Intelligent cleaning system and method for floating objects in inspection well

Abstract

The invention relates to the technical field of municipal pipeline cleaning, and discloses an inspection well floating object intelligent cleaning system and method.The cleaning system comprises a sensing and recognizing unit, a control unit, an agent adding unit and a mechanical cleaning unit, and the sensing and recognizing unit is arranged in an inspection well; the image acquisition module is used for acquiring water surface images and identifying distribution areas of solid floaters, oil stains and adhesion mixtures thereof; the control unit generates a time sequence operation instruction that the medicament adding unit acts prior to the mechanical cleaning unit according to the recognition result; the agent adding unit is used for adding a chemical treatment agent into the oil stain and adhesion mixture area; and the mechanical cleaning unit grabs and removes the solid floating objects and the treated residues. According to the invention, a sequential operation mode of first chemical decomposition and then physical grabbing is adopted, so that the problems of oil stain diffusion and incomplete cleaning of adhesive substances are solved, and efficient, complete and automatic cleaning of solid-oil mixed pollutants in the inspection well is realized.

Description

Technical Field

[0001] This invention relates to the field of municipal pipeline cleaning technology, specifically to an intelligent cleaning system and method for floating debris in inspection wells. Background Technology

[0002] Municipal drainage inspection wells are important nodes in urban underground pipe networks, and their internal condition affects drainage efficiency and flood control safety. During operation and maintenance, solid waste such as plastic bags and beverage bottles often float on the water surface inside the wells, and these solids adhere to the viscous oil sludge from catering waste, forming a dense heterogeneous mixture. These pollutants severely obstruct visibility, making it impossible for maintenance personnel to observe the condition of the pipe openings and the water level, necessitating regular cleaning.

[0003] Currently, cleanup work largely relies on manual labor using tools such as nets and rakes to retrieve debris from wells or at the wellhead. This method is inefficient, labor-intensive, and carries safety risks such as poisoning from toxic gases and falls. Furthermore, it is ineffective at removing adhesive oil slicks. While mechanical cleaning equipment can be used for retrieval, its functionality is limited, typically targeting only discrete solid waste. It cannot identify or handle oil slicks and is ill-suited to dealing with complex situations involving adhered solid oil contaminants. Moreover, the mechanical actions during the cleaning phase can easily cause oil to spread, resulting in secondary pollution. Summary of the Invention

[0004] In view of this, the present invention provides an intelligent cleaning system and method for floating debris in inspection wells to solve the problems mentioned in the background art.

[0005] In a first aspect, the present invention provides an intelligent cleaning system for floating debris in inspection wells, comprising: A sensing and identification unit is suitable for installation inside an inspection well; the sensing and identification unit is used to collect image information of the water surface inside the well to identify the distribution areas of solid floating objects, oil stains, and mixtures of the two adhering to each other; The control unit is communicatively connected to the sensing and recognition unit; The agent dosing unit is communicatively connected to the control unit; the agent dosing unit is used to add chemical treatment agents to the oily area or the distribution area of ​​the mixture of solid floating matter and oil. A mechanical cleaning unit is communicatively connected to the control unit; the mechanical cleaning unit is used to grasp and remove the solid floating objects and the treated adhesive mixture. The control unit is configured to generate timing operation instructions based on the distribution area, prioritizing the operation of the chemical dosing unit over the mechanical cleaning unit.

[0006] Beneficial effects: This application provides a technical solution for automatically treating heterogeneous solid-oil mixtures of contaminants in inspection wells; the system identifies the contamination distribution area through a sensing and identification unit, and then generates a time-series operation command based on the contamination distribution area through a control unit, and ensures that the chemical agent is added before the mechanical cleaning action, so that the system operates by first chemically decomposing the oil and then physically grabbing the solids; it effectively solves the problem of oil diffusion and secondary pollution caused by the direct action of existing mechanical cleaning methods on the mixture, and overcomes the problem of solid waste being difficult to remove due to oil adhesion, thus achieving efficient and complete cleaning of heterogeneous mixtures.

[0007] In some embodiments, the sensing and recognition unit includes: The main camera is configured to capture visible light images to identify solid floating objects; An auxiliary camera is configured to collect the spectrum of oil stains in order to identify them.

[0008] Beneficial effects: The sensing and recognition unit is equipped with a main camera and an auxiliary camera that respectively collect visible light images and oil stain characteristic spectra, enabling the system to simultaneously acquire morphological information of solid waste and distribution information of oil stains; this camera sensing method can effectively identify transparent or dark oil stains in dim and reflective environments inside the well, thereby providing an accurate and multi-dimensional data foundation for the generation of subsequent zoning and time-series operation instructions.

[0009] In some embodiments, the control unit includes: The scene understanding module is used to divide the water surface in the well into a solid waste area, an oil-covered area, and an adhesive mixture area based on the image information collected by the perception and recognition unit. The decision-making module is used to plan the sequential operation instructions based on the area division of the scenario understanding module, which first chemically treats the oil-covered area and the area of ​​adhering mixture through the agent dosing unit, and then cleans all solid floating objects through the mechanical cleaning unit.

[0010] Beneficial effects: This solution uses a scene understanding module to perform semantic segmentation on image information, forming a structured map of solid waste areas, oil-covered areas, and areas with adhering mixtures. The decision-making module then plans the optimal operation sequence for different areas based on this map. This upgrades the system's response from simple trigger execution to strategy-based operation based on environmental semantic understanding, enabling precise and efficient configuration of agents and mechanical actions during the cleaning phase.

[0011] In some embodiments, the drug dosing unit includes: A pharmaceutical storage compartment is used to store chemical treatment agents; A quantitative dosing device, connected to the reagent tank, is used to deliver a quantitative amount of chemical treatment agent to the area covered by oil and / or the area of ​​the adhesive mixture, according to the timed operation instructions.

[0012] Beneficial effects: The combination of the reagent tank and the quantitative dosing device enables the system to accurately deliver a fixed amount of chemical treatment agent to the designated area according to instructions. On the one hand, this ensures that the agent can directly act on the target contaminated area, improving reaction efficiency and avoiding waste; on the other hand, quantitative control avoids secondary environmental problems that may be caused by excessive use of the agent, making the chemical treatment process more environmentally friendly and controllable.

[0013] In some embodiments, the chemical treatment agent is a demulsifier or a biological enzyme preparation, and the quantitative dosing device is a catapult or a metering pump.

[0014] Beneficial effects: The chemical treatment agent is a demulsifier or a biological enzyme preparation, providing a chemical treatment method that can effectively decompose catering oil stains; the quantitative dosing device is a catapult device or a quantitative pump, which can be adapted to accurately deliver solid particles or liquid agents to the target area.

[0015] In some embodiments, the mechanical cleaning unit includes: The sliding rod is installed vertically inside the inspection well; A motion module is mounted on the slide bar; the motion module is adapted to move up and down on the slide bar, and the motion module is provided with a gripping device.

[0016] Beneficial effects: The mechanical cleaning unit adopts a vertically installed slide bar and a motion module that can be raised and lowered along the slide bar; the mechanical cleaning unit can utilize a small space to achieve a wide range of stable vertical movement of the motion module, providing favorable mechanical support for the grabbing device to reach any position on the water surface for operation.

[0017] In some embodiments, the gripping device is a net bag, which is rotatably mounted on the motion module.

[0018] Beneficial effects: The gripping device is designed as a rotatable net bag; the flexibility of the net bag can adapt to solid waste of different shapes and sizes, reducing the difficulty of gripping; the rotatable design of the net bag increases the flexibility of the gripping angle, which can effectively scoop up or wrap the target object from the water, thereby improving the success rate and capacity of a single gripping and improving the efficiency of mechanical cleaning.

[0019] In some embodiments, the control unit further includes a safety interlock module, configured to: During the operation phase of the agent dosing unit, the mechanical cleaning unit is locked; and / or; Beneficial effect: When the water level is detected to exceed the safety threshold, the mechanical cleaning unit is controlled to be raised to a safe height.

[0020] The safety interlock module is configured to lock the mechanical cleaning unit during the chemical dosing stage to ensure that the chemical treatment process is not disturbed by mechanical means and to ensure the treatment effect. The safety interlock module is also configured to forcibly raise the mechanical cleaning unit to a safe height when the water level exceeds the limit. This setting provides active protection against sudden water level rises, avoids damage to the mechanical cleaning unit, and improves the reliability of the cleaning system in long-term unattended operation.

[0021] Secondly, the present invention also provides a cleaning method for an intelligent cleaning system for floating debris in inspection wells, comprising: Identification steps: The sensing and identification unit collects image information of the water surface inside the well to identify the distribution areas of solid floating objects, oil slicks, and the mixture of the two adhering to each other; Cleaning steps: Based on the distribution area, generate a timed operation instruction that prioritizes the action of the chemical dosing unit over the mechanical cleaning unit; according to the timed operation instruction, first control the chemical dosing unit to add chemical treatment agent to the oil-covered area and / or the area with adhering mixture, and after a preset reaction time, control the mechanical cleaning unit to grab and remove the solid floating objects; After the cleaning process is completed, images are captured by the sensing and recognition unit to assess the status of residual pollutants; If the assessment results do not meet the preset standards, then the cleanup steps will be performed.

[0022] Beneficial effects: This cleaning method provides an automated cleaning process that can be updated, verified, and iteratively optimized. Through a closed-loop process of identification, execution, evaluation, and re-execution, it achieves initial automated cleaning. After each cleaning, it also introduces an effect evaluation step based on perception and identification, enabling the system to automatically determine whether the cleaning is thorough and to autonomously start a new round of targeted cleaning if the target is not met, thus ensuring the final quality of the cleaning task and realizing the transition from automated execution to intelligent processing.

[0023] In some embodiments, the cleaning method further includes: Upload the process data, images, and results of the identification and cleaning steps to the cloud platform.

[0024] Beneficial effects: By adding the step of uploading process data to the cloud platform, the operation process can be digitally recorded and remotely visualized, which facilitates remote monitoring and historical review by operation and maintenance personnel. It can provide a data foundation for subsequent pipeline status analysis, pollution source tracing, and preventive maintenance, thereby extending the value of a single cleaning operation to the operation and maintenance management of the pipeline system. Attached Figure Description

[0025] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the intelligent cleaning system for floating debris in inspection wells according to an embodiment of the present invention; Figure 2 This is an assembly diagram of the control unit, the agent dosing unit, and the sensing and identification unit in the intelligent cleaning system for floating debris in inspection wells according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the mechanical cleaning unit in the intelligent cleaning system for floating debris in inspection wells according to an embodiment of the present invention; Figure 4 This is a control diagram of the intelligent cleaning system for floating debris in inspection wells according to an embodiment of the present invention; Figure 5 This is a flowchart of the decision-making algorithm for the intelligent cleaning system for floating debris in inspection wells, according to an embodiment of the present invention.

[0027] Explanation of reference numerals in the attached figures: 1. Manhole cover; 2. Control unit; 3. Chemical dosing unit; 4. Sensing and identification unit; 5. Sliding rod; 6. Motion module; 7. Actuating rotary joint; 8. Net bag; 9. Chemical tank; 10. Ejector device. Detailed Implementation

[0028] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0029] The following is combined Figures 1 to 4 The following describes embodiments of the present invention.

[0030] According to an embodiment of the present invention, in one aspect, an intelligent cleaning system for floating debris in inspection wells is provided. The cleaning system includes a sensing and identification unit 4, a control unit 2, a chemical dosing unit 3, and a mechanical cleaning unit. The control unit 2 is communicatively connected to the sensing and identification unit 4, the chemical dosing unit 3 is communicatively connected to the control unit 2, and the mechanical cleaning unit is communicatively connected to the control unit 2. The sensing and identification unit 4 is adapted to be installed inside the inspection well; the sensing and identification unit 4 is used to collect image information of the water surface inside the well to identify the distribution areas of solid floating debris, oil stains, and mixtures of the two. The sensing and identification unit 4 and the chemical dosing unit 3 are installed in adjacent areas, and the mechanical cleaning unit is spaced apart from the chemical dosing unit 3. A well cover 1 is provided on the inspection well, and the well cover 1 can serve as a connection base for assembling the mechanical cleaning unit.

[0031] In this embodiment, the chemical dosing unit 3 is used to add chemical treatment agents to the oily area or the distribution area of ​​the mixture of solid floating matter and oil. The mechanical cleaning unit is used to grab and remove the solid floating matter and the treated adhesive mixture. The control unit 2 is configured to generate timing operation instructions that prioritize the operation of the chemical dosing unit 3 over the mechanical cleaning unit, based on the distribution area.

[0032] In a specific embodiment, the perception and recognition unit 4 includes a main camera (not shown in the figure) and an auxiliary camera (not shown in the figure), which are integrated together. The main camera is configured as a high-resolution RGB camera, which acquires visible light images and identifies solid floating objects by their shape and texture features. The auxiliary camera is configured as a specific-band infrared (NIR) camera, which acquires the spectrum of oil slicks and identifies oil slicks by their significant absorption characteristics in the near-infrared band (e.g., 900-1000 nm). In this embodiment, the perception and recognition unit 4 is equipped with a main camera and an auxiliary camera that respectively acquire visible light images and oil slick characteristic spectra, enabling the system to simultaneously acquire morphological information of solid waste and distribution information of oil slicks. This dual-camera perception method can effectively identify transparent or dark oil slicks in dim or reflective environments inside the well, thereby providing a precise and multi-dimensional data foundation for the subsequent generation of zoning and time-series operation instructions.

[0033] For the adhesive mixture of solid floating objects and oil, the system makes a joint judgment based on the recognition results of the main camera and the auxiliary camera. Specifically, areas that show solid outlines in the visible light image but simultaneously show strong oil signals in the corresponding oil spectral image, and where the edge regions of the two signals highly overlap or intertwine, are identified as adhesive mixture areas.

[0034] In a specific embodiment, the control unit 2 includes a scene understanding module and a decision-making module. The scene understanding module is used to divide the water surface in the well into a solid waste area, an oil-covered area, and an adhesive mixture area based on the image information collected by the perception and recognition unit 4, and can generate a structured scene map including the boundary coordinates, area, and pollutant type of each area. The decision-making module is used to plan a time-series operation instruction based on the area division by the scene understanding module. First, the oil-covered area and the adhesive mixture area are chemically treated by the chemical dosing unit 3. After the chemical treatment agent reaches the preset reaction time, the mechanical cleaning unit will clean up all solid floating objects and solid residues detached due to oil decomposition. The time-series operation instruction includes at least the start sequence, target coordinates, and dosage of the chemical dosing unit 3, as well as the subsequent grabbing path and sequence of the mechanical cleaning unit.

[0035] The specific recognition algorithm process is as follows: I. Image Acquisition and Preprocessing: Dual cameras are triggered synchronously to acquire registered RGB and NIR images. White balance correction and contrast enhancement are performed on the RGB images; dark current correction and noise filtering are performed on the NIR images.

[0036] 2. Water Surface Region Extraction: By detecting image edges, the region of interest (ROI) of the water surface in the image is determined, eliminating interference from the well wall.

[0037] III. Solid Waste Detection: A lightweight convolutional neural network (such as MobileNetV3-SSD) is used as the object detection model. The ROI regions of the RGB image are input into the model, and a series of bounding boxes are output. And its category confidence. Boundary box regions with a confidence score greater than the threshold of 0.7 are marked as solid waste candidate regions.

[0038] IV. Oil Smudge Region Segmentation: Calculate the Normalized Oil Index (NDOI) for each pixel in the image: .in It is the intensity of the waveband. It is the intensity of the red channel in an RGB image. A small constant is used to prevent zero-rounding. The NDOI value of oily areas is significantly higher than that of clean water. Adaptive thresholding (such as the Otsu algorithm) is used on the NDOI image to obtain a binarized mask. The white pixels (with a value of 1) represent areas suspected of being oily. Perform geometry opening operations to remove small noise points and form connected oil stain candidate regions.

[0039] V. Determination of Adhesive Mixtures: Input the set of candidate bounding boxes for solid waste { }, oil stain candidate mask { For each bounding box Calculate the region The proportion of white pixels .like If the solid waste is completely covered by oil, this area is marked as an area of ​​adhering mixture. If so, it is determined to be partial adhesion and is also marked as an adhesion mixture area. If If so, it is marked as a pure solid waste area. China has not been subject to any Coverage is analyzed by performing connectivity region analysis, and each connectivity region is marked as an oil spill coverage area.

[0040] This solution uses a scene understanding module to perform semantic segmentation on image information, forming a structured map of solid waste areas, oil-covered areas, and areas with adhering mixtures. The decision-making module then plans the optimal operation sequence for different areas based on this map. This upgrades the system's response from simple trigger execution to strategy-based operation based on environmental semantic understanding, enabling precise and efficient configuration of agents and mechanical actions during the cleaning phase.

[0041] In a specific embodiment, the chemical dosing unit 3 includes a chemical tank 9 and a quantitative dosing device. The chemical tank 9 is used to store chemical treatment agents. The quantitative dosing device is connected to the chemical tank 9 and is used to deliver a quantitative amount of chemical treatment agent to the oil-covered area and / or the area of ​​the adhering mixture area according to the time-series operation instructions. The quantitative dosing device can deliver a quantitative amount of chemical treatment agent to the area covered by oil and / or the area of ​​the adhering mixture through the built-in control valve or meter.

[0042] Figure 5 The flowchart illustrates the decision-making algorithm. Through the combination of the reagent tank 9 and the quantitative dosing device, the system can accurately deliver a fixed amount of chemical treatment agent to the designated area based on movement and dosing instructions. This ensures, on the one hand, that the agent directly acts on the target contaminated area, improving reaction efficiency and avoiding waste; on the other hand, quantitative control avoids secondary environmental problems that may result from excessive use of the agent, making the chemical treatment process more environmentally friendly and controllable.

[0043] In a specific embodiment, the chemical treatment agent is a demulsifier or a biological enzyme preparation, and the metering device is a catapult 10 or a metering pump. The chemical treatment agent, being a demulsifier or a biological enzyme preparation, provides a chemical treatment method that can effectively decompose catering grease; the metering device, being a catapult 10 or a metering pump, can be adapted to accurately deliver solid particles or liquid agents to the target area.

[0044] When using the ejection device 10, the ejection distance and landing point of the solid agent are controlled by adjusting the compressed air pressure or the spring force.

[0045] When using a metering pump, the preset volume of liquid medicine is delivered by adjusting and controlling the pump speed and duration.

[0046] Specifically, for animal and vegetable oils and kitchen waste, biological enzyme preparations can be added. Their main components are lipase, protease, surfactant, pH buffer, and carrier. The mechanism of action is that lipase hydrolyzes large oil molecules into hydrophilic glycerol and fatty acids at room temperature, while protease decomposes adhering organic residues. During operation, apply at 0.2-0.3 m... 2 For oily contaminants, add one water-soluble capsule containing 10 mL of the undiluted solution. For mineral oils such as lubricating oils and light fuel oils, a demulsifier can be added. Its main components are cationic surfactants such as polyacrylamide derivatives. Its mechanism of action is through charge neutralization and bridging, destabilizing dispersed oil droplets and suspended solids, causing them to aggregate into larger flocs that float or sink, thus achieving oil-water separation. During operation, add one water-soluble capsule every 0.2-0.25 mL of the solution. 2 Add one water-soluble capsule to the oil stains. The capsule contains 5 mL of the original solution.

[0047] In a specific embodiment, the mechanical cleaning unit can be configured as a robotic arm, specifically including a slide bar 5, a motion module 6, and a gripping device. The slide bar 5 is vertically installed inside the inspection well, and the motion module 6 is mounted on the slide bar 5. The motion module 6 is adapted to move up and down on the slide bar 5, and the gripping device is mounted on the motion module 6. In this solution, the mechanical cleaning unit uses a vertically installed slide bar 5 and a motion module 6 that can move up and down along the slide bar 5. The mechanical cleaning unit can utilize a small space to achieve a large range of stable vertical movement of the motion module 6, providing favorable mechanical support for the gripping device to reach any position on the water surface for operation.

[0048] In a specific embodiment, the gripping device is a net bag 8, which is rotatably mounted on the motion module 6. The flexible nature of the net bag allows it to adapt to solid waste of different shapes and sizes, reducing the difficulty of gripping. The rotatable design of the net bag 8 increases the flexibility of the gripping angle, enabling it to effectively scoop up or wrap the target object from the water, thereby improving the success rate and capacity of a single gripping operation and enhancing mechanical cleaning efficiency.

[0049] In the specific implementation process, the net bag component 8 rotates around a horizontal axis, with a rotation angle of 45-180°. The net bag component 8 is rotatably mounted on the motion module 6 via a motor-driven rotary joint 7. Figure 1 For example, the slide bar 5 is set on the inner wall side of the inspection well. During the retrieval and cleaning, the motion module 6 drives the net bag 8 downward toward the water surface. After reaching the designated position, the motion module 6 drives the net bag 8 to rotate. During the retrieval process, the rotation angle of the net bag 8 is 90 degrees.

[0050] The cleaning system provided in this embodiment operates in the following mode: In Mode 1, solid waste is the primary factor, and the mechanical cleaning unit is controlled to move. The identified solid waste is precisely retrieved through the net bag component 8. The retrieved waste is then raised to the highest point of the slide bar 5 to await manual removal and replacement of the net bag.

[0051] Mode 2, dominated by oil pollution, controls the agent addition unit 3, which adds targeted water-soluble packaged biological enzyme preparations, such as lipase and protease, according to the area covered by the oil pollution; or adds demulsifiers, such as cationic surfactants.

[0052] Mode 3, for mixed pollution, prioritize the addition of demulsifiers to decompose or coagulate the oil; after it is separated or combined with solid waste, mechanical cleaning units are then activated for unified dredging.

[0053] The cleaning system provided in this embodiment can automatically handle heterogeneous solid-oil mixtures of contaminants in inspection wells. The system identifies the contamination distribution area through the sensing and identification unit 4, and then generates a time-series operation command based on the contamination distribution area through the control unit 2. The chemical agent is added before the mechanical cleaning action, so that the system operates in a sequence of first chemically decomposing the oil and then physically grabbing the solids. It effectively solves the problems of oil diffusion and secondary pollution caused by the direct action of existing mechanical cleaning methods on the mixture, and overcomes the problem of solid waste that is difficult to remove due to oil adhesion, thus achieving efficient and complete cleaning of heterogeneous mixtures.

[0054] The cleaning system provided in this embodiment combines mechanical retrieval with chemical treatment, effectively overcoming the limitations of a single technical approach and capable of handling complex contamination conditions within inspection wells. The mechanical cleaning unit enables precise positioning and retrieval, avoiding blind operations; the quantitative addition of chemicals ensures effectiveness while preventing waste and secondary pollution. The cleaning system enables automated operation, eliminating the need for maintenance personnel to enter the well or operate at the wellhead for extended periods, reducing safety risks and labor intensity. The cleaning system requires only the installation of the sensing and identification unit 4 and the mechanical cleaning unit within the inspection well; no other complex construction or excavation is required.

[0055] In a specific embodiment, the control unit 2 also includes a safety interlock module, which is configured to lock the mechanical cleaning unit during the operation phase of the agent dosing unit 3 to ensure that the chemical treatment process is not disturbed by mechanical means and to ensure the treatment effect.

[0056] In a specific embodiment, the safety interlock module is configured to control the mechanical cleaning unit to rise to a safe height when the water level exceeds a safety threshold. This configuration provides proactive protection against sudden water level rises, preventing damage to the mechanical cleaning unit and improving the reliability of the cleaning system during long-term unattended operation. The safety interlock module includes a water level sensor to detect the water level in the inspection well; when the water level exceeds the safety threshold, it controls the mechanical cleaning unit to rise to a safe height and suspends operation.

[0057] According to an embodiment of the present invention, in another aspect, a cleaning method for an intelligent cleaning system for floating debris in inspection wells is also provided, comprising: Identification steps: The image information of the water surface in the well is collected by the sensing and identification unit 4 to identify the distribution area of ​​solid floating objects, oil sludge, and the mixture of the two adhering to each other; Cleaning steps: Based on the distribution area, generate a timed operation instruction that prioritizes the action of the chemical dosing unit 3 over the mechanical cleaning unit; according to the timed operation instruction, first control the chemical dosing unit 3 to add chemical treatment agent to the oil-covered area and / or the area with adhering mixture, and after a preset reaction time, such as 10-30 minutes, control the mechanical cleaning unit to grab and remove the solid floating objects; After the cleaning step is completed, images are collected by the sensing and recognition unit 4. Specifically, the residual pollutant status is assessed by comparing the differences between the images before and after cleaning; alternatively, the characteristics of the residual pollutants can be analyzed. If the assessment results do not meet the preset standards, the cleaning step is then performed.

[0058] This cleanup method provides an automated cleanup process that can be updated, verified, and iteratively optimized. Through a closed-loop process of identification, execution, evaluation, and re-execution, it achieves initial automated cleanup. After each cleanup, it also introduces an effect evaluation step based on perception and identification, enabling the system to automatically determine whether the cleanup is thorough and to autonomously start a new round of targeted cleanup if the target is not met, thus ensuring the final quality of the cleanup task and realizing the transition from automated execution to intelligent processing.

[0059] In a specific embodiment, the cleaning method further includes uploading the process data, images, and results of the identification and cleaning steps to a cloud platform.

[0060] By adding the step of uploading process data to the cloud platform, the operation process can be digitally recorded and remotely visualized, which facilitates remote monitoring and historical review by operation and maintenance personnel. This provides a data foundation for subsequent pipeline status analysis, pollution source tracing, and preventive maintenance, thereby extending the value of a single cleaning operation to the operation and maintenance management of the pipeline system.

[0061] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. An intelligent cleaning system for floating debris in inspection wells, characterized in that, include: The sensing and identification unit (4) is suitable for being installed inside the inspection well; the sensing and identification unit (4) is used to collect image information of the water surface inside the well in order to identify the distribution area of ​​solid floating objects, oil stains and the mixture of the two adhering together. The control unit (2) is communicatively connected to the sensing and identification unit (4); The agent dosing unit (3) is communicatively connected to the control unit (2); the agent dosing unit (3) is used to add chemical treatment agents to the oily area or the distribution area of ​​the mixture of solid floating matter and oily contaminants; The mechanical cleaning unit is communicatively connected to the control unit (2); the mechanical cleaning unit is used to grab and remove the solid floating objects and the treated adhesive mixture. The control unit (2) is configured to generate a timing operation instruction that controls the drug dosing unit (3) to take precedence over the mechanical cleaning unit based on the distribution area.

2. The cleaning system according to claim 1, characterized in that, The sensing and recognition unit (4) includes: The main camera is configured to capture visible light images to identify solid floating objects; An auxiliary camera is configured to collect the spectrum of oil stains in order to identify them.

3. The cleaning system according to claim 2, characterized in that, The control unit (2) includes: The scene understanding module is used to divide the water surface in the well into a solid waste area, an oil-covered area and an adhesive mixture area based on the image information collected by the perception and recognition unit (4); The decision module is used to plan the sequential operation instructions based on the division of the scene understanding module, first chemically treating the oil-covered area and the adhesive mixture area through the agent dosing unit (3), and then cleaning all solid floating objects through the mechanical cleaning unit.

4. The cleaning system according to claim 2, characterized in that, The drug dosing unit (3) includes: The pharmaceutical storage compartment (9) is used to store chemical treatment agents; A quantitative dosing device, connected to the reagent tank (9), is used to deliver a quantitative amount of chemical treatment agent to the area of ​​the oil-covered area and / or the area of ​​the adhesive mixture according to the timed operation instructions.

5. The cleaning system according to claim 4, characterized in that, The chemical treatment agent is a demulsifier or a biological enzyme preparation, and the quantitative dosing device is a catapult (10) or a quantitative pump.

6. The cleaning system according to claim 1, characterized in that, The mechanical cleaning unit includes: The slide bar (5) is installed vertically inside the inspection well; A motion module (6) is installed on the slide bar (5); the motion module (6) is adapted to move up and down on the slide bar (5), and the motion module (6) is provided with a gripping device.

7. The cleaning system according to claim 6, characterized in that, The gripping device is a net bag (8), which is rotatably mounted on the motion module (6).

8. The cleaning system according to claim 1, characterized in that, The control unit (2) also includes a safety interlock module, configured as follows: During the operation phase of the agent dosing unit (3), the mechanical cleaning unit is locked; and / or; When the water level is detected to exceed the safety threshold, the mechanical cleaning unit is controlled to be raised to a safe height.

9. A cleaning method based on the intelligent cleaning system for floating debris in inspection wells according to any one of claims 1 to 8, characterized in that, include: Identification steps: The image information of the water surface in the well is collected by the sensing and identification unit (4) to identify the distribution area of ​​solid floating objects, oil stains and the mixture of the two adhering together; Cleaning steps: Based on the distribution area, generate a time-series operation instruction that prioritizes the action of the chemical dosing unit (3) over the mechanical cleaning unit; according to the time-series operation instruction, first control the chemical dosing unit (3) to add chemical treatment agent to the oil-covered area and / or the area of ​​adhering mixture, and after a preset reaction time, control the mechanical cleaning unit to grab and remove the solid floating objects; After the cleaning process is completed, images are collected by the sensing and recognition unit (4) to assess the status of residual pollutants; If the assessment results do not meet the preset standards, then the cleanup steps will be performed.

10. The cleaning method according to claim 9, characterized in that, Also includes: Upload the process data, images, and results of the identification and cleaning steps to the cloud platform.

Images