Intelligent inspection system for particle accelerator facility

Abstract

The invention relates to an intelligent inspection system for a particle accelerator facility, and the system comprises a multi-mode sensor unit which is used for collecting the temperature, audio, image, water leakage and other data of the accelerator facility; the data processing and analyzing unit performs cleaning and feature extraction on the collected data, constructs a fault model by using an artificial intelligence algorithm, and realizes fault early warning and diagnosis; the inspection report generation unit is used for generating an inspection report containing fault information and processing suggestions according to the analysis result; and the data storage and management unit is used for normatively storing the inspection data. Accelerator facility states are comprehensively sensed through a multi-mode sensor, data are deeply analyzed through an artificial intelligence algorithm, the system can accurately monitor equipment operation states in real time, potential faults are found in advance, early warning is carried out in time, and fault types and positions are accurately diagnosed. The method effectively reduces the labor cost, standardizes the data management, provides a scientific decision basis for operation and maintenance personnel, and improves the operation and maintenance level and operation efficiency of the particle accelerator.

Description

Technical Field

[0001] The present invention relates to an intelligent inspection system for particle accelerator facilities, in particular to an intelligent inspection system for particle accelerator facilities based on multimodal sensors and artificial intelligence technology. Background Art

[0002] Particle accelerators are crucial for scientific research and industrial applications, and the stable operation of their facilities is fundamental to ensuring accelerator performance. Traditional accelerator facility inspections rely on manual labor, which presents significant drawbacks. On the one hand, manual inspections are labor-intensive and time-consuming. Due to the widespread distribution of accelerator facilities and long inspection cycles, real-time monitoring is difficult to achieve. Sudden equipment failures are difficult to detect and address promptly, leading to downtime and impacting scientific research progress. On the other hand, manual inspections are limited by personnel experience and expertise, resulting in inaccurate fault diagnosis. Furthermore, data recording is irregular, prone to loss or errors, and hinders long-term analysis of equipment operating status.

[0003] With the development of technologies such as the Internet of Things, big data, and artificial intelligence, various industries are transitioning towards intelligent systems. In the field of particle accelerators, the development of intelligent inspection systems has become a trend. Multimodal sensors can collect a variety of data from accelerator facilities, including temperature, audio, images, and water leaks, providing information for a comprehensive understanding of equipment status. Artificial intelligence technology can deeply analyze massive amounts of data, uncovering failure patterns and patterns, and achieve early warning and diagnosis of faults through techniques such as fault modeling and data analysis. Therefore, there is an urgent need to develop intelligent inspection systems for particle accelerator facilities, leveraging advanced technologies to enhance operational maintenance, ensure stable and reliable accelerator operation, and overcome the shortcomings of traditional manual inspections. Summary of the Invention

[0004] In view of the limitations of traditional manual inspection methods for particle accelerator facilities, the present invention aims to provide an intelligent inspection system for particle accelerator facilities based on multimodal sensors and artificial intelligence technology.

[0005] The technical solution adopted by the present invention is: an intelligent inspection system for particle accelerator facilities, including a front-end intelligent device, an edge computing device, a server and a client;

[0006] The front-end intelligent devices are distributed at various facilities of the particle accelerator and are used to collect data from the particle accelerator facilities in real time, including temperature, audio, images, and water leakage, to achieve real-time dynamic monitoring of the particle accelerator facilities;

[0007] The edge computing device is responsible for pre-processing and preliminary analysis of real-time data transmitted from the front-end intelligent device, including audio data processing and analysis, infrared temperature measurement data processing and analysis, and can detect abnormal events in a timely manner and report abnormal information to the server;

[0008] The server receives abnormal information reported by front-end intelligent devices and edge computing devices, uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information, and determines the type of fault, scope of impact, and potential risks by establishing fault models and using data analysis and mining technologies. It then makes decisions based on the analysis results and pushes alarm information.

[0009] The client is an interface for users to interact with the system. Users can view real-time data, historical data and abnormal information of the particle accelerator facility through the client.

[0010] The front-end intelligent equipment includes an infrared temperature measuring camera, an intelligent ball camera and a water leakage transmitter;

[0011] The infrared temperature camera's thermal imaging sensor is a vanadium oxide uncooled detector with a thermal imaging pixel size of 256×192, a temperature measurement range of -20°C to 150°C, and supports POE power supply.

[0012] The smart dome camera is equipped with a 360° pan / tilt, 23x optical zoom, 100m high-definition infrared night vision, a maximum image size of 2560×1440, a built-in microphone, and supports POE power supply;

[0013] The water leakage transmitter supports positioning function with a positioning resolution of 0.1m, a maximum detection distance of 150m, a detection accuracy of 0.5m±1%FS, and also has functions such as disconnection alarm and interference alarm.

[0014] The edge computing device includes an industrial computer running the Windows or Linux operating system, which is responsible for collecting real-time data transmitted by front-end intelligent devices and pre-processing and analyzing this data; combining historical data and advanced algorithm models, it can detect abnormal events in a timely manner and report the abnormal information to the server through EPICSPV.

[0015] The edge computing device is connected to the smart dome camera through a network, collecting the audio data of the smart dome camera in real time, performing pre-processing such as noise suppression, format conversion and compression on the collected audio to improve data quality, and then using support vector machines, deep learning and other algorithms to extract audio features and match them with pre-trained fault models. Once an abnormality is detected, the alarm information will be immediately reported to the server through EPICSPV.

[0016] The edge computing device connects to the infrared temperature measurement camera through the network and collects infrared temperature measurement data in real time. It uses the PyEpics library to convert the original infrared temperature measurement data into EPICSPV for publication. It adopts a sliding window algorithm to filter repeated infrared temperature measurement alarms in a short period of time. Based on the alarm history database, it mines the false alarm correlation factors and optimizes the alarm threshold parameters. Once an abnormality is detected, the alarm information is immediately reported to the server through EPICSPV.

[0017] The server includes various service programs running in the upper-level server, which is responsible for receiving abnormal information reported by front-end intelligent devices and edge computing devices, and using advanced artificial intelligence algorithms to train and model this information, conduct in-depth analysis, make decisions based on the analysis results, and push alarm information; the server also has data storage, query and statistical analysis functions to provide decision support for management.

[0018] The server adopts a microservice architecture to ensure high availability, scalability and flexibility; it receives abnormal information sent by front-end intelligent devices and edge computing devices through the network, and verifies and cleans the received data; it then calls artificial intelligence algorithms to conduct in-depth analysis of the abnormal information, and through the establishment of fault models, data analysis and mining technologies, it determines the fault type, impact scope and potential risks, and makes decisions based on the analysis results, and publishes the alarm information as EPICSPV.

[0019] The server uses the open source software Phoebus alarm service program released by the EPICS community to build a real-time alarm push service, realizing functions such as WeChat classification push, front-end web page display, and alarm information historical data query; it uses the high-performance and easy-to-use Archiver Appliance to build a historical data archiving system, storing millions of PVs, with fast data query speed and can be configured PV by PV.

[0020] The client mainly uses the new generation control software platform Phoebus. Users can view the real-time data, historical data and abnormal information of the particle accelerator facility through the client; the client also has an alarm prompt function. When an abnormal event occurs, the client can promptly display the alarm information and prompt the user to handle it.

[0021] A working method of an intelligent inspection system for a particle accelerator facility comprises the following steps:

[0022] Front-end intelligent devices are distributed across various particle accelerator facilities, collecting data from particle accelerator facilities in real time and transmitting the data to edge computing devices;

[0023] Edge computing devices pre-process and perform preliminary analysis on the real-time data they receive, including audio data processing and analysis, infrared temperature measurement data processing and analysis, to promptly detect abnormal events and report abnormal information to the server.

[0024] The server receives abnormal information reported by front-end intelligent devices and edge computing devices, and uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information. By establishing fault models and using data analysis and mining technologies, it determines the fault type, impact scope, and potential risks. It then makes decisions based on the analysis results and pushes alarm information.

[0025] The client serves as the interface for users to interact with the system. Users can view the real-time data, historical data and abnormal information of the particle accelerator facility through the client. When an abnormal event occurs, the client will promptly display the alarm information and prompt the user to handle it.

[0026] Compared with the prior art, the present invention has the following beneficial effects:

[0027] Efficient and accurate monitoring: This system uses multimodal sensors to collect various types of data such as temperature, audio, images, water leakage, etc. from particle accelerator facilities in real time, and uses artificial intelligence algorithms to conduct in-depth analysis of these data. It can accurately identify equipment failure modes and achieve comprehensive, efficient and accurate monitoring of the operating status of accelerator facilities, effectively overcoming the defects of traditional manual inspections that are difficult to monitor in real time and comprehensively.

[0028] Fault Warning and Diagnosis: By building fault models and applying data analysis and mining techniques, the system can proactively detect potential equipment failures and issue timely warnings. It can also accurately diagnose faults, identify their type and location, and provide maintenance personnel with detailed fault information, facilitating rapid troubleshooting, reducing downtime, and improving particle accelerator operational efficiency.

[0029] Reduced labor costs: The intelligent inspection system is highly automated and can replace most manual inspection tasks, reducing reliance on large numbers of inspectors and lowering labor costs. Furthermore, the system operates 24 / 7, eliminating the need for manual shifts, further improving the efficiency and reliability of inspections.

[0030] Standardized Data Management: The system automatically records inspection data and stores and manages it in a unified format, eliminating the risk of data loss and errors associated with manual recording. Standardized data management provides reliable data support for subsequent equipment operating status analysis and fault trend prediction, helping to improve the scientific nature of equipment management and the accuracy of decision-making.

[0031] Improved Operation and Maintenance Decision-Making: Based on the data collected and analyzed by the system, operations and maintenance personnel can gain a deeper understanding of the equipment's operating patterns and failure characteristics, enabling them to formulate more scientific and rational operations and maintenance plans and decisions. For example, based on equipment failure predictions, repair and maintenance work can be scheduled in advance, preventing the impact of sudden equipment failures on scientific research and improving the overall operation and maintenance of the particle accelerator. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 It is a basic principle diagram of the system of the present invention.

[0033] Figure 2 It is a schematic diagram of the system structure of the present invention.

[0034] Figure 3 This is a schematic diagram of the layout of front-end intelligent equipment at an accelerator facility site in the present invention.

[0035] Figure 4 This is a flow chart of audio processing and analysis performed by the edge computing device in the present invention.

[0036] Figure 5 It is the audio feature data curve extracted by the edge computing device in the present invention.

[0037] Figure 6 It is a real-time status display interface of a certain accelerator facility site in the present invention. DETAILED DESCRIPTION

[0038] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings:

[0039] like Figure 1-6 As shown, a particle accelerator facility intelligent inspection system mainly includes front-end intelligent devices, edge computing devices, a server, and a client. The front-end intelligent devices are distributed at various facility sites within the particle accelerator to enable real-time dynamic monitoring of the particle accelerator facility. The edge computing devices are responsible for preprocessing and performing preliminary analysis on the real-time data transmitted by the front-end intelligent devices. The server receives abnormal information reported by the front-end intelligent devices and edge computing devices, and uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information. By establishing fault models and using data analysis and mining techniques, it determines the fault type, impact scope, and potential risks, makes decisions based on the analysis results, and pushes alarm information. The client is the interface for users to interact with the system, allowing users to view the real-time data, historical data, and abnormal information of the particle accelerator facility through the client.

[0040] The front-end intelligent devices mentioned above include infrared temperature cameras, smart dome cameras, and water leakage transmitters. These devices are distributed across the particle accelerator's facilities and are responsible for collecting real-time data from the particle accelerator, including temperature, audio, images, and water leakage. These devices use Internet of Things technology to transmit the collected data to edge computing devices via the network and report any abnormal information to the server.

[0041] To optimize system operation, the layout of front-end intelligent devices should follow the following strategies:

[0042] (1) Infrared temperature measurement cameras: Deploy infrared temperature measurement cameras in key equipment areas and high-temperature risk points to achieve non-contact equipment temperature monitoring and ensure equipment safety. These cameras should be placed in locations with a wide field of view and not easily obstructed to more accurately capture temperature data.

[0043] (2) Smart dome cameras: Smart dome cameras are installed in important areas, and their 360-degree rotation and zoom functions are used to achieve all-round, no-blind-angle video surveillance. The dome cameras should be placed in a location that can cover the maximum monitoring range and is easy to adjust the angle to ensure clear and comprehensive monitoring images.

[0044] (3) Water leakage transmitter: Water leakage transmitters should be installed in areas where there may be a risk of water leakage, such as cooling systems and near water pipes. Corresponding water leakage monitoring tapes should be laid under or around water seepage points to detect and report water leaks in a timely manner to prevent water damage to equipment.

[0045] The edge computing devices mentioned above include industrial computers running Windows or Linux operating systems, which are responsible for collecting real-time data transmitted by front-end intelligent devices and pre-processing and analyzing this data. Combining historical data and advanced algorithm models, edge computing devices can promptly detect abnormal events and report abnormal information to the server. The following is an introduction to the main functions of edge computing devices:

[0046] (1) Audio data processing and analysis: The smart dome camera is connected to the network to collect audio data from the smart dome camera in real time. The collected audio is pre-processed by noise suppression, format conversion and compression to improve data quality. Then, algorithms such as support vector machines and deep learning are used to extract audio features and match them with the pre-trained fault model. Once an abnormality is detected, the alarm information is immediately reported to the server through EPICSPV.

[0047] (2) Infrared temperature measurement data processing and analysis: The infrared temperature measurement camera is connected to the network and infrared temperature measurement data is collected in real time. Once an abnormality is detected, the alarm information is immediately reported to the server through EPICSPV.

[0048] The server includes various service programs running on upper-level servers. It is responsible for receiving abnormal information reported by front-end smart devices and edge computing devices, applying advanced artificial intelligence algorithms to conduct in-depth analysis of this information, making decisions based on the analysis results, and sending alarm messages. In addition, the server also provides data storage, query, and statistical analysis capabilities to provide decision support for management. The server adopts a microservices architecture to ensure high availability, scalability, and flexibility. The following is an introduction to the server's main functions:

[0049] (1) Data reception and processing: Receive abnormal information sent by front-end intelligent devices and edge computing devices through the network, and verify and clean the received data. Then, use artificial intelligence algorithms to conduct in-depth analysis of the abnormal information. By establishing fault models, data analysis and mining technologies, determine the fault type, impact range, and potential risks, and make decisions based on the analysis results.

[0050] (2) Real-time alarm push: The open source software Phoebus alarm service program released by the EPICS community is used to build a real-time alarm push service, which realizes WeChat classification push, front-end web page display, alarm information historical data query and other functions.

[0051] (3) Historical Data Archiving: Use the high-performance and easy-to-use Archiver Appliance to build a historical data archiving system. The advantages of the Archiver Appliance are that it can store millions of PVs, has fast data query speeds, and can be configured on a PV-by-PV basis.

[0052] The client, primarily based on the next-generation control software platform Phoebus, serves as the user interface for system interaction. Users can access real-time and historical data from the particle accelerator, as well as abnormality information. Furthermore, the client includes an alarm function. When an abnormality occurs, the client promptly displays an alarm message and prompts the user to address it.

[0053] Example: An intelligent inspection system for particle accelerator facilities based on multimodal sensors and artificial intelligence technology

[0054] like Figure 1 and Figure 2 As shown, a particle accelerator facility intelligent inspection system based on multimodal sensors and artificial intelligence technology mainly includes front-end intelligent devices, edge computing devices, a server, and a client. The front-end intelligent devices are distributed at various facility sites within the particle accelerator to achieve real-time dynamic monitoring of the particle accelerator facility. The edge computing devices are responsible for preprocessing and preliminary analysis of the real-time data transmitted by the front-end intelligent devices. The server receives abnormal information reported by the front-end intelligent devices and edge computing devices, and uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information. By establishing fault models and using data analysis and mining techniques, it determines the fault type, impact scope, and potential risks, makes decisions based on the analysis results, and pushes alarm information. The client is the interface for users to interact with the system, allowing users to view real-time data, historical data, and abnormal information of the particle accelerator facility through the client. This invention effectively overcomes the limitations of traditional manual inspection methods for particle accelerator facilities, achieving real-time monitoring and abnormality alarms for particle accelerator facilities, significantly improving work efficiency and equipment operation safety, and providing strong support for the stable operation of particle accelerators.

[0055] The front-end intelligent devices include infrared temperature measurement cameras, smart dome cameras and water leakage transmitters, etc. These devices are distributed at various facilities of the particle accelerator and are responsible for collecting real-time data of the particle accelerator facilities, including temperature, audio, images, water leakage, etc. The front-end intelligent device layout diagram of a certain accelerator facility site in the present invention is shown in the figure below. Figure 3 shown.

[0056] The infrared temperature camera's thermal imaging sensor is a vanadium oxide uncooled detector with a thermal imaging pixel size of 256×192. The temperature measurement range is -20°C to 150°C and it supports POE power supply.

[0057] The smart dome camera is equipped with a 360° pan / tilt, 23x optical zoom, 100m high-definition infrared night vision, a maximum image size of 2560×1440, a built-in microphone, and supports POE power supply.

[0058] The water leakage transmitter supports positioning function with a positioning resolution of 0.1m, a maximum detection distance of 150m, and a detection accuracy of 0.5m±1%FS. It also has functions such as disconnection alarm and interference alarm.

[0059] Edge computing devices, such as industrial computers running Windows or Linux operating systems, collect real-time data transmitted by front-end intelligent devices and perform pre-processing and analysis on it. By combining historical data with advanced algorithmic models, edge computing devices can promptly detect abnormal events and report them to the server.

[0060] Edge computing devices process and analyze audio data as follows: Figure 4 As shown, the workflow is as follows:

[0061] (1) Audio data acquisition: First, connect the edge computing device to the smart dome camera through the network to ensure smooth data communication. Then, use the HCNetSDK software toolkit or other interface protocols to obtain audio data from the camera in real time. This audio data includes normal audio, abnormal audio, or fault audio during device operation.

[0062] (2) Audio data preprocessing: The collected audio data is subjected to noise suppression to reduce the impact of ambient noise on fault identification. In addition, the audio data can be converted into a format suitable for subsequent processing as needed, and necessary compression can be performed to save storage space and processing time.

[0063] (3) Intelligent algorithm analysis: Use intelligent algorithms (such as support vector machines, deep learning, etc.) to extract features from pre-processed audio data. Then, match the features of the real-time audio data with the pre-trained fault recognition model. Based on the model matching results, determine whether the current audio data indicates a device fault and determine the type and extent of the fault.

[0064] (4) Alarm: Once it is determined that the audio data indicates a device failure, the edge computing device will immediately trigger the alarm mechanism and report the alarm information to the server through EPICSPV.

[0065] (5) Continuous optimization: Evaluate the performance of edge computing devices during the fault identification process, including indicators such as processing speed and recognition accuracy; and regularly update and optimize the fault identification model to improve its recognition accuracy and generalization ability.

[0066] The process of edge computing devices processing and analyzing infrared temperature measurement data is as follows:

[0067] (1) Device connection: The edge computing device establishes a network connection with the infrared temperature measurement camera through the HCNetSDK software toolkit, ensuring real-time transmission of infrared temperature measurement data once per second.

[0068] (2) Protocol conversion: Use the PyEpics library to convert the raw infrared temperature measurement data into EPICSPV for publication.

[0069] (3) Alarm deduplication: Use a sliding window algorithm to filter out duplicate alarms within a short period of time.

[0070] (4) False alarm analysis and optimization: Based on the alarm history database, the false alarm correlation factors are mined and the alarm threshold parameters are optimized.

[0071] The server includes various service programs running in the upper-level server, which is responsible for receiving abnormal information reported by front-end intelligent devices and edge computing devices, and using advanced artificial intelligence algorithms to conduct in-depth analysis of this information. By establishing fault models, data analysis and mining and other technologies, it determines the fault type, impact range and potential risks, and makes decisions based on the analysis results and pushes alarm information.

[0072] (1) Data reception and processing: Receive abnormal information sent by front-end intelligent devices and edge computing devices through the network, and verify and clean the received data to ensure the validity and accuracy of the data. Then, call the artificial intelligence algorithm to conduct in-depth analysis of the abnormal information, including but not limited to fault model training, abnormality type identification, impact range assessment, potential risk prediction, etc., and publish the alarm information as EPICSPV.

[0073] (2) Real-time alarm push: The open source software Phoebus alarm service program released by the EPICS community is used to build a real-time alarm push service, which realizes WeChat classification push, front-end web page display, alarm information historical data query and other functions.

[0074] (3) Historical data archiving: Use ArchiverAppliance with high performance and ease of use to build a historical data archiving system. The advantages of ArchiverAppliance are that it can store millions of PVs, has fast data query speed, and can be configured PV by PV. When the performance of a single point deployment cannot meet the needs, it can be run in a cluster. The audio feature data curve extracted by the edge computing device stored in ArchiverAppliance is as follows: Figure 5 shown.

[0075] The client mainly uses the new generation control software platform Phoebus, which is the interface for users to interact with the system. Users can view the real-time data, historical data and abnormal information of the particle accelerator facility through the client. The real-time status display interface of an accelerator facility site developed based on Phoebus is as follows: Figure 6 As shown, a green indicator light indicates that no fault or abnormality has been identified, and a red indicator light indicates that a fault or abnormality has been identified. In addition, the client also has an alarm prompt function. When an abnormal event occurs, the client can promptly display the alarm information and prompt the user to handle it.

Claims

1. An intelligent inspection system for particle accelerator facilities, characterized by: Including front-end intelligent devices, edge computing devices, servers and clients; The front-end intelligent devices are distributed at various facilities of the particle accelerator and are used to collect data from the particle accelerator facilities in real time, including temperature, audio, images, and water leakage, to achieve real-time dynamic monitoring of the particle accelerator facilities; The edge computing device is responsible for pre-processing and preliminary analysis of real-time data transmitted from the front-end intelligent device, including audio data processing and analysis, infrared temperature measurement data processing and analysis, and can detect abnormal events in a timely manner and report abnormal information to the server; The server receives abnormal information reported by front-end intelligent devices and edge computing devices, uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information, and determines the type of fault, scope of impact, and potential risks by establishing fault models and using data analysis and mining technologies. It then makes decisions based on the analysis results and pushes alarm information. The client is an interface for users to interact with the system. Users can view real-time data, historical data and abnormal information of the particle accelerator facility through the client.

2. The particle accelerator facility intelligent inspection system according to claim 1, characterized in that: The front-end intelligent equipment includes an infrared temperature measuring camera, an intelligent ball camera and a water leakage transmitter; The infrared temperature camera's thermal imaging sensor is a vanadium oxide uncooled detector with a thermal imaging pixel size of 256×192, a temperature measurement range of -20°C to 150°C, and supports POE power supply. The smart dome camera is equipped with a 360° pan / tilt, 23x optical zoom, 100m high-definition infrared night vision, a maximum image size of 2560×1440, a built-in microphone, and supports POE power supply; The water leakage transmitter supports positioning function with a positioning resolution of 0.1m, a maximum detection distance of 150m, a detection accuracy of 0.5m±1%FS, and also has functions such as disconnection alarm and interference alarm.

3. The particle accelerator facility intelligent inspection system according to claim 1, characterized in that: The edge computing device includes an industrial computer running the Windows or Linux operating system, which is responsible for collecting real-time data transmitted by front-end intelligent devices and pre-processing and analyzing this data; combining historical data and advanced algorithm models, it can detect abnormal events in a timely manner and report the abnormal information to the server through EPICSPV.

4. The particle accelerator facility intelligent inspection system according to claim 3, characterized in that: The edge computing device is connected to the smart dome camera through a network, collecting the audio data of the smart dome camera in real time, performing pre-processing such as noise suppression, format conversion and compression on the collected audio to improve data quality, and then using support vector machines, deep learning and other algorithms to extract audio features and match them with pre-trained fault models. Once an abnormality is detected, the alarm information will be immediately reported to the server through EPICSPV.

5. The particle accelerator facility intelligent inspection system according to claim 3, characterized in that: The edge computing device connects to the infrared temperature measurement camera through the network and collects infrared temperature measurement data in real time. It uses the PyEpics library to convert the original infrared temperature measurement data into EPICSPV for publication. It adopts a sliding window algorithm to filter repeated infrared temperature measurement alarms in a short period of time. Based on the alarm history database, it mines the false alarm correlation factors and optimizes the alarm threshold parameters. Once an abnormality is detected, the alarm information is immediately reported to the server through EPICSPV.

6. The particle accelerator facility intelligent inspection system according to claim 1, characterized in that: The server includes various service programs running in the upper-level server, which is responsible for receiving abnormal information reported by front-end intelligent devices and edge computing devices, and using advanced artificial intelligence algorithms to train and model this information, conduct in-depth analysis, make decisions based on the analysis results, and push alarm information; the server also has data storage, query and statistical analysis functions to provide decision support for management.

7. The particle accelerator facility intelligent inspection system according to claim 6, characterized in that: The server adopts a microservice architecture to receive abnormal information sent by front-end intelligent devices and edge computing devices through the network, and verify and clean the received data; Then, artificial intelligence algorithms are called to conduct in-depth analysis of the abnormal information. By establishing fault models, data analysis and mining technologies, the fault type, impact scope and potential risks are determined, and decisions are made based on the analysis results, and the alarm information is published as EPICSPV.

8. The particle accelerator facility intelligent inspection system according to claim 6, characterized in that: The server uses the open source software Phoebus alarm service program released by the EPICS community to build a real-time alarm push service, realizing functions such as WeChat classification push, front-end web page display, and alarm information historical data query; and uses Archiver Appliance to build a historical data archiving system.

9. The particle accelerator facility intelligent inspection system according to claim 1, characterized in that: The client mainly uses the new generation control software platform Phoebus. Users can view the real-time data, historical data and abnormal information of the particle accelerator facility through the client; the client also has an alarm prompt function. When an abnormal event occurs, the client can promptly display the alarm information and prompt the user to handle it.

10. A working method of a particle accelerator facility intelligent inspection system according to any one of claims 1 to 9, characterized in that: The following steps are involved: Front-end intelligent devices are distributed across various particle accelerator facilities, collecting data from particle accelerator facilities in real time and transmitting the data to edge computing devices; Edge computing devices pre-process and perform preliminary analysis on the real-time data they receive, including audio data processing and analysis, infrared temperature measurement data processing and analysis, to promptly detect abnormal events and report abnormal information to the server. The server receives abnormal information reported by front-end intelligent devices and edge computing devices, and uses advanced artificial intelligence algorithms to conduct in-depth analysis of this information. By establishing fault models and using data analysis and mining technologies, it determines the fault type, impact scope, and potential risks. It then makes decisions based on the analysis results and pushes alarm information. The client serves as the interface for users to interact with the system. Users can view the real-time data, historical data and abnormal information of the particle accelerator facility through the client. When an abnormal event occurs, the client will promptly display the alarm information and prompt the user to handle it.

Images