated: when the air humidity is lower than the target air humidity value, the spray equipment is activated and the spray frequency is adjusted to the preset spray frequency reference value; when the light intensity is lower than the target light intensity value, the supplementary lighting equipment is activated and the supplementary lighting intensity is adjusted to the preset supplementary lighting intensity upper limit value until the environmental parameter deviation is reduced to the preset allowable range. If the deviation type is equipment operating status deviation, generate an equipment operation control plan: when the fan operating current is not within the preset rated current range, adjust the fan speed to calibrate the operating current to the preset rated current range; when the water pump does not start according to the operating conditions, issue a start command and adjust the water pump operating power to ensure that the water supply meets the operating conditions standard; when the equipment has a fault code, issue a shutdown and maintenance command, and start the backup equipment to maintain the training operation. If the deviation type is crop growth morphology deviation, a water and fertilizer supply regulation plan is generated: when the plant height is lower than the target plant height value, the water and fertilizer integrated machine is adjusted to increase the nitrogen supply; when the leaf area is lower than the target leaf area value, the water and fertilizer integrated machine is adjusted to increase the phosphorus and potassium supply until the crop growth morphology deviation is reduced to the preset allowable range.

[0024] In this embodiment, the feasibility of the generated control scheme is verified. This requires precise execution of control operations based on the deviation type and detailed implementation steps to ensure the control scheme is effective and meets practical training needs. The specific process and technical effects are as follows: When an environmental parameter deviation is detected, a targeted environmental control scheme is immediately generated. If the detected air humidity is lower than the preset air humidity target value, the spraying equipment is immediately activated, and the spraying frequency is adjusted to the preset spraying frequency benchmark value. This operation can quickly increase the ambient air humidity, achieving a precise and uniform increase in humidity, avoiding sudden increases in humidity that could stimulate crop growth. Reduce ineffective energy consumption of spraying equipment to ensure efficient and energy-saving control until the air humidity deviation is reduced to the preset allowable range, ensuring a suitable humidity environment for crop growth; if the light intensity is detected to be lower than the preset light intensity target value, immediately start the supplemental lighting equipment and adjust the supplemental lighting intensity to the preset upper limit value. Quickly make up for the problem of insufficient light through high-intensity supplemental lighting, minimize the impact of insufficient light on crop photosynthesis, and avoid crop growth abnormalities such as excessive growth and yellowing leaves. Continue supplemental lighting until the light intensity deviation is reduced to the preset allowable range, providing stable and sufficient light conditions for crop growth. When a deviation type of equipment operation status deviation is detected, an equipment operation control plan is generated. If the fan operating current is detected to be outside the preset rated current range, the fan speed is adjusted in a timely manner. Through precise speed control, the fan operating current is calibrated to the preset rated current range. This effectively avoids overload or underload operation of the fan due to abnormal current, extends the fan's service life, and ensures stable fan air delivery efficiency, guaranteeing uniform air circulation in the training environment and providing a good foundation for crop growth and normal equipment operation. If the water pump is detected to have not started according to the operating conditions, a start command is immediately issued, and the speed is adjusted accordingly. The water pump's operating power is precisely controlled through power adjustment to ensure that the water supply strictly meets the training conditions. This avoids problems such as insufficient water supply leading to crop water shortage and interruption of water and fertilizer supply, or excessive water supply causing water waste and root rot, thus ensuring the continuity and stability of the training process. If a fault code is detected in the equipment, a shutdown and maintenance order is immediately issued to prevent the fault from escalating and damaging the equipment, affecting the training progress. At the same time, the backup equipment is quickly activated to ensure seamless connection and maintain the normal operation of all training operations, minimizing the impact of equipment failure on the training and ensuring the smooth progress of the training task.When a deviation type of crop growth morphology deviation is detected, a water and fertilizer supply control plan is generated. If the crop height is detected to be lower than the preset target value, the water and fertilizer system is immediately adjusted to increase the nitrogen supply. Nitrogen can effectively promote the growth of crop stems and leaves. By precisely increasing the nitrogen supply, the problem of insufficient plant height can be specifically solved, promoting rapid crop growth. At the same time, it avoids problems such as excessive growth and lodging caused by excessive nitrogen supply, thus achieving precise control of crop growth morphology. If the crop leaf area is detected to be lower than the preset target value, the water and fertilizer system is immediately adjusted to increase the phosphorus and potassium supply. Phosphorus and potassium can promote crop leaf development and enhance the photosynthetic capacity of leaves. By precisely supplementing phosphorus and potassium, the crop leaf area can be effectively expanded, the crop photosynthetic efficiency can be improved, and the accumulation of nutrients for the crop can be guaranteed. The control continues until the crop growth morphology deviation is reduced to the preset allowable range, ensuring that the crop growth meets the expected standards of the training.

[0025] Furthermore, the feasibility verification of the generated corresponding control scheme also includes: If the equipment is not operating properly, adjust the equipment operating parameters to the rated range and replace the control commands that exceed the load. If the practical training fails to meet safety standards, delete the control instructions that pose safety hazards and supplement safety protection control measures; By integrating environmental control schemes and equipment operation control schemes, clarifying the execution sequence of each control scheme, a complete and executable horticultural smart production training control strategy is formed. The training program provides standardized operation guidance for the entire process of intelligent production in facility horticulture to trainees. The operation guidance covers greenhouse inspection, environmental parameter monitoring, equipment start-up and shutdown, water and fertilizer regulation, crop growth observation and pest and disease identification, operation specifications and precautions. The virtual twin scene, which is mapped one-to-one with the physical training site through a visual interface, can realize the synchronous and real-time visualization of the environmental parameters, equipment operation status, crop growth morphology and the operation actions of the trainees at the training site. Real-time collection of training operation data of trainees throughout the entire training process, including operation time, operation steps, equipment operation parameters, number of operation errors and operation completion time, synchronously recording the operation trajectory of trainees, forming a traceable data archive of the entire training operation process; Based on the collected data and results of the entire training operation, an automated scoring system is used to conduct intelligent assessments. The system combines the scoring results, operational error analysis, and the achievement of training objectives to generate a comprehensive evaluation report that includes strengths, weaknesses, and improvement suggestions.

[0026] In this embodiment, the operating status of the training equipment is first comprehensively tested. If the test reveals that the equipment's operating status is unqualified, the parameter adjustment process is immediately initiated to precisely adjust the equipment's operating parameters to the rated range. At the same time, the control commands that exceed the equipment's load are replaced. This measure can effectively prevent equipment failures and damage caused by abnormal parameters or overload, ensuring stable and safe operation of the equipment, extending the equipment's service life, and ensuring that the equipment's operating efficiency meets the training standards, providing hardware support for the smooth conduct of training. If the test reveals that the training safety is unqualified, the control commands with potential safety hazards are deleted immediately to prevent the expansion of safety risks. At the same time, targeted safety protection and control measures are added, such as adding equipment safety warnings and operation permission control, which can effectively prevent the occurrence of various safety accidents during training, ensure the personal safety of trainees and the integrity of training facilities and equipment, and build a solid training safety defense line. Subsequently, the previously developed environmental control and equipment operation control plans were systematically integrated. Combining the needs of the practical training conditions and crop growth patterns, the execution sequence, connection points, and priorities of each control plan were clarified to avoid conflicts between different control plans. This resulted in a complete, executable, and highly targeted horticultural smart production training control strategy, achieving synergistic linkage between environmental control and equipment operation, improving control efficiency and accuracy, and ensuring the training process closely matches actual production scenarios. Next, standardized operation guidance for the entire process of facility horticulture smart production was provided to trainees. The guidance comprehensively covered core aspects such as greenhouse inspection, environmental parameter monitoring, equipment start-up and shutdown operations, water and fertilizer control, crop growth observation and pest and disease identification, operating procedures, and precautions. This helped trainees quickly master standardized operating procedures, standardize operational behavior, reduce operational errors, improve trainees' practical skills and professional competence, and ensure the standardization and uniformity of training operations. Simultaneously, a virtual twin scene, mapped one-to-one with the physical training site, is displayed through a visual interface. Leveraging virtual twin technology, this allows for the synchronous, real-time visualization of environmental parameters (such as temperature, humidity, and light), equipment operating status (such as rotation speed and current), crop growth morphology (such as plant height and leaf area), and the trainees' operational actions. This technology enables training managers to intuitively and comprehensively grasp the entire training site, promptly identify anomalies during the training process, and facilitate rapid intervention and control. It also allows trainees to clearly understand the impact of their operations on the training system, enhancing the intuitiveness and interactivity of the training. Throughout the entire training process, real-time data on the trainees' entire training operation is collected, including core data such as operation time, operation steps, equipment operating parameters, number of operational errors, and operation completion time. The system also accurately records the trainees' operational trajectories, forming a traceable and queryable data archive of the entire training operation process. This data archive completely preserves the trainees' operational traces, providing real and comprehensive data support for subsequent intelligent assessment, operational error analysis, and optimization of training effects, ensuring the fairness and objectivity of the assessment results.Finally, based on the collected data from the entire training process and the final training results, an automated scoring system is used to conduct intelligent assessments. This eliminates the subjectivity and tediousness of traditional manual assessments, improving efficiency and accuracy. Simultaneously, by combining the scoring results, in-depth analysis of operational errors, and the achievement of training objectives, a comprehensive evaluation report is generated, including strengths, weaknesses, and targeted improvement suggestions. This report helps trainees clearly understand their strengths and weaknesses in practical operations, identify areas for improvement, and provides a scientific basis for optimizing training teaching and refining training programs, effectively enhancing the teaching quality and effectiveness of intelligent horticultural production training.

[0027] like Figure 3 The flowchart shown is a digital twin-based intelligent production training method for facility horticulture provided in this application embodiment. It includes: collecting facility horticulture data from the target facility horticulture production site and executing control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data, and crop growth status data; constructing a virtual three-dimensional scene that maps one-to-one with the intelligent horticulture production training module based on digital twin technology, and synchronizing the environmental parameters, equipment operating status, and crop growth morphology of the target facility horticulture production site in real time to achieve a virtual-real mapping between physical entities and the virtual scene; generating intelligent horticulture production training control strategies based on the simulation data of the virtual twin module and preset training conditions, and issuing control commands to the intelligent horticulture production training module to support condition simulation and virtual-real interactive control; providing trainees with standardized operation guidance for the entire process of intelligent facility horticulture production, synchronous real-time visualization of the virtual twin scene, data collection and trajectory recording throughout the training operation process, and automated scoring, intelligent assessment, and comprehensive evaluation based on operational behavior and training results.

[0028] In this embodiment, the core of intelligent production training in facility horticulture is to achieve deep integration between the physical production site and the virtual training scenario, providing trainees with realistic and efficient training support. The specific implementation steps and corresponding technical effects are as follows: First, the data acquisition module is activated to comprehensively collect various facility horticulture data from the target facility horticulture production site. The collection scope covers environmental data, equipment status data, and crop growth status data. Among them, environmental data includes key parameters such as air temperature and humidity, light intensity, and CO2 concentration. Equipment status data covers information such as the operating current, speed, power, and fault status of various control devices (sprayers, supplemental lighting, fans, water pumps, etc.). Crop growth status data includes growth indicators such as plant height, leaf area, and leaf color. Through comprehensive and thorough data acquisition, it can be ensured that the acquired data is comprehensive, realistic, and accurate, providing solid data support for subsequent virtual mapping, control strategy generation, and training assessment. At the same time, it enables real-time dynamic perception of the physical production site and timely captures changes in various parameters. Secondly, based on digital twin technology, a virtual 3D scene is constructed that maps one-to-one with the horticultural smart production training module. Relying on high-precision modeling and real-time data transmission technology, the environmental parameters, equipment operating status, and crop growth morphology of the target facility horticulture production site are collected and synchronized to the virtual 3D scene in real time, realizing a precise virtual-real mapping between physical entities and virtual scenes. This technology can break the limitations of physical space, allowing trainees to intuitively observe the entire process of facility horticulture production without having to be physically present at the actual production site. At the same time, the virtual scene can accurately replicate various changes in the physical site, ensuring the authenticity and professionalism of the training scene and providing realistic scene support for work condition simulation and operation practice. Next, combining the simulated data synchronously generated by the virtual twin module with the preset training conditions, the system's built-in algorithm model scientifically generates a horticultural smart production training control strategy that aligns with the training objectives and production realities. Subsequently, corresponding control commands are issued to the horticultural smart production training module. These commands cover various operational requirements such as environmental control, equipment operation, and water and fertilizer supply. This not only supports simulation exercises under different working conditions, allowing trainees to experience the control logic in different production scenarios, but also enables virtual-real interactive control. Trainees can issue commands through virtual scene operations, synchronously linking the physical training module or virtual module to execute corresponding operations, enhancing the interactivity and practicality of the training, and helping trainees quickly master the core skills of smart production control.Finally, comprehensive training support services are provided to trainees. On the one hand, standardized operation guidance for the entire process of intelligent production in facility horticulture is pushed out, clarifying the operation specifications and procedures for each link, such as greenhouse inspection, parameter monitoring, equipment operation, and water and fertilizer regulation, to help trainees standardize their operation behavior and reduce operational errors. On the other hand, a virtual twin scene is continuously provided for synchronous real-time visualization, allowing trainees to clearly see the synchronous changes between the physical site and the virtual scene and intuitively understand the impact of their own operations. At the same time, real-time collection of trainees' operation data throughout the entire training process is conducted, accurately recording the operation trajectory and forming a complete operation data archive to provide objective basis for subsequent assessments. Based on this, based on the collected operation behavior data and the final training results, automated scoring is completed through a preset scoring model to carry out intelligent assessments, eliminating the subjectivity and cumbersomeness of traditional manual assessments, improving assessment efficiency and fairness, and finally generating a comprehensive evaluation report that includes operational advantages, shortcomings, and targeted improvement suggestions, helping trainees to clarify the direction of improvement, effectively improve training results and professional skills, and achieve a deep integration of theory and practice.

Claims

1. A smart production training system for facility horticulture based on digital twins, characterized in that, include: The training modules include: intelligent horticulture production training module, virtual twin module, intelligent control module, and training and intelligent evaluation module. The intelligent horticulture production training module is used to collect facility horticulture data from the target facility horticulture production site and execute the control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data and crop growth status data. The virtual twin module is used to construct a virtual three-dimensional scene that maps one-to-one with the horticultural smart production training module based on digital twin technology, and to synchronize the environmental parameters, equipment operating status and crop growth morphology of the target facility horticultural production site in real time, so as to realize the virtual-real mapping between physical entities and virtual scenes. The intelligent control module is used to generate a horticultural intelligent production training control strategy based on the simulation data of the virtual twin module and the preset training conditions, and to issue control commands to the horticultural intelligent production training module to support the simulation of working conditions and the virtual-real interactive control. The practical training and intelligent assessment module is used to provide trainees with standardized operation guidance for the entire process of intelligent production in facility horticulture, synchronous real-time visualization of virtual twin scenes, data collection and trajectory recording of the entire practical operation process, as well as automated scoring, intelligent assessment and comprehensive evaluation based on operational behavior and practical training results.

2. The intelligent production training system for facility horticulture based on digital twins as described in claim 1, characterized in that: The environmental data includes air temperature, relative humidity, soil temperature, relative humidity, light intensity, carbon dioxide concentration, nutrient solution EC value, and nutrient solution pH value. The equipment status data includes the start / stop status, running time, operating current, operating voltage, and fault codes of the fan, water pump, shade net, window opener, and water-fertilizer integrated machine; The crop growth status data includes plant height, stem diameter, leaf area, number of leaves, growth level, degree of disease and pest occurrence, and yield prediction parameters.

3. The intelligent production training system for facility horticulture based on digital twins as described in claim 1, characterized in that: The specific process for constructing a virtual 3D scene that maps one-to-one with the intelligent horticulture production training module is as follows: Geometric parameters of the greenhouse frame size, enclosure structure, and spatial zoning are collected; equipment models, installation coordinates, and pipeline routes are marked; and spatial modeling of cultivation ridges and planting layout is performed to complete the digital restoration of the entity, mesh modeling, texture mapping, and scene assembly, forming a high-fidelity three-dimensional twin model that is completely consistent with the entity of the target facility horticulture production site. Parametric modeling is performed based on the geometric dimensions of the target facility horticulture production site. Corresponding material attributes are assigned according to the material characteristics of the entity. The spatial position of the three-dimensional model is determined according to the physical installation coordinates. After calibration, accuracy verification and consistency matching based on the digital twin model, a digital three-dimensional model that is completely identical to the physical entity is formed.

4. The intelligent production training system for facility horticulture based on digital twins as described in claim 1, characterized in that: The virtual twin module receives environmental parameters, equipment operating status and crop growth morphology data collected by the horticultural smart production training module in real time, and drives the virtual three-dimensional scene to be dynamically updated synchronously with the above data; The specific steps for performing dynamic updates are as follows: The virtual-real data synchronization verification interval represents the period during which the virtual twin module performs consistency comparisons between the physical data of the target facility horticulture production site and the virtual model data. A pre-constructed mapping relationship between the verification interval and the refresh time interval threshold is used to characterize the quantitative correspondence and linkage adjustment rules between the virtual and real data synchronization verification interval, the verification interval critical interval, the verification interval offset, the verification interval deviation, the interval threshold upward adjustment coefficient, the interval threshold downward adjustment coefficient, and the refresh time interval benchmark threshold. If the virtual and real data synchronization verification interval is within the preset verification interval critical range, then the refresh time interval baseline threshold is maintained. The preset verification interval critical range represents the closed interval formed by the preset verification interval critical lower limit and the preset verification interval critical upper limit.

5. The intelligent production training system for facility horticulture based on digital twins as described in claim 4, characterized in that: The dynamic updating also includes: If the virtual and real data synchronization verification interval is greater than the preset verification interval critical upper limit, the verification interval offset is input into the verification interval-refresh time interval threshold mapping relationship, and the interval threshold reduction coefficient is output. The refresh time interval base threshold and the interval threshold reduction coefficient are combined to obtain the target refresh time interval threshold. The verification interval offset represents the degree of positive deviation between the virtual and real data synchronization verification interval and the preset verification interval critical upper limit. If the virtual and real data synchronization verification interval is less than the preset verification interval critical lower limit, the verification interval deviation is input into the verification interval-refresh time interval threshold mapping relationship, and the interval threshold adjustment coefficient is output. The refresh time interval base threshold and the interval threshold adjustment coefficient are combined to obtain the target refresh time interval threshold. The verification interval deviation represents the degree of negative deviation between the virtual and real data synchronization verification interval and the preset verification interval critical lower limit.

6. The intelligent production training system for facility horticulture based on digital twins as described in claim 1, characterized in that: The specific steps for generating the intelligent horticultural production training and control strategy are as follows: By calling preset training conditions, the preset training conditions include normal production training conditions, extreme environment training conditions, equipment failure training conditions, and water and fertilizer imbalance training conditions. Each of the training conditions has preset corresponding target parameters, triggering conditions, and training assessment standards. The simulation data based on the virtual twin module is compared item by item with the target parameters of the currently enabled preset training conditions. The deviation values ​​between the simulation data and the target parameters of the currently enabled preset training conditions are analyzed to clarify the type and degree of deviation. The deviation types include environmental parameter deviation, equipment operating status deviation, and crop growth morphology deviation.

7. The intelligent production training system for facility horticulture based on digital twins as described in claim 6, characterized in that: The generated horticulture intelligent production training and control strategy also includes: The method generates corresponding control schemes based on the type and degree of deviation. The control schemes include environmental control schemes, equipment operation control schemes, and water and fertilizer supply control schemes. The environmental control schemes correspond to the adjustment of parameters such as air temperature and humidity and light intensity. The equipment operation control scheme corresponds to the start-up, shutdown, and parameter adjustment of equipment such as fans and water pumps; The water and fertilizer supply control scheme corresponds to the adjustment of fertilizer supply and water supply of the integrated water and fertilizer machine; The corresponding control schemes generated are tested for feasibility to ensure that they will not cause damage to physical equipment, abnormal crop growth, or safety risks to the training. If the test is passed, the various control schemes are integrated to form a complete horticultural smart production training control strategy. If the test fails, the control parameters are iteratively optimized and readjusted until the test is passed.

8. The intelligent production training system for facility horticulture based on digital twins as described in claim 7, characterized in that: The specific process for verifying the feasibility of the generated corresponding control scheme is as follows: If the deviation type is environmental parameter deviation, an environmental control plan is generated: when the air humidity is lower than the target air humidity value, the spray equipment is activated and the spray frequency is adjusted to the preset spray frequency reference value; when the light intensity is lower than the target light intensity value, the supplementary lighting equipment is activated and the supplementary lighting intensity is adjusted to the preset supplementary lighting intensity upper limit value until the environmental parameter deviation is reduced to the preset allowable range. If the deviation type is equipment operating status deviation, generate an equipment operation control plan: when the fan operating current is not within the preset rated current range, adjust the fan speed to calibrate the operating current to the preset rated current range. When the water pump fails to start as required by the operating conditions, a start command is issued and the water pump operating power is adjusted to ensure that the water supply meets the operating conditions standard; when the equipment displays a fault code, a shutdown and maintenance command is issued and the backup equipment is started to maintain the training operation. If the deviation type is crop growth morphology deviation, a water and fertilizer supply regulation plan is generated: when the plant height is lower than the target plant height value, the water and fertilizer integrated machine is adjusted to increase the nitrogen supply; when the leaf area is lower than the target leaf area value, the water and fertilizer integrated machine is adjusted to increase the phosphorus and potassium supply until the crop growth morphology deviation is reduced to the preset allowable range.

9. The intelligent production training system for facility horticulture based on digital twins as described in claim 8, characterized in that: The feasibility verification of the generated corresponding control scheme also includes: If the equipment is not operating properly, adjust the equipment operating parameters to the rated range and replace the control commands that exceed the load. If the practical training fails to meet safety standards, delete the control instructions that pose safety hazards and supplement safety protection control measures; By integrating environmental control schemes and equipment operation control schemes, clarifying the execution sequence of each control scheme, a complete and executable horticultural smart production training control strategy is formed. The training program provides standardized operation guidance for the entire process of intelligent production in facility horticulture to trainees. The operation guidance covers greenhouse inspection, environmental parameter monitoring, equipment start-up and shutdown, water and fertilizer regulation, crop growth observation and pest and disease identification, operation specifications and precautions. The virtual twin scene, which is mapped one-to-one with the physical training site through a visual interface, can realize the synchronous and real-time visualization of the environmental parameters, equipment operation status, crop growth morphology and the operation actions of the trainees at the training site. Real-time collection of training operation data of trainees throughout the entire training process, including operation time, operation steps, equipment operation parameters, number of operation errors and operation completion time, synchronously recording the operation trajectory of trainees, forming a traceable data archive of the entire training operation process; Based on the collected data and results of the entire training operation, an automated scoring system is used to conduct intelligent assessments. The system combines the scoring results, operational error analysis, and the achievement of training objectives to generate a comprehensive evaluation report that includes strengths, weaknesses, and improvement suggestions.

10. The method of applying the digital twin-based intelligent production training system for facility horticulture as described in any one of claims 1-9, characterized in that, include: Collect facility horticulture data from the target facility horticulture production site and execute control commands issued by the intelligent control module. The facility horticulture data includes environmental data, equipment status data, and crop growth status data. Based on digital twin technology, a virtual three-dimensional scene is constructed and mapped one-to-one with the horticultural smart production training module. The environmental parameters, equipment operation status and crop growth morphology of the target facility horticultural production site are synchronized in real time to realize the virtual-real mapping between physical entities and virtual scenes. Based on the simulation data of the virtual twin module and the preset training conditions, a horticultural intelligent production training control strategy is generated, and control commands are sent to the horticultural intelligent production training module to support the simulation of working conditions and the interaction between the virtual and real systems. It provides trainees with standardized operation guidance for the entire process of intelligent production in facility horticulture, synchronous real-time visualization of virtual twin scenes, data collection and trajectory recording of the entire training operation process, as well as automated scoring, intelligent assessment and comprehensive evaluation based on operation behavior and training results.

Images