A digital twin-based intelligent inspection system and method for chemical laboratories

Abstract

This invention provides an intelligent inspection system and method for chemical laboratories based on digital twins, belonging to the field of industrial Internet of Things and intelligent safety management technology. The system includes a physical sensing layer, a digital twin engine layer, an intelligent processing layer, and an application interaction layer. The physical sensing layer collects real-time data of the laboratory environment and equipment through a sensor network; the digital twin engine layer constructs a three-dimensional virtual model corresponding to the physical laboratory and displays the data overlay; the intelligent processing layer dynamically plans inspection paths and generates alarms based on real-time and historical data through anomaly pattern recognition and reinforcement learning; the application interaction layer provides a visual inspection interface and safety emergency handling functions. The method includes data acquisition, 3D driving, intelligent analysis, and visualization steps. This invention achieves global perception of laboratory safety status, intelligent risk early warning, and automation of the inspection process, improving the real-time performance, accuracy, and intelligence level of chemical laboratory safety management.

Description

Technical Field

[0001] This invention relates to the field of industrial Internet of Things and intelligent safety management technology, specifically to an intelligent inspection system and method for chemical laboratories based on digital twins. Background Technology

[0002] In the field of safety management in chemical laboratories, traditional inspection methods mainly rely on manual, scheduled patrols. This method not only lacks real-time accuracy, making it difficult to detect safety hazards such as leaks and abnormal temperature control in a timely manner, but also poses a direct threat to the personal safety of inspection personnel in complex chemical environments with flammable and explosive properties. Furthermore, manual recording methods are inefficient, prone to data errors, and difficult to trace.

[0003] Existing laboratory monitoring systems are mostly decentralized and independent video surveillance or two-dimensional data charts, lacking a three-dimensional reconstruction and unified display of the overall laboratory space status. These systems cannot effectively integrate and analyze information such as equipment status, environmental parameters, and the location of hazardous materials, making it difficult for managers to obtain an intuitive and comprehensive safety situational awareness and failing to support refined safety management decisions.

[0004] In recent years, digital twin technology has provided a new approach to the virtualized monitoring of physical entities. However, some existing digital twin inspection systems focus primarily on model visualization, exhibiting significant shortcomings in model accuracy, deep integration with real-time data, and intelligent prediction and coordinated response to complex risks. Their applications often remain at the static display stage, failing to fully leverage the core value of digital twins in risk warning, emergency command, and intelligent decision-making. Summary of the Invention

[0005] In view of the above-mentioned technical problems in related technologies, the present invention proposes a chemical laboratory intelligent inspection system and method based on digital twins, which can overcome the above-mentioned shortcomings of the prior art.

[0006] To achieve the above-mentioned technical objectives, the technical solution of the present invention is implemented as follows: A digital twin-based intelligent inspection system for chemical laboratories; This digital twin-based intelligent inspection system for chemical laboratories includes: The physical sensing layer includes a sensor network and a data acquisition module. The sensor network is used to monitor physical and chemical safety parameters of the laboratory environment, and the data acquisition module is used to acquire real-time monitoring data from the sensor network. The digital twin engine layer is communicatively connected to the physical perception layer and is used to construct and render a three-dimensional virtual model corresponding to the physical laboratory space based on the real-time monitoring data, and to overlay and display the real-time monitoring data on the three-dimensional virtual model. The intelligent processing layer, which is communicatively connected to the digital twin engine layer, includes an intelligent inspection planning algorithm module and an alarm module. The intelligent inspection planning algorithm module is used to plan inspection paths or identify abnormal features based on the real-time monitoring data and historical data. The alarm module is used to generate alarm information when an anomaly is detected. An application interaction layer, which is communicatively connected to the intelligent processing layer, is used to provide a visual inspection interface. The visual inspection interface includes at least a display area for displaying the three-dimensional virtual model and overlay data.

[0007] Furthermore, the sensor network includes at least one of a temperature sensor, a humidity sensor, a smoke sensor, a volatile organic compound concentration sensor, a pressure sensor, a tilt sensor, a displacement sensor, and a water immersion sensor; and / or, The digital twin engine layer includes: A 3D scene rendering module is used to load and render the 3D virtual model, and establish the coordinate mapping relationship between the 3D virtual model and the physical laboratory space; and The real-time data monitoring panel module is used to overlay a transparent two-dimensional interactive layer on the three-dimensional virtual model to display the real-time monitoring data.

[0008] Furthermore, the intelligent inspection planning algorithm module is specifically used for: An anomaly pattern library is constructed or updated based on historical monitoring data. The anomaly pattern library defines at least one risk pattern among gradual equipment failure, chemical leakage and diffusion, association with non-compliance operations, and environmental stability imbalance. The real-time monitoring data is matched with the anomaly pattern library to identify anomaly features or risk patterns. Based on the identified anomaly features, risk patterns, and the real-time risk situation of the laboratory, inspection routes are dynamically planned.

[0009] Furthermore, the intelligent inspection planning algorithm module integrates a reinforcement learning model, which is used to dynamically generate inspection paths based on a state space that includes real-time risk values, equipment health, and personnel density factors, as well as a reward mechanism that includes rewards for successfully identifying anomalies and covering high-risk areas.

[0010] Furthermore, the alarm module supports a multi-level alarm mechanism, and the alarm methods include at least one of the following: providing dynamic visual prompts in the 3D virtual model of the digital twin engine layer, sending text messages, sending emails, and pushing information to mobile applications.

[0011] Furthermore, the visual inspection interface of the application interaction layer also includes: The floating label display module is used to display associated sensor parameters in real time above the device objects in the 3D virtual model; The dynamic dashboard module is used to display trend analysis of environmental indicators or statistics on equipment operating status in chart form; and The multimedia linkage module is used to respond to interactive operations on the camera model object in the three-dimensional virtual model, and to retrieve and display the corresponding real-time video monitoring screen.

[0012] Furthermore, the application interaction layer also includes a security emergency handling module, which performs the following operations upon receiving alarm information from the alarm module: According to the preset emergency plan, environmental control, channel management or equipment isolation instructions are automatically or assisted to be issued; Emergency evacuation routes are generated and displayed in the three-dimensional virtual model, and the route information is pushed to the terminal devices of relevant personnel; Initiate a remote expert collaboration session to enable remote experts to provide annotation guidance in the 3D virtual model or on-site first-person view.

[0013] Furthermore, the system also integrates an augmented reality-assisted inspection module, which is used to overlay and display device information, real-time parameters, maintenance manuals, or annotation and guidance information from remote experts in the field of vision of the augmented reality device worn by the inspection personnel.

[0014] According to another aspect of the present invention, a method for intelligent inspection of chemical laboratories based on digital twins is provided; This intelligent inspection method for chemical laboratories based on digital twins includes: Real-time monitoring data of the laboratory environment is collected through the physical sensing layer; Through the digital twin engine layer, a three-dimensional virtual model corresponding to the physical laboratory space is driven based on the real-time monitoring data, and the real-time monitoring data is superimposed and displayed on the three-dimensional virtual model; Through the intelligent processing layer, intelligent inspection planning is executed based on the real-time monitoring data and historical data to identify abnormal features, and alarm information is generated when an abnormality is identified. The interactive layer is used to display the three-dimensional virtual model, overlaid real-time monitoring data, and alarm information in the visual inspection interface.

[0015] Furthermore, the step of executing intelligent inspection planning based on the real-time monitoring data and historical data through the intelligent processing layer to identify abnormal characteristics includes: Build or update the abnormal pattern library based on historical monitoring data; Based on the real-time monitoring data and the abnormal pattern library, the current abnormal characteristics or risk patterns are identified through matching and analysis. Based on a reinforcement learning model, the inspection path is dynamically planned according to the state space consisting of real-time risk value, equipment health and personnel density factors, and a preset reward mechanism.

[0016] The beneficial effects of this invention are as follows: By constructing a three-dimensional digital twin that is precisely mapped to the physical laboratory and integrating multi-source sensor data and intelligent analysis algorithms, the laboratory environment can be monitored in a panoramic, visualized, and real-time manner, and abnormal risks can be automatically identified and assessed. This enables intelligent planning and execution of inspection tasks, as well as rapid response and precise handling of safety hazards, ultimately achieving the goal of significantly improving the initiative, refinement, and overall safety effectiveness of chemical laboratory safety management. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the 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.

[0018] Figure 1 This is an architecture diagram of a digital twin-based intelligent inspection system for chemical laboratories according to an embodiment of the present invention. Detailed Implementation

[0019] 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. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0020] like Figure 1 As shown in the figure, a digital twin-based intelligent inspection system for chemical laboratories according to an embodiment of the present invention includes: The physical sensing layer includes a sensor network and a data acquisition module. The sensor network is used to monitor physical and chemical safety parameters of the laboratory environment, and the data acquisition module is used to acquire real-time monitoring data from the sensor network. The digital twin engine layer is communicatively connected to the physical perception layer and is used to construct and render a three-dimensional virtual model corresponding to the physical laboratory space based on the real-time monitoring data, and to overlay and display the real-time monitoring data on the three-dimensional virtual model. The intelligent processing layer, which is communicatively connected to the digital twin engine layer, includes an intelligent inspection planning algorithm module and an alarm module. The intelligent inspection planning algorithm module is used to plan inspection paths or identify abnormal features based on the real-time monitoring data and historical data. The alarm module is used to generate alarm information when an anomaly is detected. An application interaction layer, which is communicatively connected to the intelligent processing layer, is used to provide a visual inspection interface. The visual inspection interface includes at least a display area for displaying the three-dimensional virtual model and overlay data.

[0021] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for a chemical laboratory, in a specific embodiment, includes a sensor network comprising at least one of a temperature sensor, a humidity sensor, a smoke sensor, a volatile organic compound concentration sensor, a pressure sensor, a tilt sensor, a displacement sensor, and a water immersion sensor; and / or, The digital twin engine layer includes: A 3D scene rendering module is used to load and render the 3D virtual model, and establish the coordinate mapping relationship between the 3D virtual model and the physical laboratory space; and The real-time data monitoring panel module is used to overlay a transparent two-dimensional interactive layer on the three-dimensional virtual model to display the real-time monitoring data.

[0022] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for a chemical laboratory, in a specific implementation, wherein the intelligent inspection planning algorithm module is specifically used for: An anomaly pattern library is constructed or updated based on historical monitoring data. The anomaly pattern library defines at least one risk pattern among gradual equipment failure, chemical leakage and diffusion, association with non-compliance operations, and environmental stability imbalance. The real-time monitoring data is matched with the anomaly pattern library to identify anomaly features or risk patterns. Based on the identified anomaly features, risk patterns, and the real-time risk situation of the laboratory, inspection routes are dynamically planned.

[0023] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for a chemical laboratory, in a specific implementation, the intelligent inspection planning algorithm module integrates a reinforcement learning model, which is used to dynamically generate inspection paths based on a state space that includes factors such as real-time risk values, equipment health, and personnel density, as well as a reward mechanism that includes rewards for successfully identifying anomalies and covering high-risk areas.

[0024] According to an embodiment of the present invention, a chemical laboratory intelligent inspection system based on digital twins is provided. In a specific implementation, the alarm module supports a multi-level alarm mechanism. The alarm methods include at least one of the following: dynamic visual prompts in the three-dimensional virtual model of the digital twin engine layer, sending text messages, sending emails, and pushing information to mobile applications.

[0025] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for a chemical laboratory, in a specific implementation, further includes the following in the visual inspection interface of the application interaction layer: The floating label display module is used to display associated sensor parameters in real time above the device objects in the 3D virtual model; The dynamic dashboard module is used to display trend analysis of environmental indicators or statistics on equipment operating status in the form of charts; The multimedia linkage module is used to respond to interactive operations on the camera model object in the three-dimensional virtual model, and to retrieve and display the corresponding real-time video monitoring screen.

[0026] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for chemical laboratories, in a specific implementation, further includes a safety emergency handling module in the application interaction layer. This safety emergency handling module performs the following operations upon receiving an alarm message from the alarm module: According to the preset emergency plan, environmental control, channel management or equipment isolation instructions are automatically or assisted to be issued; Emergency evacuation routes are generated and displayed in the three-dimensional virtual model, and the route information is pushed to the terminal devices of relevant personnel; Initiate a remote expert collaboration session to enable remote experts to provide annotation guidance in the 3D virtual model or on-site first-person view.

[0027] According to an embodiment of the present invention, a digital twin-based intelligent inspection system for chemical laboratories is provided. In a specific embodiment, the system further integrates an augmented reality-assisted inspection module, which is used to overlay and display equipment information, real-time parameters, maintenance manuals, or annotation and guidance information from remote experts in the field of vision of the augmented reality device worn by the inspector.

[0028] Secondly, according to an embodiment of the present invention, a method for intelligent inspection of a chemical laboratory based on digital twins, the method includes: Real-time monitoring data of the laboratory environment is collected through the physical sensing layer; Through the digital twin engine layer, a three-dimensional virtual model corresponding to the physical laboratory space is driven based on the real-time monitoring data, and the real-time monitoring data is superimposed and displayed on the three-dimensional virtual model; Through the intelligent processing layer, intelligent inspection planning is executed based on the real-time monitoring data and historical data to identify abnormal characteristics, and alarm information is generated when an anomaly is detected; and The interactive layer is used to display the three-dimensional virtual model, overlaid real-time monitoring data, and alarm information in the visual inspection interface.

[0029] According to an embodiment of the present invention, a method for intelligent inspection of a chemical laboratory based on digital twins, in a specific implementation, includes the step of performing intelligent inspection planning based on real-time monitoring data and historical data through an intelligent processing layer to identify abnormal features, including: Build or update the abnormal pattern library based on historical monitoring data; Based on the real-time monitoring data and the abnormal pattern library, the current abnormal characteristics or risk patterns are identified through matching and analysis. Based on a reinforcement learning model, the inspection path is dynamically planned according to the state space consisting of real-time risk value, equipment health and personnel density factors, and a preset reward mechanism.

[0030] To facilitate understanding of the above technical solutions of the present invention, the following detailed description of the above technical solutions of the present invention will be provided through specific usage methods.

[0031] The following detailed description, in conjunction with embodiments and corresponding technical details, further illustrates the intelligent inspection system and method for chemical laboratories based on digital twins provided by the present invention.

[0032] Example 1: This embodiment provides an intelligent inspection system for a chemical laboratory based on digital twins. The system includes a physical perception layer, a digital twin engine layer, an intelligent processing layer, and an application interaction layer.

[0033] The physical sensing layer includes a sensor network and a data acquisition module. The sensor network consists of various types of sensors deployed within the laboratory, such as temperature sensors, humidity sensors, smoke sensors, volatile organic compound concentration sensors, pressure sensors, tilt sensors, displacement sensors, and water immersion sensors, used to comprehensively monitor the physical and chemical safety parameters of the laboratory environment. The data acquisition module is responsible for communicating with the sensor network, for example, using the MQTT protocol to interface with the aforementioned environmental sensors and the HTTP protocol to interface with devices such as intelligent reagent cabinets, to achieve the collection and aggregation of various real-time monitoring data.

[0034] The digital twin engine layer is communicatively connected to the physical perception layer. Based on technologies such as WebGL / Three.js, this layer constructs a three-dimensional virtual model corresponding to the physical laboratory space at a 1:1 scale. It specifically includes a three-dimensional scene rendering module and a real-time data monitoring panel module. The three-dimensional scene rendering module is responsible for loading the 3D model files of the laboratory, completing scene parsing and loading, and establishing an accurate spatial coordinate mapping relationship between the virtual scene and the physical space. The real-time data monitoring panel module is used to overlay a transparent two-dimensional interaction layer on the rendered three-dimensional scene, thereby realizing the "virtual-real combination" data display. For example, the dynamic vision drive function can change the appearance of the model according to real-time data, play the corresponding animation when the fume hood is opened, or change the color of the model in a certain area to red and flash when the temperature and humidity in that area exceed the standard, so as to provide an intuitive visual warning.

[0035] The intelligent processing layer is communicatively connected to the digital twin engine layer, and its core includes an intelligent patrol planning algorithm module and an alarm module. The intelligent patrol planning algorithm module extracts and identifies abnormal features based on historical patrol data and real-time monitoring data, in combination with time series prediction models or machine learning algorithms, and can be used to plan the patrol path. The alarm module is used to generate alarm information when an abnormality is identified and supports a multi-level alarm mechanism. The alarm methods include dynamic red flashing prompts in the three-dimensional virtual model of the digital twin engine layer, and notifications sent through multiple channels such as text messages, emails, and mobile application push. When an abnormality is detected, the system can automatically associate the three-dimensional model, video surveillance footage, and dangerous goods ledger data for multi-dimensional linkage display.

[0036] The application interaction layer is communicatively connected to the intelligent processing layer and is used to provide an interface for user operations, mainly including a visual patrol interface. The visual patrol interface at least includes a display area for displaying the three-dimensional virtual model and the superimposed data. Further, this interface can also integrate a floating label display module for real-time display of associated sensor parameter labels above the device objects in the three-dimensional model; a dynamic dashboard module for drawing real-time trend curve graphs of environmental indicators, pie charts of equipment operation rates, etc. through tools such as ECharts; and a multimedia linkage module for supporting the pop-up of the corresponding real-time video surveillance footage when the user clicks on the camera model in the three-dimensional space to achieve integrated patrol.

[0037] In addition, the system can also be provided with a laboratory resource management and configuration module for realizing the association of equipment assets and three-dimensional model components, the visualization management of the storage locations of dangerous chemicals, and the configuration of system operation permissions for different role users.

[0038] Embodiment 2: This embodiment details a specific implementation process of an intelligent inspection method for chemical laboratories using the aforementioned system, mainly including the following steps: Step S1: Establish a 3D visualization model of the laboratory. Using digital twin technology, construct a lightweight 3D virtual model of the target chemical laboratory at a 1:1 scale. A spatial topology network can be built based on the laboratory layout, defining key equipment, inspection points, etc., as nodes to form a basic road network map supporting virtual inspections.

[0039] Step S2: Establish a real-time data monitoring system. Using the data acquisition module of the physical sensing layer, and employing protocols such as MQTT and HTTP, real-time data from sensors in the laboratory, including temperature, humidity, smoke, and VOC concentration, as well as the status data of the intelligent reagent cabinet, are collected. The collected data is received via a message queue and undergoes necessary smoothing processing to drive updates to the corresponding equipment status in the 3D virtual model.

[0040] Step S3: Execute intelligent inspection planning and anomaly identification. The intelligent inspection planning algorithm module of the intelligent processing layer operates based on historical data and real-time monitoring data. For example, the ARIMA model can be used to predict equipment data trends. When the predicted value exceeds a preset safety threshold, it is confirmed that the equipment has a risk of failure. At the same time, the algorithm can perform matching analysis on real-time data based on the built-in anomaly pattern library to identify risk patterns such as chemical leaks and spread, and associations with violations.

[0041] Step S4: Perform intelligent inspection and visualization. The system can virtually roam in the digital twin space according to a preset or dynamically planned path to check the data of each monitoring point. When the data of a certain sensor exceeds a threshold, the system automatically switches the three-dimensional view to the abnormal coordinate point and can retrieve the associated real-time video monitoring footage. All data is displayed through the visualization interface of the application interaction layer, including overlaying data panels on the three-dimensional scene and drawing trend charts.

[0042] Step S5: Alarm and Fault Handling. Upon identifying an anomaly or predicting a fault risk, the alarm module triggers multi-level alarms. Simultaneously, the system can initiate a safety emergency response process. For example, based on the road network model, the optimal path from the virtual inspector's current location to the risk point is calculated, and the system guides the response. Relevant alarm information, handling suggestions, and evacuation routes can be pushed to the mobile terminals of safety personnel.

[0043] Example 3: This embodiment further elaborates on the core design of the system from the perspective of functional modules, especially detailing the algorithm design of the intelligent processing layer.

[0044] The intelligent inspection planning algorithm module, as the core decision engine, is designed with a multi-layered architecture: The first layer is scenario-based feature recognition based on a "chemical laboratory anomaly pattern library." This library defines various typical risk patterns, such as gradual equipment failure, chemical spill and diffusion, correlation of unauthorized operations, and environmental stability imbalance. The algorithm combines models such as ARIMA and LSTM to learn the normal baseline of the equipment and achieves early and accurate risk identification through methods such as monitoring data drift, multi-sensor concentration gradient correlation, and operation and status linkage analysis.

[0045] The second layer is dynamic risk path planning based on reinforcement learning. The algorithm meshes the laboratory's digital twin space, defines a state space that includes factors such as real-time risk values, equipment health, and personnel density, and designs a multi-objective reward function with positive rewards for successfully identifying anomalies and covering high-risk areas. Through a reinforcement learning model, it calculates and generates dynamic inspection paths that maximize overall safety benefits in real time, rather than fixed shortest paths.

[0046] The third layer involves the closed-loop self-evolution and knowledge accumulation of the algorithm. The system drives incremental optimization of the model by recording alarm feedback results and uses unsupervised learning to mine potential new patterns in historical data. After confirmation, these patterns are expanded to the abnormal pattern library, enabling the system to continuously learn and adapt to different scenarios.

[0047] The alarm module implements a multi-level alarm mechanism to ensure that abnormal information can reach relevant personnel in a timely manner through various means such as dynamic prompts in the 3D scene, SMS, email, and APP push.

[0048] The application interaction layer provides a visual inspection interface that integrates functions such as floating tags, dynamic dashboards, and multimedia linkage, offering users an immersive monitoring and operation experience. The safety emergency response function can be activated during major risks to execute or assist in the execution of emergency plans such as automatically controlling ventilation systems, displaying evacuation routes, and initiating remote expert AR collaboration.

[0049] The laboratory resource management and configuration module supports the association of equipment assets, visualization of hazardous materials storage locations, and fine-grained user permission management of the system base, ensuring the accuracy of system data and operational safety.

[0050] In summary, by utilizing the technical solution of this invention, a three-dimensional digital twin precisely mapped to the physical laboratory is constructed, and multi-source sensor data and intelligent analysis algorithms are integrated. This enables panoramic, visualized, and real-time monitoring of the laboratory environment, and automated identification and assessment of abnormal risks. Consequently, intelligent planning and execution of inspection tasks are achieved, as well as rapid response and precise handling of safety hazards. Ultimately, this significantly enhances the initiative, refinement, and overall safety effectiveness of chemical laboratory safety management.

[0051] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A digital twin-based intelligent inspection system for chemical laboratories, characterized in that, include: The physical sensing layer includes a sensor network and a data acquisition module. The sensor network is used to monitor physical and chemical safety parameters of the laboratory environment, and the data acquisition module is used to acquire real-time monitoring data from the sensor network. The digital twin engine layer is communicatively connected to the physical perception layer and is used to construct and render a three-dimensional virtual model corresponding to the physical laboratory space based on the real-time monitoring data, and to overlay and display the real-time monitoring data on the three-dimensional virtual model. The intelligent processing layer, which is communicatively connected to the digital twin engine layer, includes an intelligent inspection planning algorithm module and an alarm module; the intelligent inspection planning algorithm module is used to plan inspection paths or identify abnormal features based on the real-time monitoring data and historical data. The alarm module is used to generate alarm information when an anomaly is detected; An application interaction layer, which is communicatively connected to the intelligent processing layer, is used to provide a visual inspection interface. The visual inspection interface includes at least a display area for displaying the three-dimensional virtual model and overlay data.

2. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 1, characterized in that, The sensor network includes at least one of the following: temperature sensor, humidity sensor, smoke sensor, volatile organic compound concentration sensor, pressure sensor, tilt sensor, displacement sensor, and water immersion sensor; and / or, The digital twin engine layer includes: A 3D scene rendering module is used to load and render the 3D virtual model, and establish the coordinate mapping relationship between the 3D virtual model and the physical laboratory space; and The real-time data monitoring panel module is used to overlay a transparent two-dimensional interactive layer on the three-dimensional virtual model to display the real-time monitoring data.

3. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 1, characterized in that, The intelligent inspection planning algorithm module is specifically used for: An anomaly pattern library is constructed or updated based on historical monitoring data. The anomaly pattern library defines at least one risk pattern among gradual equipment failure, chemical leakage and spread, association with non-compliance operations, and environmental stability imbalance. The real-time monitoring data is matched with the abnormal pattern library to identify abnormal features or risk patterns. And based on the identified abnormal features, risk patterns, and real-time risk status of the laboratory, the inspection route is dynamically planned.

4. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 3, characterized in that, The intelligent inspection planning algorithm module integrates a reinforcement learning model, which dynamically generates inspection paths based on a state space that includes real-time risk values, equipment health, and personnel density factors, as well as a reward mechanism that includes rewards for successfully identifying anomalies and covering high-risk areas.

5. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 1, characterized in that, The alarm module supports a multi-level alarm mechanism, and the alarm methods include at least one of the following: providing dynamic visual prompts in the 3D virtual model of the digital twin engine layer, sending text messages, sending emails, and pushing information to mobile applications.

6. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 1, characterized in that, The visual inspection interface of the application interaction layer also includes: The floating label display module is used to display associated sensor parameters in real time above the device objects in the 3D virtual model; The dynamic dashboard module is used to display trend analysis of environmental indicators or statistics on equipment operating status in chart form; and The multimedia linkage module is used to respond to interactive operations on the camera model object in the three-dimensional virtual model, and to retrieve and display the corresponding real-time video monitoring screen.

7. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 1, characterized in that, The application interaction layer also includes a security emergency handling module, which performs the following operations upon receiving an alarm message from the alarm module: According to the preset emergency plan, environmental control, channel management or equipment isolation instructions are automatically or assisted to be issued; Emergency evacuation routes are generated and displayed in the three-dimensional virtual model, and the route information is pushed to the terminal devices of relevant personnel; Initiate a remote expert collaboration session to enable remote experts to provide annotation guidance in the 3D virtual model or on-site first-person view.

8. The intelligent inspection system for a chemical laboratory based on digital twins according to claim 7, characterized in that, The system also integrates an augmented reality-assisted inspection module, which overlays and displays device information, real-time parameters, maintenance manuals, or annotations and guidance from remote experts in the field of vision of the augmented reality device worn by the inspectors.

9. A method for intelligent inspection of a chemical laboratory based on digital twins, characterized in that, The method, applied to the system as described in any one of claims 1-8, comprises: Real-time monitoring data of the laboratory environment is collected through the physical sensing layer; Through the digital twin engine layer, a three-dimensional virtual model corresponding to the physical laboratory space is driven based on the real-time monitoring data, and the real-time monitoring data is superimposed and displayed on the three-dimensional virtual model; Through the intelligent processing layer, intelligent inspection planning is executed based on the real-time monitoring data and historical data to identify abnormal features, and alarm information is generated when an abnormality is identified. The interactive layer is used to display the three-dimensional virtual model, overlaid real-time monitoring data, and alarm information in the visual inspection interface.

10. A method for intelligent inspection of a chemical laboratory based on digital twins according to claim 9, characterized in that, The intelligent inspection plan, executed by the intelligent processing layer based on the real-time monitoring data and historical data to identify abnormal features, includes: Build or update the abnormal pattern library based on historical monitoring data; Based on the real-time monitoring data and the abnormal pattern library, the current abnormal characteristics or risk patterns are identified through matching and analysis. Based on a reinforcement learning model, the inspection path is dynamically planned according to the state space consisting of real-time risk value, equipment health and personnel density factors, and a preset reward mechanism.

Images