Nuclear power station flow channel combined dredging robot applying image recognition algorithm

Abstract

The invention provides a nuclear power station flow channel joint dredging robot applying an image recognition algorithm. The robot comprises an image recognition sensing system, a robot moving navigation system, a dredging operation system and a safety monitoring system. The image recognition sensing system is used for collecting an image in the flow channel through a high-resolution camera and analyzing and recognizing sludge and sundries in the flow channel according to the image; the robot mobile navigation system is used for realizing positioning, path planning and autonomous movement of the robot; the desilting operation system is used for removing and collecting sludge and sundries in the flow channel; the safety monitoring system is used for monitoring the running state and the working environment of the robot in real time and ensuring the safety and the reliability of the dredging process of the robot; by integrating image recognition, automatic navigation and a safety monitoring system, efficient, accurate and safe flow channel dredging operation is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of nuclear power plant dredging, and in particular to a nuclear power plant flow channel joint dredging robot applying an image recognition algorithm. Background Art

[0002] During the operation of nuclear power plants, a large amount of silt and debris often accumulates in the flow channels, which not only affects the patency of the flow channels, but may also cause equipment failure and safety hazards. Currently, dredging operations mostly rely on manual operations, which are not only labor-intensive and inefficient, but also involve high operational risks. Especially in narrow and complex flow channel environments, traditional dredging methods are difficult to ensure the thoroughness and accuracy of dredging. Therefore, the development of an automated and intelligent dredging robot that can efficiently and accurately complete the dredging task in the flow channel while ensuring safety has important practical significance and application value.

[0003] According to the related disclosed technical solutions, the technology with publication number CN113713485A discloses an automated desilting system for a nuclear power plant and a desilting system for a rotary filter screen, comprising a mixing tank arranged at the bottom of the rotary filter screen, a cam pump arranged at the bottom of the rotary filter screen and connected to the outlet of the mixing tank, a siltation storage tank arranged on the ground near the top of the rotary filter screen, a solid-liquid vibrating screen arranged above the siltation storage tank, a return pipe connected between the inlet of the mixing tank and the siltation storage tank, and a water supply pipe connected between the cam pump and the solid-liquid vibrating screen; the solution After the marine biological garbage at the bottom of the rotating filter forms a solid-liquid mixture, it is powered by a cam pump and lifted to the solid-liquid vibrating screen on the ground for solid-liquid separation, thereby achieving odor-free cleaning of the garbage and separation of sludge and marine organisms, reducing operational risks and labor intensity; it is suitable for installation and use in water-scarce or waterless conditions and narrow on-site spaces; however, this solution is mainly suitable for dredging operations at fixed locations and cannot flexibly adapt to sludge accumulation at different locations in the flow channel; in addition, the solution lacks real-time monitoring and dynamic adjustment functions based on image recognition and three-dimensional models, making it difficult to achieve efficient and accurate dredging management. Summary of the invention

[0004] The purpose of the present invention is to propose a nuclear power plant flow channel joint dredging robot using an image recognition algorithm in response to the current deficiencies.

[0005] The present invention adopts the following technical solution:

[0006] A nuclear power plant flow channel joint dredging robot using an image recognition algorithm, the robot comprising an image recognition sensor system, a robot mobile navigation system, a dredging operation system and a safety monitoring system; the image recognition sensor system is used to collect images in the flow channel through a high-resolution camera, and identify silt and debris in the flow channel based on image analysis; the robot mobile navigation system is used to achieve robot positioning, path planning and autonomous movement; the dredging operation system is used to complete the removal and collection of silt and debris in the flow channel; the safety monitoring system is used to monitor the robot's operating status and working environment in real time to ensure the safety and reliability of the robot's dredging process;

[0007] The image recognition sensor system includes an image acquisition module and an image processing and recognition module; the image acquisition module is used to collect images in the flow channel of the nuclear power plant in real time through a high-resolution camera; the image processing and recognition module is used to process and analyze the collected images to identify silt and debris in the flow channel;

[0008] The robot mobile navigation system includes a navigation positioning module and a drive control module; the navigation positioning module uses laser radar technology to locate the position of the robot in the flow channel in real time and plans the movement path of the robot; the drive control module is used to perform motion control of the robot and control the movement and walking of the robot according to the movement path instructions of the navigation positioning module;

[0009] The desilting operation system includes a desilting operation device, a desilting tool and a sludge processing and collecting device; the desilting operation device is specifically a multifunctional mechanical arm, and the multifunctional mechanical arm can operate the desilting tool to remove sludge and debris in the flow channel; the sludge processing and collecting device is used to collect and process the removed sludge and debris;

[0010] The safety monitoring system includes a real-time monitoring module and an alarm module; the real-time monitoring module is used to monitor the robot's operating status and working environment in real time through a variety of sensors, and the alarm module is used to issue an alarm message when an abnormality is detected in the robot's operating status or working environment, so as to remind the staff to deal with potential problems in a timely manner;

[0011] A method for desilting a flow channel of a nuclear power plant using an image recognition algorithm and a robot, the method comprising the following steps:

[0012] S11: Use laser radar technology to locate the robot in the flow channel in real time, and plan the optimal movement path based on the positioning results;

[0013] S12: transmitting the planned moving path instruction to the drive control module, and the drive control module controls the robot to move along the planned moving path according to the moving path instruction;

[0014] S13: When the robot moves on the moving path, a high-resolution camera installed at the front end of the robot is used to collect images in the flow channel in real time;

[0015] S14: collecting the acquired image data, and comparing and analyzing the image data with a preset silt and debris recognition model to identify the location of the silt and debris in the flow channel;

[0016] S15: Execute the dredging process: start the dredging operation device in the dredging operation system, and use the multifunctional mechanical arm to manipulate the dredging tool to clean and collect the identified silt and debris;

[0017] S16: After the dredging is completed, the image of the silt and debris position identified in step S14 is collected again, and the dredging effect is checked in combination with the silt and debris accumulation before the dredging. If the check fails, the dredging process of step S15 is repeated until the test is qualified;

[0018] S17: After the dredging inspection is completed, steps S12 to S16 are repeated until the robot moves to the end of the planned moving path;

[0019] Furthermore, in step S11, the robot movement path is planned by:

[0020] S111: Scan the robot's surrounding environment data through the laser radar, and construct point cloud data in the flow channel according to the data collected by the laser radar;

[0021] S112: constructing an initial map of the flow channel according to the point cloud data obtained in step S111, wherein the initial map is a basic representation of the environment in the flow channel;

[0022] S113: performing path planning according to the path planning algorithm combined with the initial map, and generating a moving path instruction to transmit to the drive control module;

[0023] S114: Setting a path update cycle, re-execute steps S111 to S113 to update the path planning at each path update cycle time point or after the dredging operation is completed in the subsequent steps, and generate a new moving path instruction to transmit to the drive control module;

[0024] Furthermore, in step S14, the identification of sludge and debris in the flow channel is completed by the following method:

[0025] S141: Acquire an image in the flow channel, and perform a preprocessing operation on the acquired image; the preprocessing operation includes image denoising and image enhancement processing;

[0026] S142: Compare the preprocessed image with a preset silt and debris recognition model to quickly and preliminarily detect areas where silt and debris exist; the silt and debris recognition model is specifically a model obtained by training with a large amount of labeled silt and debris image data; the specific process of the rapid preliminary detection meets the following requirements:

[0027] The preprocessed image is divided into multiple detection areas, and for each detection area, the following is performed:

[0028]

[0029] Among them, F d is a feature vector set of a certain detection area, wherein the feature vectors in the feature vector set include shape feature vectors, texture feature vectors and color feature vectors; F m is the feature vector set in the sludge debris recognition model; Sim(F d ,F m ) is the feature matching degree between a certain detection area in the image and the silt debris recognition model, and a matching degree threshold is set. When the feature matching degree between a certain detection area and the silt debris recognition model is higher than the matching degree threshold, it is marked that the detection area has silt debris;

[0030] S143: When an area with silt and debris is detected, the robot stops moving and adjusts the angle of the camera so that the camera can focus and obtain an image of the center position of the silt and debris area; the specific implementation process is as follows:

[0031] S1431: Detect the edge of the silt and debris by using an edge detection algorithm, and control the position and direction of the camera to slowly turn to the edge of the silt and debris; calculate the rate of increase of edge pixels of the image newly entering the camera range during the camera rotation process, and when the rate of increase of edge pixels is lower than the set rate of increase threshold, stop the rotation of the camera;

[0032] S1432: re-detecting the silt and debris area within the camera's field of view, calculating the centroid position of the silt and debris area, and aligning the camera center with the centroid position to obtain a center position image of the silt and debris area;

[0033] S144: Calculating the edge and shape features of the silt debris area according to the center position image of the silt debris area, so as to guide the subsequent dredging process;

[0034] Furthermore, in the step S16, the dredging effect is checked by:

[0035] K=α1·M h +α2·L h ;

[0036] Where K is the dredging inspection coefficient, α1 and α2 are preset weight coefficients, satisfying α1+α2=1; M h is the area of ​​silt and debris in the silt and debris position image after dredging, L h is the perimeter of the silt and debris area in the silt and debris position image after dredging;

[0037] By comparing the silt removal inspection coefficient with the set silt removal inspection threshold, when the silt removal inspection coefficient is less than the set silt removal inspection threshold, it is determined that the silt removal effect is good, otherwise the silt removal operation in step S15 is repeated.

[0038] The beneficial effects achieved by the present invention are:

[0039] The present invention uses a high-resolution camera to collect and analyze images in the flow channel in real time, so as to accurately identify the location and shape characteristics of silt and debris; uses laser radar technology to perform real-time positioning and path planning to ensure that the robot can move autonomously and efficiently in the flow channel, thereby achieving precise dredging operations; uses a multi-functional robotic arm and dredging tools to perform automated dredging, thereby greatly reducing manual labor intensity and operational risks; after dredging is completed, the dredging effect is checked by setting a dredging inspection threshold to ensure that each dredging meets the expected standards, thereby improving the long-term patency and safety of the flow channel. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present invention can be further understood from the following description in conjunction with the accompanying drawings. The components in the figures are not necessarily drawn to scale, but the emphasis is placed on illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0041] Figure 1 It is a schematic diagram of the overall system module of the robot of the present invention.

[0042] Figure 2 The present invention is a schematic flow chart of a method for desilting a nuclear power plant flow channel using a joint robot using an image recognition algorithm.

[0043] Figure 3 The present invention is a schematic flow chart of a method for obtaining the sludge accumulation condition in a flow channel of a nuclear power plant. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with its embodiments; it should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention; for those skilled in the art, other systems, methods and / or features of the present embodiment will become apparent after reviewing the following detailed description; it is intended that all such additional systems, methods, features and advantages are included in this specification; included within the scope of the present invention and protected by the appended claims; additional features of the disclosed embodiments are described in the following detailed description, and these features will be apparent from the following detailed description.

[0045] The same or similar numbers in the drawings of the embodiments of the present invention correspond to the same or similar parts; in the description of the present invention, it should be understood that if the terms "upper", "lower", "left", "right" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the drawings, it is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or component referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, the terms describing the positional relationship in the drawings are only used for illustrative purposes and cannot be understood as limitations on this patent. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0046] Embodiment 1:

[0047] like Figure 1 As shown, this embodiment provides a nuclear power plant flow channel joint dredging robot using an image recognition algorithm, the robot comprising an image recognition sensor system, a robot mobile navigation system, a dredging operation system and a safety monitoring system; the image recognition sensor system is used to collect images in the flow channel through a high-resolution camera, and identify silt and debris in the flow channel based on image analysis; the robot mobile navigation system is used to achieve robot positioning, path planning and autonomous movement; the dredging operation system is used to complete the removal and collection of silt and debris in the flow channel; the safety monitoring system is used to monitor the robot's operating status and working environment in real time to ensure the safety and reliability of the robot's dredging process;

[0048] The image recognition sensor system includes an image acquisition module and an image processing and recognition module; the image acquisition module is used to collect images in the flow channel of the nuclear power plant in real time through a high-resolution camera; the image processing and recognition module is used to process and analyze the collected images to identify silt and debris in the flow channel;

[0049] Furthermore, the image acquisition module specifically includes a camera installed at the front end of the robot, and the camera can be adjusted at different angles and positions to ensure the acquisition of clear image data;

[0050] The robot mobile navigation system includes a navigation positioning module and a drive control module; the navigation positioning module uses laser radar technology to locate the position of the robot in the flow channel in real time and plans the movement path of the robot; the drive control module is used to perform motion control of the robot and control the movement and walking of the robot according to the movement path instructions of the navigation positioning module;

[0051] The desilting operation system includes a desilting operation device, a desilting tool and a sludge processing and collecting device; the desilting operation device is specifically a multifunctional mechanical arm, and the multifunctional mechanical arm can operate the desilting tool to remove sludge and debris in the flow channel; the sludge processing and collecting device is used to collect and process the removed sludge and debris;

[0052] The safety monitoring system includes a real-time monitoring module and an alarm module; the real-time monitoring module is used to monitor the robot's operating status and working environment in real time through a variety of sensors, and the alarm module is used to issue an alarm message when an abnormality is detected in the robot's operating status or working environment, so as to remind the staff to deal with potential problems in a timely manner;

[0053] like Figure 2 As shown, this embodiment also provides a method for desilting a flow channel of a nuclear power plant using an image recognition algorithm and a robot, the method comprising the following steps:

[0054] S11: Use laser radar technology to locate the robot in the flow channel in real time, and plan the optimal movement path based on the positioning results;

[0055] S12: transmitting the planned moving path instruction to the drive control module, and the drive control module controls the robot to move along the planned moving path according to the moving path instruction;

[0056] S13: When the robot moves on the moving path, a high-resolution camera installed at the front end of the robot is used to collect images in the flow channel in real time;

[0057] S14: collecting the acquired image data, and comparing and analyzing the image data with a preset silt and debris recognition model to identify the location of the silt and debris in the flow channel;

[0058] S15: Execute the dredging process: start the dredging operation device in the dredging operation system, and use the multifunctional mechanical arm to manipulate the dredging tool to clean and collect the identified silt and debris;

[0059] S16: After the dredging is completed, the image of the silt and debris position identified in step S14 is collected again, and the dredging effect is checked in combination with the silt and debris accumulation before the dredging. If the check fails, the dredging process of step S15 is repeated until the test is qualified;

[0060] S17: After the dredging inspection is completed, steps S12 to S16 are repeated until the robot moves to the end of the planned moving path;

[0061] Furthermore, in step S11, the robot movement path is planned specifically by:

[0062] S111: Scan the robot's surrounding environment data through the laser radar, and construct point cloud data in the flow channel according to the data collected by the laser radar;

[0063] S112: constructing an initial map of the flow channel according to the point cloud data obtained in step S111, wherein the initial map is a basic representation of the environment in the flow channel;

[0064] S113: performing path planning according to the path planning algorithm combined with the initial map, and generating a moving path instruction to transmit to the drive control module;

[0065] S114: Setting a path update cycle, re-execute steps S111 to S113 to update the path planning at each path update cycle time point or after the dredging operation is completed in the subsequent steps, and generate a new moving path instruction to transmit to the drive control module;

[0066] Furthermore, in step S14, the identification of sludge and debris in the flow channel is completed specifically in the following manner:

[0067] S141: Acquire an image in the flow channel, and perform a preprocessing operation on the acquired image; the preprocessing operation includes image denoising and image enhancement processing;

[0068] S142: Compare the preprocessed image with a preset silt and debris recognition model to quickly and preliminarily detect areas where silt and debris exist; the silt and debris recognition model is specifically a model obtained by training with a large amount of labeled silt and debris image data; the specific process of the rapid preliminary detection meets the following requirements:

[0069] The preprocessed image is divided into multiple detection areas, and for each detection area, the following is performed:

[0070]

[0071] Among them, F d is a feature vector set of a certain detection area, wherein the feature vectors in the feature vector set include shape feature vectors, texture feature vectors and color feature vectors; F m is the feature vector set in the sludge debris recognition model; Sim(F d ,F m) is the feature matching degree between a certain detection area in the image and the silt debris recognition model, and a matching degree threshold is set. When the feature matching degree between a certain detection area and the silt debris recognition model is higher than the matching degree threshold, it is marked that the detection area has silt debris;

[0072] S143: When an area with silt and debris is detected, the robot stops moving and adjusts the angle of the camera so that the camera can focus and obtain an image of the center position of the silt and debris area; the specific implementation process is as follows:

[0073] S1431: Detect the edge of the silt and debris by using an edge detection algorithm, and control the position and direction of the camera to slowly turn to the edge of the silt and debris; calculate the rate of increase of edge pixels of the image newly entering the camera range during the camera rotation process, and when the rate of increase of edge pixels is lower than the set rate of increase threshold, stop the rotation of the camera;

[0074] S1432: re-detecting the silt and debris area within the camera's field of view, calculating the centroid position of the silt and debris area, and aligning the camera center with the centroid position to obtain a center position image of the silt and debris area;

[0075] S144: Calculating the edge and shape features of the silt debris area according to the center position image of the silt debris area, so as to guide the subsequent dredging process;

[0076] Furthermore, in step S16, the desilting effect is checked by:

[0077] K=α1·M h +α2·L h ;

[0078] Where K is the dredging inspection coefficient, α1 and α2 are preset weight coefficients, satisfying α1+α2=1; M h is the area of ​​silt and debris in the silt and debris position image after dredging, L h is the perimeter of the silt and debris area in the silt and debris position image after dredging;

[0079] By comparing the silt removal inspection coefficient with the set silt removal inspection threshold, when the silt removal inspection coefficient is less than the set silt removal inspection threshold, it is determined that the silt removal effect is good, otherwise the silt removal operation in step S15 is repeated;

[0080] Further, the dredging inspection threshold is set in the following manner;

[0081]

[0082] Among them, Q is the dredging inspection threshold, q is the preset pollution inspection benchmark value, and M gis the area of ​​silt and debris in the silt and debris position image before dredging, L g is the perimeter of the silt debris area in the silt debris position image after dredging, β is the reduction coefficient, and its value range is (0.5,1);

[0083] Because the more serious the accumulation of silt and debris in an area, the more likely it is that new silt will accumulate. After dredging is completed, the dredging inspection threshold is set according to the accumulation of silt and debris in the area before dredging, so as to complete the inspection of the dredging effect. This can ensure that more stringent inspections are carried out in areas with serious silt accumulation, ensure the dredging effect, prevent new silt accumulation, and improve the long-term patency and safety of the flow channel.

[0084] Embodiment 2:

[0085] This embodiment should be understood to include at least all the features of any of the above embodiments, and further improve upon them;

[0086] In this embodiment, the combined dredging robot can also obtain the silt accumulation situation in the flow channel of the nuclear power plant according to the historical dredging situation; Figure 3 As shown, this embodiment provides a method for obtaining the sludge accumulation condition in the flow channel of a nuclear power plant, and the method comprises the following steps:

[0087] S21: Acquire historical desilting data of n consecutive times in the flow channel of the nuclear power plant, wherein the historical desilting data includes edge and shape feature information obtained from the center position image of each silt and debris area in the flow channel before desilting;

[0088] S22: For each historical desilting data, obtain the centroid position coordinates of all silt and debris areas, and calculate the silt and debris cumulative degree value of the centroid position coordinates:

[0089] L i =α1·M g,i +α2·L g,i ;

[0090] Among them, L i is the cumulative degree of silt and debris at a certain centroid position coordinate in the i-th historical desilting data, M g,i is the area of ​​the silt debris region in the silt debris position image before silting corresponding to the centroid position coordinates in the i-th historical silting data, L g,i is the perimeter of the silt debris area in the silt debris position image before silting corresponding to the centroid position coordinates in the i-th historical silting data;

[0091] S23: establishing a simulated three-dimensional model of the flow channel of the nuclear power plant; substituting the coordinates of each centroid position in each historical dredging data into the simulated three-dimensional model, and rendering the simulated three-dimensional model with each centroid position coordinate as the center and radiating to the surrounding areas;

[0092] S24: providing the rendered simulated three-dimensional model to the user, and repeating steps S21 to S23 to update the simulated three-dimensional model each time new historical dredging data is obtained;

[0093] Furthermore, in step S23, the divergence area and the divergence color intensity on each centroid position coordinate are determined by calculating the divergence index of each centroid position coordinate; the specific calculation method of the divergence index is as follows:

[0094]

[0095] Among them, S i is the divergence index of the coordinates of a centroid position in the i-th historical desilting data. The larger i is, the closer the desilting data is to the current time. γ is a preset adjustment parameter with a value range of (0, 0.5).

[0096] Furthermore, the divergence area is determined by the following formula:

[0097] A i = k·S i ;

[0098] Among them, A i is the divergence area of ​​the coordinates of a centroid position in the i-th historical desilting data, k is the mapping proportional constant, which is determined according to the pre-experimental determination;

[0099] Further, the divergent color intensity is determined by the following formula:

[0100]

[0101] Among them, C i is the divergent color intensity of a centroid position coordinate in the i-th historical desilting data, with a value range of [0,1], representing the change of color from light to deep; S max is the preset normalization coefficient;

[0102] This embodiment constructs a simulated three-dimensional model of the nuclear power plant flow channel and uses historical dredging data analysis to render the simulated three-dimensional model, so that the staff can intuitively view the silt accumulation in the flow channel; by calculating the divergence index of the silt debris area, the divergence area and the divergence color intensity are determined, so that the visualization of the accumulation situation is more accurate and detailed; the simulated three-dimensional model can not only help the staff identify areas prone to accumulation and optimize dredging strategies, but also provide an important basis for the repair and maintenance of the flow channel.

[0103] The contents disclosed above are only preferred feasible embodiments of the present invention, and do not limit the protection scope of the present invention. Therefore, all equivalent technical changes made using the contents of the present invention specification and drawings are included in the protection scope of the present invention. In addition, the elements therein can be updated as technology develops.

Claims

1. A nuclear power plant flow channel joint dredging robot using an image recognition algorithm, characterized in that: The robot includes an image recognition sensor system, a robot mobile navigation system, a silt removal operation system and a safety monitoring system; the image recognition sensor system is used to collect images in the flow channel through a high-resolution camera, and identify silt and debris in the flow channel based on image analysis; the robot mobile navigation system is used to achieve robot positioning, path planning and autonomous movement; the silt removal operation system is used to complete the removal and collection of silt and debris in the flow channel; the safety monitoring system is used to monitor the robot's operating status and working environment in real time to ensure the safety and reliability of the robot's silt removal process; The image recognition sensor system includes an image acquisition module and an image processing and recognition module; the image acquisition module is used to collect images in the flow channel of the nuclear power plant in real time through a high-resolution camera; the image processing and recognition module is used to process and analyze the collected images to identify silt and debris in the flow channel; The robot mobile navigation system includes a navigation and positioning module and a drive control module; The navigation and positioning module uses laser radar technology to locate the position of the robot in the flow channel in real time and plans the movement path of the robot; the drive control module is used to perform motion control of the robot and control the movement of the robot according to the movement path instructions of the navigation and positioning module; The desilting operation system includes a desilting operation device, a desilting tool and a sludge processing and collecting device; the desilting operation device is specifically a multifunctional mechanical arm, and the multifunctional mechanical arm can operate the desilting tool to remove sludge and debris in the flow channel; the sludge processing and collecting device is used to collect and process the removed sludge and debris; The security monitoring system includes a real-time monitoring module and an alarm module; The real-time monitoring module is used to monitor the robot's operating status and working environment in real time through a variety of sensors. The alarm module is used to issue an alarm message when an abnormality is detected in the robot's operating status or working environment, reminding the staff to deal with potential problems in a timely manner.

2. A method for desilting a nuclear power plant flow channel combined with a robot using an image recognition algorithm, applied to a nuclear power plant flow channel combined with a robot using an image recognition algorithm as claimed in claim 1, characterized in that: The method comprises the following steps: S11: Use laser radar technology to locate the robot in the flow channel in real time, and plan the optimal movement path based on the positioning results; S12: transmitting the planned moving path instruction to the drive control module, and the drive control module controls the robot to move along the planned moving path according to the moving path instruction; S13: When the robot moves on the moving path, a high-resolution camera installed at the front end of the robot is used to collect images in the flow channel in real time; S14: collecting the acquired image data, and comparing and analyzing the image data with a preset silt and debris recognition model to identify the location of the silt and debris in the flow channel; S15: Execute the dredging process: start the dredging operation device in the dredging operation system, and use the multifunctional mechanical arm to manipulate the dredging tool to clean and collect the identified silt and debris; S16: After the dredging is completed, the image of the silt and debris position identified in step S14 is collected again, and the dredging effect is checked in combination with the silt and debris accumulation before the dredging. If the check fails, the dredging process of step S15 is repeated until the test is qualified; S17: After the dredging inspection is completed, steps S12 to S16 are repeated until the robot moves to the end of the planned moving path.

3. A method for desilting a nuclear power plant flow channel using an image recognition algorithm and a robot according to claim 2, characterized in that: In step S11, the robot movement path is planned by: S111: Scan the robot's surrounding environment data through the laser radar, and construct point cloud data in the flow channel according to the data collected by the laser radar; S112: constructing an initial map of the flow channel according to the point cloud data obtained in step S111, wherein the initial map is a basic representation of the environment in the flow channel; S113: performing path planning according to the path planning algorithm combined with the initial map, and generating a moving path instruction to transmit to the drive control module; S114: Setting a path update cycle, re-execute steps S111 to S113 to update the path planning at each path update cycle time point or after the dredging operation is completed in the subsequent steps, and generate a new moving path instruction to transmit to the drive control module.

4. The method for desilting a nuclear power plant flow channel using an image recognition algorithm and a robot according to claim 2, characterized in that: In step S14, the identification of sludge and debris in the flow channel is completed in the following manner: S141: Acquire an image in the flow channel, and perform a preprocessing operation on the acquired image; the preprocessing operation includes image denoising and image enhancement processing; S142: Compare the preprocessed image with a preset silt and debris recognition model to quickly and preliminarily detect areas where silt and debris exist; the silt and debris recognition model is specifically a model obtained by training with a large amount of labeled silt and debris image data; the specific process of the rapid preliminary detection meets the following requirements: The preprocessed image is divided into multiple detection areas, and for each detection area, the following is performed: Among them, F d is a feature vector set of a certain detection area, wherein the feature vectors in the feature vector set include shape feature vectors, texture feature vectors and color feature vectors; F m is the feature vector set in the sludge debris recognition model; Sim(F d ,F m ) is the feature matching degree between a certain detection area in the image and the silt debris recognition model, and a matching degree threshold is set. When the feature matching degree between a certain detection area and the silt debris recognition model is higher than the matching degree threshold, it is marked that the detection area has silt debris; S143: When an area with silt and debris is detected, the robot stops moving and adjusts the angle of the camera so that the camera can focus and obtain an image of the center position of the silt and debris area; the specific implementation process is as follows: S1431: Detect the edge of the silt and debris by using an edge detection algorithm, and control the position and direction of the camera to slowly turn to the edge of the silt and debris; calculate the rate of increase of edge pixels of the image newly entering the camera range during the camera rotation process, and when the rate of increase of edge pixels is lower than the set rate of increase threshold, stop the rotation of the camera; S1432: re-detecting the silt and debris area within the camera's field of view, calculating the centroid position of the silt and debris area, and aligning the camera center with the centroid position to obtain a center position image of the silt and debris area; S144: Calculate the edge and shape features of the silt and debris area based on the center position image of the silt and debris area, so as to guide the subsequent dredging process.

5. The method for desilting a nuclear power plant flow channel using an image recognition algorithm and a robot according to claim 2, characterized in that: In the step S16, the desilting effect is checked by: K=α1·M h +α2·L h ; Where K is the dredging inspection coefficient, α1 and α2 are preset weight coefficients, satisfying α1+α2=1; M h is the area of ​​silt and debris in the silt and debris position image after dredging, L h is the perimeter of the silt and debris area in the silt and debris position image after dredging; By comparing the silt removal inspection coefficient with the set silt removal inspection threshold, when the silt removal inspection coefficient is less than the set silt removal inspection threshold, it is determined that the silt removal effect is good, otherwise the silt removal operation in step S15 is repeated.

Images