Inspection centralized control platform system and control method thereof

Abstract

The invention provides an inspection centralized control platform system and a control method thereof. The inspection centralized control platform system comprises a server, an inspection robot and a camera, the server supports a plurality of standard industrial protocols, is in communication connection with the inspection robot, the camera and the DCS platform, and obtains process system data of the DCS platform; the cameras comprise a macro meter camera, a network camera and a fixed-point camera; the inspection robot performs inspection according to the inspection route set by the server, shoots an entity meter of the measurement equipment through a macro meter camera to obtain a meter count value, shoots a robot inspection image through a network camera, and sends the meter count value and the robot inspection image to the server; the fixed-point camera shoots a fixed-point image and sends the image to the server; and the server performs fault detection based on at least two of the process system data, the meter count value, the robot inspection image and the fixed point image. According to the invention, double-platform linkage monitoring, inspection and the like are realized, and the inspection efficiency and accuracy are improved.

Description

Technical Field

[0001] This disclosure relates to the fields of nuclear power monitoring and Internet of Things technology, and in particular to an inspection and control platform system and its control method. Background Technology

[0002] With the development of industrial automation and intelligent technologies, various IoT (Internet of Things) devices are increasingly being used in industrial inspection, security monitoring, and other fields. However, traditional inspection methods suffer from problems such as low efficiency, incomplete data collection, and slow response speed. At the same time, IoT devices from different manufacturers often use different communication protocols, making it difficult to achieve unified management and data integration. Summary of the Invention

[0003] The technical problem to be solved by this disclosure is the aforementioned deficiencies in the prior art, and a patrol inspection centralized control platform system and its control method are provided.

[0004] This disclosure solves the above-mentioned technical problems through the following technical solution:

[0005] This disclosure provides an inspection and control platform system, which includes: a server, an inspection robot, and cameras deployed in a central control room;

[0006] The server supports multiple standard industrial protocols for communication with the inspection robot and the camera, respectively.

[0007] The server is also connected to the DCS (Distributed Control System) platform to acquire process system data from the DCS platform; wherein the process system data includes measurement data sent by several measuring devices.

[0008] The cameras include a macro meter camera and a network camera mounted on the inspection robot, as well as a fixed-point camera set at a fixed location; the network camera is a wide-angle camera and supports an open application platform;

[0009] The inspection robot is used to perform inspections according to the inspection route set by the server. It obtains meter images by taking pictures of the physical meters of the measuring equipment on the inspection route using the macro meter camera to obtain meter count values. It also takes robot inspection images of a first designated area around the inspection robot using the network camera and sends the meter count values ​​and robot inspection images to the server.

[0010] The fixed-point camera is used to capture a fixed-point image of a matching preset area and send the fixed-point image to the server;

[0011] The server is also used to perform fault detection based on at least two of the process system data, the table count values, the robot inspection images, and the fixed-point images.

[0012] Optionally, the fixed-point camera includes a high-altitude spherical eagle-eye camera;

[0013] The server is also communicatively connected to the high-altitude spherical eagle-eye camera.

[0014] The high-altitude spherical eagle-eye camera is installed on the roof and is used to monitor objects and people in the distance;

[0015] And / or,

[0016] The fixed-point camera includes a helmet detection camera;

[0017] The server is also communicatively connected to the helmet detection camera.

[0018] The helmet detection camera is used to detect whether the first target person is wearing a helmet;

[0019] And / or,

[0020] The fixed-point camera includes a face recognition camera;

[0021] The server is also communicatively connected to the face recognition camera;

[0022] The face recognition camera is used to identify the face of a second target person in order to verify their identity;

[0023] And / or,

[0024] The fixed-point camera includes a personnel detection camera;

[0025] The server is also communicatively connected to the personnel detection camera;

[0026] The personnel detection camera is used to monitor the target location and to collect statistics on at least one of passenger flow, number of people, and personnel density.

[0027] And / or,

[0028] The fixed-point camera includes a behavior recognition camera;

[0029] The server is also communicatively connected to the behavior recognition camera;

[0030] The behavior recognition camera is used to recognize the body movements of third-party target personnel, and supports behavior analysis of falling, violent exercise, abnormal spacing, lingering, abnormal number of people, crossing the warning line, wandering, running, and leaving the post;

[0031] And / or,

[0032] The inspection and control platform system also includes a lidar mounted on the inspection robot;

[0033] The server is also communicatively connected to the lidar;

[0034] The lidar is used to acquire laser data of a second designated area around the inspection robot.

[0035] Optionally, the inspection and control platform system also includes a meter inspection host;

[0036] The server is also communicatively connected to the meter inspection host.

[0037] The meter inspection host is used to store the meter count values ​​obtained from the meter images;

[0038] And / or,

[0039] The inspection and control platform system also includes a hard disk video recorder;

[0040] The server is also communicatively connected to the hard disk recorder;

[0041] The hard disk recorder is used to store images captured by the camera;

[0042] And / or,

[0043] The inspection and control platform system also includes an inspection knowledge base host;

[0044] The server is also connected to the inspection knowledge base host.

[0045] The inspection knowledge base host is used to store the processing results of each inspection.

[0046] And / or,

[0047] The inspection and control platform system also includes interactive devices;

[0048] The server is also communicatively connected to the interactive device;

[0049] The interactive device is used to send interactive data to the server and display the graphic and text data sent by the server.

[0050] Optionally, the server uses Spark Streaming (Spark real-time stream processing framework) as the real-time streaming data processing engine and constructs a logical relationship graph between measurement devices within the DCS platform;

[0051] The inspection and control platform system adopts a unified data format and standard, and the data within the system uses a unified timestamp, adding semantic tags to the data within the system.

[0052] This disclosure also provides a control method for an inspection centralized control platform system, the control method being applied to the aforementioned inspection centralized control platform system, the control method comprising:

[0053] The server determines the first target equipment to be inspected based on the DCS fault detection data in the process system data of the DCS platform, and sets the inspection route of the inspection robot based on the first target equipment.

[0054] The inspection robot performs inspections according to the inspection route, obtains meter images by taking pictures of the physical meters of the first target device with a macro meter camera to obtain the first meter count value, and takes first robot inspection images of a first designated area around the inspection robot with a network camera, and sends the first meter count value and the first robot inspection images to the server.

[0055] The fixed-point cameras set along the inspection route send the captured first images to the server.

[0056] The server obtains inspection results based on the DCS fault detection data, the first table count value, the first robot inspection image, and the first image; wherein, the inspection results include whether a fault has occurred in the inspection area and the cause of the fault when it occurs.

[0057] Optionally, the control method further includes:

[0058] The server sets the inspection route for the inspection robot;

[0059] The inspection robot performs inspections according to the inspection route set by the server. It obtains second meter images by taking pictures of the physical meters of all measuring devices along the inspection route using the macro meter camera, and obtains second meter count values. It also takes second robot inspection images of a first designated area around the inspection robot using the network camera, and sends the second meter count values ​​and the second robot inspection images to the server.

[0060] The fixed-point cameras set along the inspection route send the captured second images to the server.

[0061] The server compares the second table count value, the second robot inspection image, and the second image with the process system data of the DCS platform to obtain the inspection result; wherein, the inspection result also includes whether the measuring equipment in the DCS platform has malfunctioned.

[0062] Optionally, the DCS fault detection data is obtained by the DCS controller layer in the DCS platform through lightweight diagnostics based on the process system data;

[0063] The server obtains inspection results based on the DCS fault detection data, the first table count value, the first robot inspection image, and the first image, including:

[0064] The server takes the DCS fault detection data, the first table count value, the first robot inspection image, and the first image as input data, and outputs the inspection results through the equipment health assessment model; wherein, the equipment health assessment model is constructed based on a deep learning predictive analysis model.

[0065] Optionally, the server compares the second table count value, the second robot inspection image, and the second image with the process system data of the DCS platform to obtain the inspection result, including:

[0066] The server takes the second table count value, the second robot inspection image, the second image, and the process system data as input data, and outputs the inspection results through the equipment health assessment model.

[0067] Optionally, the server determines the first target equipment to be inspected based on the DCS fault detection data in the process system data of the DCS platform, and sets the inspection route of the inspection robot according to the first target equipment, including:

[0068] Based on the DCS fault detection data and the fault knowledge data stored in the inspection knowledge base host, the server determines the first target device that makes up the inspection point and the inspection content for the first target device, and sets the inspection route according to the location of the first target device and the connection relationship of related devices.

[0069] Optionally, the server sets the inspection route for the inspection robot, including:

[0070] The server determines the inspection points and the detection content of the measuring devices at the inspection points based on the fault knowledge data stored in the inspection knowledge base host, and sets the inspection route according to the location of the measuring devices at the inspection points and the connection relationship of related devices.

[0071] Based on common knowledge in the field, the above-mentioned preferred conditions can be combined arbitrarily to obtain various preferred embodiments of this disclosure.

[0072] The positive advancements of this disclosure are as follows: By communicating with various IoT (Internet of Things) devices such as inspection robots and macro-meter cameras, which support multiple standard industrial protocols, unified management and data integration are achieved within the system. Process system data is dynamically loaded while the DCS platform is running. A deep learning-based equipment health assessment model analyzes the process system data and the inspection data from the inspection control platform system, monitoring faults in the industrial field. This enables dual-platform linkage monitoring, control, alarm, and inspection functions between the DCS platform and the inspection control platform system. DCS fault detection can trigger automatic inspections, enabling rapid response and fault handling. Furthermore, by linking inspection route planning with fault knowledge data, inspection routes are optimized, improving inspection efficiency and accuracy, reducing manual intervention, and lowering maintenance costs. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of a patrol control platform system provided in Embodiment 1 of this disclosure;

[0074] Figure 2 This is a schematic diagram of another inspection and control platform system provided in Embodiment 1 of this disclosure;

[0075] Figure 3 This is a schematic diagram of the internal data flow of another inspection and control platform system provided in Embodiment 1 of this disclosure;

[0076] Figure 4 A flowchart of a control method for an inspection centralized control platform system provided in Embodiment 2 of this disclosure;

[0077] Figure 5 This is a flowchart of a specific implementation of a control method for an inspection centralized control platform system provided in Embodiment 2 of this disclosure. Detailed Implementation

[0078] The present disclosure is further illustrated below by way of embodiments, but the present disclosure is not limited to the scope of the embodiments described herein.

[0079] The prefixes such as "first" and "second" used in this disclosure are merely for distinguishing different descriptive objects and do not limit the position, order, priority, quantity, or content of the described objects. The use of ordinal numbers and other prefixes used to distinguish descriptive objects in this disclosure does not constitute a limitation on the described objects. The description of the described objects is given in the claims or the context of the embodiments, and should not be construed as an unnecessary limitation. Furthermore, in the description of this embodiment, unless otherwise stated, "multiple" means two or more.

[0080] Example 1

[0081] Figure 1 This is a schematic diagram of a patrol control platform system provided as an exemplary embodiment of the present disclosure. The patrol control platform system includes: a server 1, a patrol robot 2, and a camera 3 deployed in the central control room.

[0082] Server 1 supports multiple standard industrial protocols for communication with inspection robot 2 and camera 3 respectively.

[0083] Server 1 also communicates with the DCS platform to acquire process system data from the DCS platform. This process system data includes measurement data sent by several measuring devices.

[0084] Camera 3 includes a macro meter camera 31 and a network camera 32 mounted on the inspection robot 2, as well as a fixed-point camera 33 set at a fixed location. The network camera 32 is a wide-angle camera and supports open application platforms.

[0085] The inspection robot 2 is used to perform inspections according to the inspection route set by the server 1. It uses a macro meter camera 31 to capture images of the physical meters of the measuring devices along the inspection route to obtain meter count values. It also uses a network camera 32 to capture robot inspection images of a first designated area around the inspection robot 2 and sends the meter count values ​​and robot inspection images to the server 1.

[0086] The fixed-point camera 33 is used to capture a fixed-point image of a matching preset area and send the fixed-point image to the server 1.

[0087] Server 1 is also used for fault detection based on at least two of the following: process system data, meter counts, robot inspection images, and fixed-point images.

[0088] Among them, the DCS (Distributed Control System) platform is a nuclear power plant control system deployed in the industrial site. It is used to control the production (e.g., total power generation, hourly power generation, etc.) and process management (e.g., adjusting steam temperature, adjusting valve opening, etc.) in the industrial site. It receives measurement data (e.g., steam temperature, pipeline pressure, flow rate, valve opening, pump status, etc.) from various measuring devices (e.g., temperature sensors, pressure sensors, etc.). It is existing technology and will not be described in detail here.

[0089] Both the inspection robot 2 and the camera 3 are inspection devices. Non-contact sensors can be installed on the inspection robot 2 to collect environmental data, such as temperature, humidity, air quality, and noise. The inspection robot 2 can then send this environmental data to the server, providing data support for fault monitoring. The data acquired by the inspection devices is collectively referred to as inspection data (e.g., images captured by the camera, environmental data acquired by the non-contact sensors), and this inspection data is one of the data foundations for the server's fault monitoring.

[0090] The inspection robot 2 can be a wheeled robot, such as a robot with a four-wheel and eight-wheel drive chassis, which supports wireless communications such as Wi-Fi (Wireless Fidelity) and 5G (the fifth-generation mobile communication technology), and is equipped with a lifting arm. Both the macro-meter camera 31 and the network camera 32 can be set on the lifting arm to obtain better shooting positions and perspectives through the lifting arm. The server 1 can directly control the inspection robot and the lifting arm through instructions; the server 1 can also send instructions to the built-in controller of the inspection robot, and the built-in controller controls the inspection robot and the lifting arm. Controlling the robot to move, rotate, and controlling the lifting arm to perform three-dimensional movement and end rotation are all existing technologies and will not be elaborated here.

[0091] There are many problems in the existing inspection system, such as data isolation and no interaction with the nuclear power plant control system, the inspection tasks are carried out according to the pre-set plan, lack of intelligence, and the application integration of heterogeneous sensing devices. To solve these problems, the inspection centralized control platform system integrates a server, an inspection robot, and multiple cameras. The server supports multiple standard industrial protocols such as TCP / IP (Transmission Control Protocol / Internet Protocol), RTSP (Real-Time Streaming Protocol), and GB28181 (National Standard 28181) according to the device type; the server can support the access of IoT (Internet of Things) devices such as inspection robots, macro-meter cameras, network cameras, and fixed-point cameras through a gateway or switch adapter; the server obtains process system data from the DCS platform and supports the parsing and structured processing of historical trend data, alarm records, control instructions, etc. The DCS fault detection triggers automatic inspection to achieve functions such as dual-platform linkage monitoring, control, alarm, and inspection; in regular inspections, the server can also perform intelligent fault monitoring in combination with process system data; establish a nuclear power plant intelligent inspection centralized control platform with the linkage function of intelligent inspection equipment and the nuclear power plant control system to solve the data island problem and achieve the functions of less-attended and intelligent management and control of inspection work.

[0092] The macro-meter camera 31 captures the meter image of the physical meter of the measurement device. The meter image needs to be clear enough to identify the meter value. Preferably, the macro-meter camera 31 has more than 5 million pixels.

[0093] The network camera 32 can be a HEOP (Open Application Platform) network camera, which supports the development platform HEOP and the AI (Artificial Intelligence) development platform and can import third-party intelligent applications.

[0094] The fixed-point camera 33 can adopt an aerial spherical eagle-eye camera, a safety helmet detection camera, a face recognition camera, a personnel detection camera, a behavior recognition camera, etc. to monitor and identify different contents.

[0095] Various cameras can perform visual monitoring of equipment and personnel to identify potential hazards, ongoing malfunctions (such as pipeline leaks), non-compliant attire, abnormal gatherings and movements of people, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0096] The front-end of the inspection and control platform system can be developed using Vue (a progressive JavaScript front-end framework), including view, router, and component components. The control layer, business service layer, and data access layer implement the control logic for each basic business function. The data access layer is used for reading and writing to the database, which can be MySQL (a database software). In the runtime environment, the inspection and control platform system software is deployed on a server. The inspection and control platform system can dynamically bind physical equipment inspection paths to the DCS logic control process configuration diagram, enabling interactive technology such as "clicking the DCS screen → retrieving the corresponding equipment inspection data."

[0097] Specifically, the inspection and control platform system dynamically loads process system data while the DCS platform is running. The inspection data and process system data acquired by the inspection and control platform system itself are displayed simultaneously. A lightweight data engine can be used to achieve seamless integration of real-time process system data from the DCS platform with inspection data from the inspection and control platform system. Figure 3 The diagram illustrates the internal data flow of the inspection and control platform system that realizes DCS process implementation, specifically in the following aspects:

[0098] (1) Multi-source data acquisition

[0099] It comprehensively collects process system data (data reflecting the process flow, i.e., process data) from the DCS system, and interfaces with the DCS system through standard protocols such as OPC DA / UA (OPC Data Access / OPC Unified Architecture), Modbus (a protocol developed by Modicon), and Profibus (process fieldbus) to collect key process parameters (such as temperature, pressure, flow rate, valve opening, pump status, etc.) in real time. It supports the parsing and structured processing of historical trend data, alarm records, and control commands.

[0100] Meanwhile, intelligent inspection equipment, such as cameras and various non-contact sensors, collects on-site environmental data, including temperature, humidity, air quality, and noise. This data is transmitted in real time to the intelligent inspection big data storage and processing platform via wired or wireless communication, supporting the uploading of multimodal data such as images, audio, video, and sensor data.

[0101] (2) Data integration and storage

[0102] The intelligent inspection big data storage and processing platform employs a lightweight streaming data processing engine (such as Spark Streaming) to integrate the collected process system data and inspection data from the DCS platform. Data standardization is achieved by unifying the format of heterogeneous data from the DCS platform and inspection equipment. Timestamp alignment is implemented using an event-driven mechanism to synchronize data timelines. Data tagging is used to add multi-dimensional tags such as source, equipment type, spatial location, and inspection task ID. Dynamic data fusion constructs a logical association graph of "process variables - equipment status," supporting dynamic querying and joint analysis. A unified database is established, utilizing distributed storage technology, specifically the Hadoop Distributed File System (HDFS), to ensure efficient data storage and reliable management.

[0103] (3) Process visualization and inspection linkage display module

[0104] The unified visualization interface allows for a dual-view linkage mechanism between the process view and the inspection view. The process view displays the DCS process flow and real-time variables in the form of a flowchart or P&I (Pipeline & Instrumentation) diagram. The inspection view displays inspection data in the form of equipment topology diagrams, image monitoring, heat maps, etc.

[0105] Linkage mechanism: Selecting a process node (e.g., via mouse or touchscreen operation) will automatically display the inspection status and historical trends of related equipment; when the inspection system detects an anomaly, it will automatically highlight the corresponding process path and the scope of impact.

[0106] It supports issuing inspection tasks or DCS control strategy suggestions directly through a graphical interface. Inspection tasks can include the target equipment to be inspected and the inspection content, as well as the inspection route determined based on the target equipment and the inspection content. The target equipment to be inspected can be electrical equipment (such as sensors, steam generators, electric valves, etc.) or non-electrical equipment (such as pipelines, storage tanks, manual valves, etc.).

[0107] (4) Method for implementing automatic inspection triggered by DCS fault detection

[0108] A deep learning-based data fusion algorithm is proposed, using process system data (such as pressure and flow) from the DCS platform as the main feature and inspection perception data (such as images or infrared thermal imaging) as auxiliary features to establish an equipment health assessment model. A hierarchical distributed computing architecture is designed to implement lightweight diagnostics (anomaly detection of key parameters in process system data) at the DCS controller layer, and to complete plant-wide equipment degradation prediction in the cloud, establishing a closed-loop logic between inspection results and DCS control correction.

[0109] A hierarchical distributed computing architecture is adopted: Edge layer (DCS controller in DCS platform): Deploy lightweight anomaly detection model, such as 1D-CNN (one-dimensional convolutional neural network) + LSTM (long short-term memory network), to monitor key parameters (pressure, flow, etc.) in real time, with a response time ≤100ms, and trigger the automatic threshold calibration function.

[0110] The inspection and control platform system backend employs a multimodal data fusion model to integrate process system data and inspection data from the DCS platform. It models equipment topology relationships using a graph neural network (GNN) and stores structured expert knowledge rules (emphasizing IF-THEN rules; supporting attribute-value pair-based condition combinations; and supporting rule priority, confidence level, and scope of application). The system backend receives status data, invokes knowledge rules for reasoning, and outputs diagnostic results and risk levels. Based on the diagnostic results, it dynamically generates inspection task plans, including inspection time, personnel, inspection locations, tools, and precautions. The system integrates templated inspection schemes through software and triggers automatic inspection execution, particularly supporting multi-dimensional retrieval and intelligent analysis of the operating status of key equipment in nuclear power plants (such as turbines, boilers, main pumps, and valves).

[0111] The inspection and control platform system, based on fused multi-source data, constructs an equipment health assessment model using predictive analytics. It employs machine learning algorithms (such as LSTM, random forest, and Bayesian networks) to predict the lifespan and failure probability of key equipment; combines process variable trends to deduce potential fault propagation paths; and links DCS control parameters (such as setpoint and threshold adjustments) to achieve preventative regulation. It automatically generates inspection strategies: dynamically adjusting inspection frequency, task priority, and inspection methods based on prediction results; supporting templated matching and adaptive optimization of inspection tasks; and allowing inspection task suggestions to be fed back to the DCS platform via OPC DA / UA or API (Application Programming Interface).

[0112] In this embodiment, a server supporting multiple standard industrial protocols communicates with various IoT devices such as inspection robots and macro-meter cameras to achieve unified management and data integration within the system. Process system data is dynamically loaded while the DCS platform is running. A deep learning-based equipment health assessment model analyzes the process system data and the inspection data from the inspection control platform system to monitor faults in the industrial field. This enables dual-platform linkage monitoring, control, alarm, and inspection functions between the DCS platform and the inspection control platform system. DCS fault detection can trigger automatic inspections, enabling rapid response and fault handling. Furthermore, the system correlates inspection route planning with fault knowledge data to optimize inspection routes, improve inspection efficiency and accuracy, reduce manual intervention, and lower maintenance costs.

[0113] In an optional embodiment, refer to Figure 2 The fixed-point camera 33 includes a high-altitude spherical eagle-eye camera 331.

[0114] Server 1 is also connected to the high-altitude spherical eagle-eye camera 331.

[0115] The high-altitude spherical eagle-eye camera 331 is installed on the roof to monitor distant objects and people.

[0116] Among them, the high-altitude spherical eagle-eye camera 331 can output large-angle (e.g., 270°) large-scene stitching images, and the panoramic images can support distortion correction of the area of ​​interest. It also features a built-in 45x optical image stabilization zoom lens.

[0117] In this embodiment, the high-altitude spherical eagle-eye camera focuses on monitoring distant locations, such as whether there are leaks in distant pipelines. It can perform visual monitoring of both equipment and personnel to identify potential dangers, ongoing malfunctions (such as pipeline leaks), non-compliant attire, abnormal gatherings and movements of people, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0118] In an optional embodiment, refer to Figure 2 The fixed-point camera 33 includes a safety helmet detection camera 332.

[0119] Server 1 is also in communication connection with helmet detection camera 332.

[0120] The helmet detection camera 332 is used to detect whether the first target person is wearing a helmet.

[0121] The helmet detection camera 332 can be configured as a helmet detection camera + speaker kit. This kit includes one camera and one speaker, supporting the detection of people not wearing helmets. Helmets can be detected in red, orange, yellow, blue, and white colors, and the input / output switches are linked to trigger the speaker to broadcast announcements. The helmet detection camera 332 also supports switching between the following modes: helmet detection mode, face capture mode, perimeter mode, and road monitoring mode. It can also support the independent simultaneous operation of helmet detection and face capture algorithms.

[0122] In this embodiment, the safety helmet detection camera focuses on detecting non-compliant safety helmet attire. It can perform visual monitoring of both equipment and personnel to identify potential hazards, ongoing malfunctions (such as pipeline leaks), non-compliant attire, abnormal personnel gatherings and movements, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0123] In an optional embodiment, refer to Figure 2The fixed-point camera 33 includes a face recognition camera 333.

[0124] Server 1 is also connected to face recognition camera 333.

[0125] The face recognition camera 333 is used to identify the face of a second target person in order to verify their identity.

[0126] Among them, the face recognition camera 333 can extract target features by itself to form a deep face image that can be learned, which greatly improves the detection rate of the target face.

[0127] In this embodiment, the face recognition camera focuses on identifying personnel and can perform visual monitoring of personnel to identify potential dangers, non-compliant personnel operations, abnormal personnel gatherings and movements, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0128] In an optional embodiment, refer to Figure 2 The fixed-point camera 33 includes the personnel detection camera 334.

[0129] Server 1 also communicates with personnel detection camera 334.

[0130] The 334 personnel detection camera is used to monitor target locations and to collect statistics on at least one of the following: passenger flow, number of people, and personnel density.

[0131] Among them, the personnel detection camera 334 can accurately count passenger flow, number of people, and personnel density.

[0132] People counting: Supports simultaneous operation of area attention (abnormal number of people, timeout alarm, number of people change reporting, etc.), job duty detection, heat map and smart lighting functions.

[0133] Inclined Passenger Flow: Based on pedestrian path analysis, it statistically analyzes the entry, exit, and passage of target personnel within a specified scenario; it supports forward and backward deduplication; and it supports flow rate analysis.

[0134] Crowd density and congestion detection: It can detect the crowding situation of people in a specified scene, configure the density level according to the number of people and duty cycle, and support up to 8 recognition areas; for large scenes, crowd density can support up to 1000 targets.

[0135] In this embodiment, the personnel detection camera can visually monitor personnel to identify potential dangers, abnormal personnel gatherings and movements, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0136] In an optional embodiment, refer to Figure 2The fixed-point camera 33 includes a behavior recognition camera 335.

[0137] Server 1 is also in communication connection with behavior recognition camera 335.

[0138] The behavior recognition camera 335 is used to recognize the body movements of third-party personnel, and supports behavior analysis of falling, vigorous movement, abnormal spacing, lingering, abnormal number of people, crossing the warning line, wandering, running, and leaving the post.

[0139] In this embodiment, the behavior recognition camera focuses on recognizing people's body movements to identify potential dangers and abnormal personnel movements, thereby reducing or avoiding safety accidents and improving production and personnel safety.

[0140] In an optional embodiment, refer to Figure 2 The inspection and control platform system also includes a lidar 4 mounted on the inspection robot 2.

[0141] Server 1 is also connected to LiDAR 4.

[0142] The lidar 4 is used to acquire laser data of a second designated area around the inspection robot 2.

[0143] In this embodiment, the laser data acquired by the lidar can be used to identify people and objects, thereby determining whether they are moving along the inspection route, and helping the inspection robot avoid people and objects when moving along the inspection route.

[0144] In an optional embodiment, refer to Figure 2 The inspection and control platform system also includes the meter inspection host 5.

[0145] Server 1 is also connected to the meter inspection host 5.

[0146] The meter inspection host 5 is used to store the meter count values ​​obtained from the meter images.

[0147] In this embodiment, the meter count values ​​obtained from the meter images are stored by the meter inspection host for easy access by the server later.

[0148] In an optional embodiment, refer to Figure 2 The inspection and control platform system also includes a hard disk video recorder 6.

[0149] Server 1 is also connected to hard disk recorder 6.

[0150] The hard disk recorder 6 is used to store images captured by the camera 3.

[0151] In this embodiment, images captured by each camera are stored using a hard disk recorder for easy retrieval by the server later.

[0152] In an optional embodiment, refer to Figure 2 The inspection and control platform system also includes the inspection knowledge base host 7.

[0153] Server 1 is also connected to the inspection knowledge base host 7.

[0154] The inspection knowledge base host 7 is used to store the processing results of each inspection.

[0155] In this embodiment, the processing results of each inspection are stored on the inspection knowledge base host to facilitate subsequent calls by the server.

[0156] In an optional embodiment, refer to Figure 2 The inspection and control platform system also includes interactive devices 8.

[0157] Server 1 also communicates with interactive device 8.

[0158] Interactive device 8 is used to send interactive data to server 1 and display the graphic and text data sent by server 1.

[0159] Interactive devices can include keyboards, mice, touchscreens, etc., and can be configured according to actual needs.

[0160] In this embodiment, staff can operate and manage the inspection and control platform system through interactive devices.

[0161] In an optional embodiment, server 1 uses Spark Streaming as a real-time streaming data processing engine and constructs a logical relationship graph between measurement devices within the DCS platform.

[0162] The inspection and control platform system adopts a unified data format and standard, and uses a unified timestamp for the data within the system, adding semantic tags to the data within the system.

[0163] In this embodiment, the continuous data stream (such as Kafka, Flume, Socket, etc.) is divided into micro-batch by the Spark real-time stream processing framework, and then handed over to the Spark engine for batch execution. This reuses Spark's in-memory computing, fault tolerance, scheduling and ecosystem advantages to achieve high throughput and fault-tolerant near real-time processing.

[0164] Example 2

[0165] This embodiment provides a control method for an inspection centralized control platform system, which is applied to the inspection centralized control platform system in Embodiment 1.

[0166] Figure 4A flowchart illustrating a control method for an inspection centralized control platform system provided as an exemplary embodiment of this disclosure, the control method comprising:

[0167] S11. The server determines the first target equipment to be inspected based on the DCS fault detection data in the process system data of the DCS platform, and sets the inspection route of the inspection robot based on the first target equipment.

[0168] S12. The inspection robot performs inspections according to the inspection route, captures the physical meter of the first target device with a macro meter camera to obtain the meter image and obtain the first meter count value, captures the first robot inspection image of the first designated area around the inspection robot with a network camera, and sends the first meter count value and the first robot inspection image to the server.

[0169] S13. The fixed-point cameras set up along the inspection route send the first image captured to the server.

[0170] S14. The server obtains the inspection results based on the DCS fault detection data, the first table count value, the first robot inspection image, and the first image. The inspection results include whether a fault occurred within the inspection area and the determined cause of the fault if it did occur.

[0171] Among them, the DCS (Distributed Control System) platform is a nuclear power plant control system deployed in the industrial site. It is used to control the production (e.g., total power generation, hourly power generation, etc.) and process management (e.g., adjusting steam temperature, adjusting valve opening, etc.) in the industrial site. It receives measurement data (e.g., steam temperature, pipeline pressure, flow rate, valve opening, pump status, etc.) from various measuring devices (e.g., temperature sensors, pressure sensors, etc.). It is existing technology and will not be described in detail here.

[0172] Both inspection robots and cameras are inspection devices. Non-contact sensors can be installed on inspection robots to collect environmental data such as temperature, humidity, air quality, and noise. The inspection robot can then send this environmental data to a server, providing data support for fault monitoring. The data acquired by these inspection devices is collectively referred to as inspection data (e.g., images captured by cameras, environmental data acquired by non-contact sensors), and this inspection data is one of the data foundations for the server's fault monitoring.

[0173] In one specific embodiment, in a scenario where non-contact sensors are installed on the inspection robot to collect on-site environmental data, the inspection robot also sends the environmental data to a server. The server obtains the inspection results based on DCS fault detection data, first meter count value, first robot inspection image, environmental data, and the first image.

[0174] The inspection robot can be a wheeled robot, such as a robot with a four-wheel eight-drive chassis, which supports wireless communications such as Wi-Fi (Wireless Fidelity) and 5G (5th Generation Mobile Communication Technology), and is equipped with a lifting arm. Both the macro-meter camera and the network camera can be set on the lifting arm to obtain better shooting positions and perspectives through the lifting arm. Server 1 can directly control the inspection robot and the lifting arm through instructions; Server 1 can also send instructions to the built-in controller of the inspection robot, and the built-in controller controls the inspection robot and the lifting arm. Controlling the robot to move, rotate, and controlling the lifting arm to perform three-dimensional movement and end rotation are all existing technologies and will not be elaborated here.

[0175] There are many problems in the existing inspection system, such as data isolation and no interaction between the inspection system and the nuclear power plant control system, the inspection tasks are carried out according to the pre-plan, the intelligence is insufficient, and there are problems such as the application integration of heterogeneous sensing devices. To solve these problems, the inspection centralized control platform system integrates servers, inspection robots and various cameras. The server supports multiple standard industrial protocols such as TCP / IP (Transmission Control Protocol / Internet Protocol), RTSP (Real-Time Streaming Protocol), and GB28181 (National Standard 28181) according to the device type; the server can be adapted through a gateway or a switch to support the access of IoT (Internet of Things devices) such as inspection robots, macro-meter cameras, network cameras, and fixed-point cameras; the server obtains process system data from the DCS platform and supports the parsing and structured processing of historical trend data, alarm records, control instructions, etc. The DCS fault detection triggers automatic inspection to realize functions such as dual-platform linkage monitoring, control, alarm, and inspection; in the regular inspection, the server can also combine the process system data for intelligent fault monitoring; establish a nuclear power plant intelligent inspection centralized control platform with the linkage function of intelligent inspection equipment and the nuclear power plant control system to solve the data island problem and realize the few-person duty and intelligent management and control of the inspection work.

[0176] Specifically, the inspection centralized control platform system dynamically loads the process system data in the running state of the DCS platform, and the inspection data obtained by the inspection centralized control platform system itself and the process system data are displayed double. The seamless integration of the real-time process system data of the DCS platform and the inspection data of the inspection centralized control platform system can be realized through a lightweight data engine. Figure 3 The schematic diagram of the internal data flow of the inspection centralized control platform system realizing the DCS process embodiment is shown, which is specifically reflected in the following aspects:

[0177] (1) Multi-source data acquisition

[0178] It comprehensively collects process system data (data reflecting the process flow, i.e., process data) from the DCS system, and interfaces with the DCS system through standard protocols such as OPC DA / UA (OPC Data Access / OPC Unified Architecture), Modbus (a protocol developed by Modicon), and Profibus (process fieldbus) to collect key process parameters (such as temperature, pressure, flow rate, valve opening, pump status, etc.) in real time. It supports the parsing and structured processing of historical trend data, alarm records, and control commands.

[0179] Meanwhile, intelligent inspection equipment, such as cameras and various non-contact sensors, collects on-site environmental data, including temperature, humidity, air quality, and noise. This data is transmitted in real time to the intelligent inspection big data storage and processing platform via wired or wireless communication, supporting the uploading of multimodal data such as images, audio, video, and sensor data.

[0180] (2) Data integration and storage

[0181] The intelligent inspection big data storage and processing platform employs a lightweight streaming data processing engine (such as Spark Streaming) to integrate the collected process system data and inspection data from the DCS platform. Data standardization is achieved by unifying the format of heterogeneous data from the DCS platform and inspection equipment. Timestamp alignment is implemented using an event-driven mechanism to synchronize data timelines. Data tagging is used to add multi-dimensional tags such as source, equipment type, spatial location, and inspection task ID. Dynamic data fusion constructs a logical association graph of "process variables - equipment status," supporting dynamic querying and joint analysis. A unified database is established, utilizing distributed storage technology, specifically the Hadoop Distributed File System (HDFS), to ensure efficient data storage and reliable management.

[0182] (3) Process visualization and inspection linkage display module

[0183] The unified visualization interface allows for a dual-view linkage mechanism between the process view and the inspection view. The process view displays the DCS process flow and real-time variables in the form of a flowchart or P&I (Pipeline & Instrumentation) diagram. The inspection view displays inspection data in the form of equipment topology diagrams, image monitoring, heat maps, etc.

[0184] Linkage mechanism: Selecting a process node (e.g., via mouse or touchscreen operation) will automatically display the inspection status and historical trends of related equipment; when the inspection system detects an anomaly, it will automatically highlight the corresponding process path and the scope of impact.

[0185] It supports issuing inspection tasks or DCS control strategy suggestions directly through a graphical interface. Inspection tasks can include the target equipment to be inspected and the inspection content, as well as the inspection route determined based on the target equipment and the inspection content. The target equipment to be inspected can be electrical equipment (such as sensors, steam generators, electric valves, etc.) or non-electrical equipment (such as pipelines, storage tanks, manual valves, etc.).

[0186] (4) Method for implementing automatic inspection triggered by DCS fault detection

[0187] A deep learning-based data fusion algorithm is proposed, using process system data (such as pressure and flow) from the DCS platform as the main feature and inspection perception data (such as images or infrared thermal imaging) as auxiliary features to establish an equipment health assessment model. A hierarchical distributed computing architecture is designed to implement lightweight diagnostics (anomaly detection of key parameters in process system data) at the DCS controller layer, and to complete plant-wide equipment degradation prediction in the cloud, establishing a closed-loop logic between inspection results and DCS control correction.

[0188] A hierarchical distributed computing architecture is adopted: Edge layer (DCS controller in DCS platform): Deploy lightweight anomaly detection model, such as 1D-CNN (one-dimensional convolutional neural network) + LSTM (long short-term memory network), to monitor key parameters (pressure, flow, etc.) in real time, with a response time ≤100ms, and trigger the automatic threshold calibration function.

[0189] The inspection and control platform system backend employs a multimodal data fusion model to integrate process system data and inspection data from the DCS platform. It models equipment topology relationships using a graph neural network (GNN) and stores structured expert knowledge rules (emphasizing IF-THEN rules; supporting attribute-value pair-based condition combinations; and supporting rule priority, confidence level, and scope of application). The system backend receives status data, invokes knowledge rules for reasoning, and outputs diagnostic results and risk levels. Based on the diagnostic results, it dynamically generates inspection task plans, including inspection time, personnel, inspection locations, tools, and precautions. The system integrates templated inspection schemes through software and triggers automatic inspection execution, particularly supporting multi-dimensional retrieval and intelligent analysis of the operating status of key equipment in nuclear power plants (such as turbines, boilers, main pumps, and valves).

[0190] The inspection and control platform system, based on fused multi-source data, constructs an equipment health assessment model using predictive analytics. It employs machine learning algorithms (such as LSTM, random forest, and Bayesian networks) to predict the lifespan and failure probability of key equipment; combines process variable trends to deduce potential fault propagation paths; and links DCS control parameters (such as setpoint and threshold adjustments) to achieve preventative regulation. It automatically generates inspection strategies: dynamically adjusting inspection frequency, task priority, and inspection methods based on prediction results; supporting templated matching and adaptive optimization of inspection tasks; and allowing inspection task suggestions to be fed back to the DCS platform via OPC DA / UA or API (Application Programming Interface).

[0191] In this embodiment, a server supporting multiple standard industrial protocols communicates with various IoT devices such as inspection robots and macro-meter cameras to achieve unified management and data integration within the system. Process system data is dynamically loaded while the DCS platform is running. A deep learning-based equipment health assessment model analyzes the process system data and the inspection data from the inspection control platform system to monitor faults in the industrial field. This enables dual-platform linkage monitoring, control, alarm, and inspection functions between the DCS platform and the inspection control platform system. DCS fault detection can trigger automatic inspections, enabling rapid response and fault handling. Furthermore, the system correlates inspection route planning with fault knowledge data to optimize inspection routes, improve inspection efficiency and accuracy, reduce manual intervention, and lower maintenance costs.

[0192] In an optional embodiment, refer to Figure 5 The control methods also include:

[0193] S15. Configure the inspection route for the inspection robot on the server.

[0194] S16. The inspection robot performs inspections according to the inspection route set by the server. It obtains second meter images by taking pictures of the physical meters of all measuring devices along the inspection route using a macro meter camera, and obtains second meter count values. It also takes second robot inspection images of a first designated area around the inspection robot using a network camera, and sends the second meter count values ​​and second robot inspection images to the server.

[0195] S17. The fixed-point cameras set up along the inspection route send the captured second images to the server.

[0196] S18. The server compares the count value of the second table, the inspection image of the second robot, and the second image with the process system data of the DCS platform to obtain the inspection result. The inspection result also includes whether the measuring equipment in the DCS platform is malfunctioning.

[0197] Even without DCS fault detection to trigger automatic inspections, regular inspections are still necessary to eliminate potential hazards and faults; this means routine daily inspections need to be scheduled. During routine inspections, intelligent fault monitoring can be implemented using process system data, allowing for the setting of inspection routes, inspection content, and other inspection tasks.

[0198] In one specific embodiment, in a scenario where non-contact sensors are installed on the inspection robot to collect on-site environmental data, the inspection robot also sends the environmental data to the server. The server compares the second table count value, the second robot inspection image, the environmental data, and the second image with the process system data of the DCS platform to obtain the inspection result.

[0199] In this embodiment, the server can also perform intelligent fault monitoring by combining process system data during routine inspections; a nuclear power plant intelligent inspection centralized control platform with the function of linking intelligent inspection equipment with the nuclear power plant control system is established to solve the problem of data silos and realize less manned operation and intelligent management of inspection work.

[0200] In an optional embodiment, the DCS fault detection data is obtained by the DCS controller layer in the DCS platform through lightweight diagnostics based on process system data.

[0201] Step S14 includes:

[0202] The server takes DCS fault detection data, first table count values, first robot inspection images, and first images as input data, and outputs inspection results through the equipment health assessment model. The equipment health assessment model is built based on a deep learning-based predictive analytics model.

[0203] The inspection and control platform system, based on fused multi-source data, constructs an equipment health assessment model using predictive analytics models. It employs machine learning algorithms (such as LSTM, random forest, and Bayesian networks) to predict the lifespan and failure probability of key equipment; combines process variable trends to deduce potential fault propagation paths; and links DCS control parameters (such as setpoint and threshold adjustments) to achieve preventative regulation. It automatically generates inspection strategies: dynamically adjusting inspection frequency, task priority, and inspection methods based on prediction results; supporting templated matching and adaptive optimization of inspection tasks; and allowing inspection task suggestions to be fed back to the DCS platform via OPCDA / UA or API (Application Programming Interface).

[0204] In this embodiment, the equipment health assessment model uses machine learning algorithms (such as LSTM, random forest, Bayesian network) to predict the lifespan and failure probability of key equipment; combines process variable trends to deduce potential fault propagation paths; and links DCS control parameters (such as setpoints, threshold adjustments, etc.) to achieve preventative regulation.

[0205] In an optional embodiment, step S18 includes:

[0206] The server takes the count value of the second table, the second robot inspection image, and the process system data as input data, and outputs the inspection results through the equipment health assessment model.

[0207] In this embodiment, during routine inspections, the server can also perform intelligent fault monitoring by combining process system data and obtain inspection results through the equipment health assessment model; a nuclear power plant intelligent inspection centralized control platform with the function of linking intelligent inspection equipment with the nuclear power plant control system is established to solve the problem of data silos and realize less manned operation and intelligent management of inspection work.

[0208] In an optional embodiment, step S11 includes:

[0209] Based on DCS fault detection data and fault knowledge data stored in the inspection knowledge base host, the server determines the first target device that makes up the inspection point and the inspection content of the first target device, and sets the inspection route according to the location of the first target device and the connection relationship of related devices.

[0210] Among them, the inspection and control platform system can adopt a multimodal data fusion model to integrate the process system data and inspection data of the DCS platform. It can model the topological relationship of equipment through graph neural networks, store structured expert knowledge rules, receive status data, call knowledge rules for reasoning, and output diagnostic results and risk levels. Then, based on the diagnostic results, it can dynamically generate inspection tasks (including inspection routes and inspection content).

[0211] In this embodiment, when DCS fault detection triggers automatic inspection, the fault detection data is combined with fault knowledge data composed of historical data to optimize the inspection route and arrange inspection tasks in a more targeted manner, thereby improving inspection efficiency and accuracy, reducing manual intervention, and lowering operation and maintenance costs.

[0212] In an optional embodiment, step S15 includes:

[0213] Based on the fault knowledge data stored in the inspection knowledge base host, the server determines the inspection points and the detection content of the measuring devices at the inspection points, and sets the inspection route according to the location of the measuring devices at the inspection points and the connection relationship of related devices.

[0214] Among them, the inspection and control platform system can adopt a multimodal data fusion model to integrate the process system data and inspection data of the DCS platform. It can model the topological relationship of equipment through graph neural networks, store structured expert knowledge rules, receive status data, call knowledge rules for reasoning, and output diagnostic results and risk levels. Then, based on the diagnostic results, it can dynamically generate inspection tasks (including inspection routes and inspection content).

[0215] In this embodiment, in the absence of DCS fault detection that triggers automatic inspection, the fault knowledge data composed of DCS fault detection data and historical data is combined during routine inspections to optimize inspection routes and schedule inspection tasks in a more targeted manner, thereby improving inspection efficiency and accuracy, reducing manual intervention, and lowering operation and maintenance costs.

[0216] While specific embodiments of this disclosure have been described above, those skilled in the art should understand that these are merely illustrative examples, and the scope of protection of this disclosure is defined by the appended claims. Those skilled in the art can make various changes or modifications to these embodiments without departing from the principles and essence of this disclosure, but all such changes and modifications fall within the scope of protection of this disclosure.

Claims

1. A centralized inspection and control platform system, characterized in that, The inspection and control platform system includes: a server, an inspection robot, and cameras deployed in the central control room; The server supports multiple standard industrial protocols for communication with the inspection robot and the camera, respectively. The server is also connected to the DCS platform to acquire process system data from the DCS platform; wherein the process system data includes measurement data sent by several measuring devices. The cameras include a macro meter camera and a network camera mounted on the inspection robot, as well as a fixed-point camera set at a fixed location; the network camera is a wide-angle camera and supports an open application platform; The inspection robot is used to perform inspections according to the inspection route set by the server. It obtains meter images by taking pictures of the physical meters of the measuring equipment on the inspection route using the macro meter camera to obtain meter count values. It also takes robot inspection images of a first designated area around the inspection robot using the network camera and sends the meter count values ​​and robot inspection images to the server. The fixed-point camera is used to capture a fixed-point image of a matching preset area and send the fixed-point image to the server; The server is also used to perform fault detection based on at least two of the process system data, the table count values, the robot inspection images, and the fixed-point images.

2. The inspection and control platform system as described in claim 1, characterized in that, The fixed-point camera includes a high-altitude spherical eagle-eye camera; The server is also communicatively connected to the high-altitude spherical eagle-eye camera. The high-altitude spherical eagle-eye camera is installed on the roof and is used to monitor objects and people in the distance; And / or, The fixed-point camera includes a helmet detection camera; The server is also communicatively connected to the helmet detection camera. The helmet detection camera is used to detect whether the first target person is wearing a helmet; And / or, The fixed-point camera includes a face recognition camera; The server is also communicatively connected to the face recognition camera; The face recognition camera is used to identify the face of a second target person in order to verify their identity; And / or, The fixed-point camera includes a personnel detection camera; The server is also communicatively connected to the personnel detection camera; The personnel detection camera is used to monitor the target location and to collect statistics on at least one of passenger flow, number of people, and personnel density. And / or, The fixed-point camera includes a behavior recognition camera; The server is also communicatively connected to the behavior recognition camera; The behavior recognition camera is used to recognize the body movements of third-party target personnel, and supports behavior analysis of falling, violent exercise, abnormal spacing, lingering, abnormal number of people, crossing the warning line, wandering, running, and leaving the post; And / or, The inspection and control platform system also includes a lidar mounted on the inspection robot; The server is also communicatively connected to the lidar; The lidar is used to acquire laser data of a second designated area around the inspection robot.

3. The inspection and control platform system as described in claim 1, characterized in that, The inspection and control platform system also includes a meter inspection host; The server is also communicatively connected to the meter inspection host. The meter inspection host is used to store the meter count values ​​obtained from the meter images; And / or, The inspection and control platform system also includes a hard disk video recorder; The server is also communicatively connected to the hard disk recorder; The hard disk recorder is used to store images captured by the camera; And / or, The inspection and control platform system also includes an inspection knowledge base host; The server is also connected to the inspection knowledge base host. The inspection knowledge base host is used to store the processing results of each inspection. And / or, The inspection and control platform system also includes interactive devices; The server is also communicatively connected to the interactive device; The interactive device is used to send interactive data to the server and display the graphic and text data sent by the server.

4. The inspection and control platform system as described in claim 1, characterized in that, The server uses Spark Streaming as the real-time streaming data processing engine and constructs a logical relationship graph between measurement devices within the DCS platform. The inspection and control platform system adopts a unified data format and standard, and the data within the system uses a unified timestamp, adding semantic tags to the data within the system.

5. A control method for an inspection centralized control platform system, characterized in that, The control method is applied to the inspection centralized control platform system as described in any one of claims 1-4, and the control method includes: The server determines the first target equipment to be inspected based on the DCS fault detection data in the process system data of the DCS platform, and sets the inspection route of the inspection robot based on the first target equipment. The inspection robot performs inspections according to the inspection route, obtains meter images by taking pictures of the physical meters of the first target device with a macro meter camera to obtain the first meter count value, and takes first robot inspection images of a first designated area around the inspection robot with a network camera, and sends the first meter count value and the first robot inspection images to the server. The fixed-point cameras set along the inspection route send the captured first images to the server. The server obtains inspection results based on the DCS fault detection data, the first table count value, the first robot inspection image, and the first image; wherein, the inspection results include whether a fault has occurred in the inspection area and the cause of the fault when it occurs.

6. The control method of the inspection centralized control platform system as described in claim 5, characterized in that, The control method further includes: The server sets the inspection route for the inspection robot; The inspection robot performs inspections according to the inspection route set by the server. It obtains second meter images by taking pictures of the physical meters of all measuring devices along the inspection route using the macro meter camera, and obtains second meter count values. It also takes second robot inspection images of a first designated area around the inspection robot using the network camera, and sends the second meter count values ​​and the second robot inspection images to the server. The fixed-point cameras set along the inspection route send the captured second images to the server. The server compares the second table count value, the second robot inspection image, and the second image with the process system data of the DCS platform to obtain the inspection result; wherein, the inspection result also includes whether the measuring equipment in the DCS platform has malfunctioned.

7. The control method of the inspection centralized control platform system as described in claim 6, characterized in that, The DCS fault detection data is obtained by the DCS controller layer in the DCS platform through lightweight diagnosis based on the process system data. The server obtains inspection results based on the DCS fault detection data, the first table count value, the first robot inspection image, and the first image, including: The server takes the DCS fault detection data, the first table count value, the first robot inspection image, and the first image as input data, and outputs the inspection results through the equipment health assessment model; wherein, the equipment health assessment model is constructed based on a deep learning predictive analysis model.

8. The control method of the inspection centralized control platform system as described in claim 7, characterized in that, The server compares the second table count value, the second robot inspection image, and the second image with the process system data of the DCS platform to obtain the inspection result, including: The server takes the second table count value, the second robot inspection image, the second image, and the process system data as input data, and outputs the inspection results through the equipment health assessment model.

9. The control method of the inspection centralized control platform system as described in claim 5, characterized in that, The server determines the first target equipment to be inspected based on the DCS fault detection data in the process system data of the DCS platform, and sets the inspection route of the inspection robot according to the first target equipment, including: Based on the DCS fault detection data and the fault knowledge data stored in the inspection knowledge base host, the server determines the first target device that makes up the inspection point and the inspection content for the first target device, and sets the inspection route according to the location of the first target device and the connection relationship of related devices.

10. The control method of the inspection centralized control platform system as described in claim 6, characterized in that, The server sets the inspection route for the inspection robot, including: The server determines the inspection points and the detection content of the measuring devices at the inspection points based on the fault knowledge data stored in the inspection knowledge base host, and sets the inspection route according to the location of the measuring devices at the inspection points and the connection relationship of related devices.

Images