An automatic floating dross cleaning system for hyperspectral water quality monitoring and a control method thereof

Abstract

This application discloses an automatic scum removal system and its control method for hyperspectral water quality monitoring, belonging to the field of wastewater treatment. The system includes a hyperspectral monitoring module, a scum identification and interference judgment module, a control system module, and an automatic scum removal execution module. The hyperspectral monitoring module continuously acquires water surface spectral signals to obtain water quality parameters; the scum identification and interference judgment module determines the interference level; the control system module generates control commands based on the interference level, automatically triggering the scum removal program in cases of severe interference; the automatic scum removal execution module completes scum removal through water flow disturbance, collection, and discharge, with secondary imaging verification performed by a cleaning effect verification unit. This application achieves fully automated management from scum identification to cleaning completion verification, with fast response speed, high cleaning timeliness, and ensures the continuity and integrity of monitoring data. It is applicable to horizontal flow sedimentation tanks and other water treatment structures with surface scum problems.

Description

Technical Field

[0001] This invention relates to the field of wastewater treatment technology, and in particular to an automatic scum removal system and its control method for hyperspectral water quality monitoring. Background Technology

[0002] Water environment monitoring is a crucial step in wastewater treatment processes. The accuracy and reliability of monitoring data directly affect the stable attainment of effluent quality standards and the optimization of process parameters. In the advanced treatment stage of wastewater treatment plants, horizontal flow sedimentation tanks, as the core structure for solid-liquid separation, play a vital role in removing suspended solids and reducing turbidity. The quality of their effluent directly impacts the treatment effect of subsequent disinfection processes and the final discharge standards. In recent years, with the rapid development of optical sensing technology and spectral analysis algorithms, hyperspectral water quality monitoring technology has been increasingly widely used in wastewater treatment plants due to its advantages in simultaneously acquiring multi-dimensional spectral characteristics of water bodies and achieving synchronous inversion of multiple parameters. This technology, by collecting water surface spectral data and combining it with preset regression models or machine learning algorithms, can obtain key water quality parameters such as chemical oxygen demand (COD), total nitrogen (TNO), total phosphorus (TP), and suspended solids concentration in real time, providing strong data support for process operation and control.

[0003] However, in practical engineering applications, the surface conditions of horizontal flow sedimentation tanks are often quite complex. Algae growth, oil accumulation, and flocculent suspended matter generated by water flow disturbance easily accumulate on the surface under the influence of wind and water current, forming a scum layer. This scum buildup not only disrupts the light reflection balance of the water surface but also creates a blocking effect within the field of view of the hyperspectral imager. When the scum enters the spectral acquisition area, its own spectral response characteristics superimpose and confuse with the water spectral signal, leading to a significant decrease in the signal-to-noise ratio of the spectral data, thus affecting the accuracy of water quality parameter inversion. Especially under conditions of high scum density, the water quality parameters acquired by the hyperspectral imager may show significant deviations, or even lead to misjudgments. This is an unacceptable technical risk for operation and management personnel who need to make process decisions based on monitoring data.

[0004] Regarding the treatment of scum on the surface of sedimentation tanks, the industry currently generally adopts a cleaning mode that combines manual inspection and manual dredging. Maintenance personnel, according to a predetermined inspection cycle or based on feedback of abnormal effluent quality, carry specialized dredging tools to the sedimentation tank site to manually dredge and collect the accumulated scum. This passive cleaning method has several inherent drawbacks: First, manual inspection is limited by personnel allocation and inspection frequency, making it difficult to achieve continuous real-time monitoring of the tank surface conditions. There is a significant time lag between the generation and discovery of scum, during which time the hyperspectral imager has been continuously interfered with, compromising the historical integrity and continuity of monitoring data. Second, the generation of scum is random and sudden, especially after rain, during algal blooms, or when there are large fluctuations in influent water quality, the amount of scum generated may increase dramatically, and the response speed and processing capacity of manual cleaning cannot meet the requirements for timely removal. Third, manual cleaning operations are limited by weather conditions and safety factors; in severe weather or when working space is limited at the edge of the tank, the timeliness and thoroughness of the cleaning work cannot be guaranteed.

[0005] Further analysis reveals a deeper contradiction in the existing technical solutions: hyperspectral water quality monitoring technology is essentially a high-precision, high-sensitivity optical detection method, and its technical characteristics dictate that it has almost stringent requirements for the stability and cleanliness of the monitoring environment. Only when there is no obstruction or interference within the field of view and the water surface light reflection conditions are consistent can the hyperspectral analyzer fully utilize its multi-parameter synchronous detection advantages to output reliable water quality data. However, as an open structure in wastewater treatment processes, the surface environment of a horizontal flow sedimentation tank inevitably produces scum. This fundamental contradiction between the high-precision monitoring requirement and the complex operating environment cannot be fundamentally resolved simply by increasing the frequency of manual inspections or adopting a passive, reactive cleaning strategy. Manual cleaning is essentially a post-hoc remedial mechanism; it cannot eliminate the interference and data distortion already caused by scum in the monitoring process.

[0006] Therefore, how to establish a proactive intervention and real-time response technical mechanism between the accuracy requirements of hyperspectral water quality monitoring and the complex working environment of advection sedimentation tanks, so as to achieve timely identification, rapid treatment and effect verification of scum interference, and thus ensure that the hyperspectral instrument always works in an effective field of view, has become a key challenge and an urgent technical problem for those skilled in the art. Summary of the Invention

[0007] The purpose of this invention is to provide an automatic slag removal system and its control method for hyperspectral water quality monitoring, so as to solve the technical problems in the prior art where floating slag on the surface of the water in the horizontal sedimentation tank interferes with the accuracy of hyperspectral water quality monitoring and manual cleaning is slow and untimely.

[0008] To achieve the above-mentioned objectives, the present invention adopts the following technical solution: An automatic sludge removal system for hyperspectral water quality monitoring includes a hyperspectral monitoring module, a scum identification and interference judgment module, a control system module, and an automatic sludge removal execution module. The hyperspectral monitoring module is fixedly installed above a horizontal sedimentation tank and is used to continuously collect water surface spectral signals and acquire water quality parameters. The scum identification and interference judgment module is located inside the control system module and is used to receive data transmitted by the hyperspectral monitoring module, execute scum identification and interference judgment algorithms, and output decision signals. The control system module uses a programmable logic controller as the core control unit and is electrically connected to the hyperspectral monitoring module, the scum identification and interference judgment module, and the automatic sludge removal execution module. The automatic sludge removal execution module is located inside the horizontal sedimentation tank and includes a moving track extending along the length of the tank, a sludge removal execution mechanism installed on the moving track, and a scum collection and discharge component connected to the sludge removal execution mechanism.

[0009] Furthermore, the hyperspectral monitoring module includes a hyperspectral water quality monitor, a monitor mounting bracket, and a first communication interface. The hyperspectral water quality monitor is fixedly installed above the horizontal sedimentation tank at a preset position via the monitor mounting bracket. The preset position is located in the middle of the tank's length and near the outlet. The first communication interface is located on the housing of the hyperspectral water quality monitor and is used to establish a data communication connection with the control system module. The hyperspectral water quality monitor operates in a wavelength range of 400nm to 1000nm, with a spectral resolution of no more than 5nm, and a continuously adjustable monitoring frequency range of 1 minute to 60 minutes. The monitor mounting bracket is made of stainless steel and includes a vertical column, a horizontal beam, and adjustable connectors for adjusting the installation angle and height of the hyperspectral water quality monitor.

[0010] Furthermore, the scum identification and interference judgment module includes a camera image acquisition and monitoring unit, a data storage unit, a scum identification algorithm unit, an interference judgment algorithm unit, and a decision logic unit. The camera image acquisition and monitoring unit uses an industrial digital camera to synchronously acquire images within the monitoring field of view of the hyperspectral water quality monitor. The data storage unit stores a scum sample image database and preset interference judgment threshold parameters. The scum sample image database includes multiple sets of scum image samples acquired under different working conditions, each set of image samples being labeled with corresponding scum type, density level, and distribution morphology characteristics. The scum types include algae aggregation type, oil sludge floating type, flocculent suspended matter accumulation type, and mixed type. The density level is divided into three levels: light, moderate, and heavy, according to the proportion of scum covering the area. The principle for setting the interference judgment threshold is as follows: when the proportion of scum covering the area is less than the first preset threshold, it is judged as an interference-free state; when the proportion of scum covering the area is greater than or equal to the first preset threshold and less than the second preset threshold, it is judged as a light interference state; when the proportion of scum covering the area is greater than or equal to the second preset threshold, it is judged as a heavy interference state.

[0011] Furthermore, the control system module includes a programmable logic controller (PLC) unit, a signal input module, a signal output module, a communication module, and a human-machine interface. The PLC unit integrates timing control logic blocks, action sequence control logic blocks, and safety protection logic blocks to manage the time coordination between modules, define the motion sequence of the automatic slag removal execution module, and monitor abnormal status signals during system operation. The human-machine interface is located on the control cabinet panel and is used to display system operating status parameters, alarm information, and manual operation control buttons.

[0012] Furthermore, the automatic slag removal execution module includes a moving track system, a slag removal execution mechanism, and a slag collection and discharge assembly; the moving track system includes two parallel track beams, a drive motor, a transmission device, and a position sensor; the slag removal execution mechanism includes an installation frame, a walking mechanism, a retrieval mechanism, and a water flow disturbance mechanism; the retrieval mechanism includes a retrieval arm, a drive cylinder, and a retrieval claw for grabbing and collecting slag; the water flow disturbance mechanism is used to disturb the water flow in the slag accumulation area before the retrieval operation, causing the dispersed slag to converge towards the retrieval area; the slag collection and discharge assembly includes a collection tank, a conveying pipe, a discharge pump, and a slag collection box for conveying the retrievald slag to the slag collection box.

[0013] Furthermore, the automatic slag removal execution module also includes a cleaning effect verification unit; the cleaning effect verification unit includes a second camera and image analysis software, which is used to perform secondary imaging acquisition on the original monitoring field of view after the slag removal is completed, and to judge the residual slag after cleaning through image processing analysis.

[0014] Furthermore, the present invention also includes a control method for an automatic sludge removal system for hyperspectral water quality monitoring, comprising the following steps: Step 1, the hyperspectral monitoring module continuously acquires spectral signals from the surface of the horizontal sedimentation tank, obtains real-time water quality parameter data, and transmits the data to the control system module; Step 2, the control system module simultaneously activates the camera image acquisition monitoring unit to acquire images of the monitoring field of view of the hyperspectral monitoring module; Step 3, the sludge identification algorithm unit preprocesses, extracts features, and performs pattern matching on the images acquired by the camera, identifies the presence of sludge in the images, and determines the type and distribution location of the sludge; Step 4, the interference judgment algorithm unit calculates the proportion of the sludge occlusion area within the monitoring field of view based on the sludge identification results, and compares it with a preset interference judgment threshold to determine the current interference level; Step 5, the decision logic unit generates corresponding control commands based on the interference level: when the interference level is no interference, the system maintains normal operation mode; when the interference level is mild interference, the system enters early warning mode, triggers an audible and visual alarm to alert maintenance personnel, and shortens the execution cycle of the monitoring loop; when the interference level is severe interference, the system immediately starts the sludge removal execution program.

[0015] Further, the slag removal execution procedure described in step five includes the following sub-steps: the slag removal execution mechanism starts from the initial position and moves along the moving track to the preset area to be cleaned; the water flow disturbance mechanism starts to disturb the water flow in the slag accumulation area; the drive cylinder of the retrieval mechanism pushes the retrieval arm to extend horizontally, and the retrieval claw enters the slag accumulation area to perform a grabbing and collection action; after the retrieval is completed, the retrieval mechanism resets, and the collection tank receives the slag released by the retrieval claw; the discharge pump starts to pump the slag mixture in the collection tank to the slag collection box through the conveying pipeline; the slag removal execution mechanism returns to the initial position along the moving track or moves to the next area to be cleaned, repeating the slag removal action until all areas to be cleaned are processed; the cleaning effect verification unit starts to perform secondary imaging acquisition on the original monitoring field of view and judge the cleaning effect.

[0016] Furthermore, the control method also includes a safety protection monitoring step: during the entire system operation, the safety protection logic block continuously monitors the operating status parameters of each module, including the current and temperature signals of the drive motor, the position signals of the position sensors, the pressure signals of the cylinders or hydraulic cylinders, and the communication status between each module; when any parameter is detected to exceed the preset safety range, the safety protection logic block immediately executes protection actions, including stopping the operation of the drive motor, locking the position of the retrieval mechanism, cutting off the power supply to the discharge pump, and triggering the corresponding alarm signal.

[0017] Furthermore, as a preferred embodiment of the present invention, the hyperspectral monitoring module first performs self-test and calibration operations after system startup, including light source intensity calibration, spectral response calibration, and camera parameter calibration. After the self-test and calibration are completed, the system enters normal operation mode and executes monitoring cycles according to the preset monitoring period. When the water quality of the influent to the horizontal sedimentation tank fluctuates or enters the algae bloom period, the control system module can automatically adjust the setting of the interference judgment threshold according to the upper computer monitoring system or historical data trend analysis, so that the system can respond more sensitively to scum interference. The scum collection box is equipped with a liquid level sensor and a scum quantity sensor to monitor the amount of scum collected and the sewage level in the scum collection box. When the amount of scum in the scum collection box reaches the preset full load threshold or the liquid level reaches the preset overflow threshold, the system automatically triggers an alarm signal.

[0018] Furthermore, the automatic sludge removal system and control method for hyperspectral water quality monitoring of the present invention, by establishing a closed-loop control mechanism between hyperspectral monitoring, sludge identification, interference judgment, and automatic sludge removal, achieves fully automated management of the entire process from the identification of sludge interference to the completion of cleaning and verification of recovery. Compared with the existing manual inspection and manual retrieval modes, the system of the present invention has a faster response speed, higher cleaning timeliness, and better guarantee of the continuity and integrity of monitoring data. The system of the present invention adopts a modular design concept, and the functional modules interact with each other through standardized communication interfaces and data formats, which facilitates subsequent functional expansion and system upgrades. The application scope of the present invention is not limited to the horizontal sedimentation tank of sewage treatment plants, but can also be extended to other water treatment structures or natural water body monitoring scenarios that require hyperspectral water quality monitoring and have surface sludge problems. Attached Figure Description

[0019] Figure 1 This is a three-dimensional schematic diagram of the overall structure of the present invention; Figure 2 This is a schematic cross-sectional view of the overall structure of the present invention; Figure 3 This is a block diagram of the scum identification and interference judgment module of the present invention; Figure 4 This is a block diagram of the control system module of the present invention; Figure 5 This is a block diagram of the cleaning effect verification unit of the present invention; In the diagram, 1. Hyperspectral monitoring module; 101. Hyperspectral water quality monitor; 102. Monitor mounting bracket; 1021. Vertical column; 1022. Horizontal beam; 1023. Adjustable connector; 103. First communication interface; 2. Scum identification and interference judgment module; 201. Camera image acquisition monitoring unit; 202. Data storage unit; 203. Scum identification algorithm unit; 204. Interference judgment algorithm unit; 205. Decision logic unit; 3. Control system module; 301. Programmable logic controller unit; 302. Signal input module; 303. Signal output module; 304. Communication module; 305. Human-machine interface; 306. Timing control logic block; 307. Action sequence control logic block; 308. Safety protection logic block; 309. Historical data storage unit; 4. Self 401. Dynamic slag removal execution module; 4011. Moving track system; 4012. Track beam; 4013. Drive motor; 4014. Transmission device; 4015. Position sensor; 402. Slag removal execution mechanism; 4021. Mounting frame; 4022. Walking mechanism; 4023. Salvage mechanism; 40231. Salvage arm; 40232. Drive cylinder; 40233. Salvage claw; 4024. Water flow disturbance mechanism; 40241. Nozzle; 40242. Small submersible pump; 403. Slag collection and discharge assembly; 4031. Collection tank; 4032. Conveying pipeline; 4033. Discharge pump; 4034. Slag collection box; 4035. Liquid level sensor; 4036. Slag quantity sensor; 5. Cleaning effect verification unit; 501. Second camera; 502. Image analysis software; 6. Horizontal flow sedimentation tank. Detailed Implementation 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, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0020] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0021] It should be noted that, unless otherwise specified, the embodiments and features described in this invention can be combined with each other.

[0022] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0023] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship commonly used when the product of this invention is in use, or the orientation or positional relationship commonly understood by those skilled in the art. They are only used for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention. In addition, the terms "first," "second," etc., are only used to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0024] In the description of this invention, it should also be noted that, unless otherwise explicitly specified and limited, the terms "set," "install," "connect," and "link" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0025] The hyperspectral water quality monitoring automatic scum removal system of the present invention is arranged above and inside a horizontal sedimentation tank, and the various functional modules are coordinated and operated through a control system module. In a specific embodiment, the system consists of a hyperspectral monitoring module 1, a scum identification and interference judgment module 2, a control system module 3, an automatic scum removal execution module 4, and a cleaning effect verification unit 5. The modules are connected through a standardized communication interface and a data bus to form a complete closed-loop control architecture.

[0026] The hyperspectral monitoring module 1 is positioned above the horizontal flow sedimentation tank 6 to continuously acquire spectral signals from the water surface and obtain water quality parameters. This module includes a hyperspectral water quality monitor 101, a monitor mounting bracket 102, and a first communication interface 103. The hyperspectral water quality monitor 101 is fixedly mounted on the horizontal flow sedimentation tank 6 at a predetermined position above it via the monitor mounting bracket 102. This predetermined position is located in the middle of the tank's length and near the outlet end to ensure that the monitoring field of view covers the main monitoring area of ​​the horizontal flow sedimentation tank.

[0027] The hyperspectral water quality monitor 101 operates in the wavelength range of 400 nm to 1000 nm, with a spectral resolution of no more than 5 nm, and a continuously adjustable monitoring frequency range of 1 minute to 60 minutes. When the monitoring frequency is set to 1 minute, the hyperspectral water quality monitor 101 can continuously acquire spectral signals from the water surface within the monitoring field of view, obtaining water quality parameter data in real time. When the monitoring frequency is set to 60 minutes, the hyperspectral water quality monitor 101 performs periodic spectral acquisition according to a preset time interval, suitable for applications with lower monitoring frequency requirements. In practical applications, the monitoring frequency setting can be adjusted according to the process operating conditions of the horizontal flow sedimentation tank and changes in the influent water quality to balance the integrity of monitoring data and system operating efficiency.

[0028] The monitoring instrument mounting bracket 102 is made of stainless steel, possessing excellent corrosion resistance and structural strength, and can adapt to the humid and corrosive working environment of a wastewater treatment plant. The mounting bracket 102 includes a vertical column 1021, a horizontal beam 1022, and an adjusting connector 1023. The lower end of the vertical column 1021 is fixedly connected to the upper edge of the wall of the horizontal sedimentation tank 6, and reliably secured using pre-embedded flanges and chemical anchors, ensuring that the bracket does not loosen or deform during long-term operation. One end of the horizontal beam 1022 is fixedly connected to the upper end of the vertical column 1021, and the other end extends horizontally into the tank body, forming a stable cantilever structure. The adjusting connector 1023 is located at the end of the horizontal beam 1022 and employs an adjustable connection structure to adjust the installation angle and height of the hyperspectral water quality monitor 101, ensuring that the central axis of the monitoring field of view forms a preset incident angle with the water surface. In one specific embodiment, the incident angle is set between 45 and 60 degrees, at which point the water surface spectral reflectance signal has the best acquisition efficiency and signal-to-noise ratio.

[0029] The first communication interface 103 is located on the housing of the hyperspectral water quality monitor 101. It adopts a standard Ethernet interface or RS485 interface and is used to establish a data communication connection with the control system module 3 to realize high-speed transmission of spectral data and synchronous control of the system.

[0030] The scum identification and interference judgment module 2 is located inside the control system module 3. As the core algorithm module of this invention, it is responsible for receiving data transmitted by the hyperspectral monitoring module 1, executing the scum identification and interference judgment algorithm, and outputting a decision signal to the control system module 3. This module 2 includes a camera image acquisition and monitoring unit 201, a data storage unit 202, a scum identification algorithm unit 203, an interference judgment algorithm unit 204, and a decision logic unit 205.

[0031] The camera image acquisition and monitoring unit 201 is positioned beside the hyperspectral water quality monitor 101, and the two are mounted together via a fixed bracket to ensure consistency in the field of view. The camera image acquisition and monitoring unit 201 is used for synchronous imaging and acquisition of the monitoring field of view of the hyperspectral water quality monitor 101. Its imaging chip is a complementary metal-oxide-semiconductor or charge-coupled device, with an effective pixel count of no less than two million pixels and a frame rate of no less than thirty frames per second, meeting the requirements of real-time imaging monitoring. The lens of the camera image acquisition and monitoring unit 201 is a fixed-focus lens with a focal length selectable from eight millimeters to twenty-five millimeters. The specific focal length is selected based on the coverage area of ​​the monitoring field of view and the imaging distance, ensuring that the lens's field of view can cover the entire monitoring field of view of the hyperspectral water quality monitor 101. In a specific embodiment, a fixed-focus lens with a focal length of twelve millimeters is selected, with a horizontal field of view of approximately sixty degrees, which can completely cover the monitoring field of view of the hyperspectral water quality monitor 101.

[0032] Data storage unit 202 is used to store a scum sample image database and preset interference judgment threshold parameters. The scum sample image database includes multiple sets of scum image samples collected under different working conditions. Each set of image samples is labeled with corresponding scum type, density level, and distribution morphology characteristics. Scum types include algae aggregation type, oil sludge floating type, flocculent suspended matter accumulation type, and mixed type. The density level is divided into three levels: light, moderate, and heavy, according to the proportion of scum covering the area. The principle for setting the interference judgment threshold is as follows: when the proportion of scum covering the area is less than the first preset threshold, it is judged as an interference-free state, and the system maintains normal operation mode; when the proportion of scum covering the area is greater than or equal to the first preset threshold and less than the second preset threshold, it is judged as a light interference state, and the system starts the warning mode; when the proportion of scum covering the area is greater than or equal to the second preset threshold, it is judged as a heavy interference state, and the system immediately starts the scum cleaning execution program. In a specific embodiment, the first preset threshold is set to 15%, and the second preset threshold is set to 30%. This threshold setting has been repeatedly tested and verified, and can achieve accurate judgment and timely response to scum interference under different working conditions.

[0033] The scum identification algorithm unit 203 is used to preprocess, extract features, and perform pattern matching on the images acquired by the camera image acquisition and monitoring unit 201 to identify the presence of scum and determine its type and distribution location. Image preprocessing steps include noise reduction, contrast enhancement, and edge sharpening to improve the accuracy of subsequent feature extraction. The feature extraction process extracts texture, color, and morphological features from the scum image to form a feature vector. The pattern matching process compares the extracted feature vector with the scum sample image database stored in the data storage unit 202 to determine the type and distribution location of the scum.

[0034] The interference judgment algorithm unit 204 calculates the proportion of the area obstructed by scum within the monitoring field of view based on the recognition result of the scum recognition algorithm unit 203, compares the calculation result with the preset interference judgment threshold in the data storage unit 202, and outputs an interference level signal. The method for calculating the proportion of the obstructed area is as follows: the ratio of the number of pixels in the scum region to the total number of pixels in the monitoring field of view image is statistically expressed as a percentage.

[0035] The decision logic unit 205 generates corresponding control commands based on the interference level signal output by the interference judgment algorithm unit 204, and sends the control commands to the control system module 3. When the interference level is no interference, the system maintains normal operation mode and continues to execute the monitoring cycle; when the interference level is mild interference, the system enters early warning mode, triggers an audible and visual alarm to alert maintenance personnel, and shortens the execution cycle of the monitoring cycle; when the interference level is severe interference, the system immediately starts the slag removal execution program and calls the automatic slag removal execution module 4 to clean the floating slag.

[0036] The control system module 3 uses a programmable logic controller (PLC) as the core control unit, and is electrically connected to the hyperspectral monitoring module 1, the scum identification and interference judgment module 2, and the automatic scum removal execution module 4 to achieve coordinated control of the entire system. This control system module 3 includes a PLC unit 301, a signal input module 302, a signal output module 303, a communication module 304, and a human-machine interface 305.

[0037] The programmable logic controller (PLC) unit 301 is a modular controller with both digital and analog input / output capabilities. Its CPU has a processing power of at least 32 bits and a clock frequency of at least 100 MHz, capable of meeting the real-time computation requirements of complex control algorithms. Internally, the PLC unit 301 integrates a timing control logic block 306, an action sequence control logic block 307, a safety protection logic block 308, and a historical data storage unit 309. The timing control logic block 306 manages the time coordination between modules, ensuring that the hyperspectral data acquisition, scum identification, and scum removal processes are executed sequentially according to a preset time sequence. The action sequence control logic block 307 defines the motion sequence of the automatic scum removal module 4, including track movement, retrieval actions, collection actions, and reset actions. The safety protection logic block 308 monitors abnormal status signals during system operation, immediately stopping all actions and triggering an alarm upon detecting an over-limit signal. The historical data storage unit 309 records all monitoring data, identification results, interference judgment records, scum removal execution records, and fault records during system operation, facilitating subsequent data analysis and process optimization.

[0038] The signal input module 302 receives spectral data from the hyperspectral monitoring module 1 and decision signals from the scum identification and interference judgment module 2. The signal output module 303 sends control commands to the automatic scum removal execution module 4. The communication module 304 enables data exchange with the host computer monitoring system or remote management platform, supporting both Ethernet and wireless communication. The human-machine interface 305 is located on the control cabinet panel and uses a touch screen display to show system operating status parameters, alarm information, and manual operation control buttons. The control system module 3 also includes a manual operation mode, allowing maintenance personnel to directly control the movement and retrieval actions of the scum removal execution mechanism via manual operation control buttons when necessary, to handle special working conditions or perform equipment maintenance and repair.

[0039] The automatic scum removal module 4 is installed inside the horizontal flow sedimentation tank 6 and is used to automatically remove the identified scum. This module 4 includes a moving track system 401, a scum removal actuator 402, and a scum collection and discharge assembly 403.

[0040] The moving track system 401 includes two parallel track beams 4011, a drive motor 4012, a transmission device 4013, and position sensors 4014. The two track beams 4011 extend along the length of the horizontal sedimentation tank 6 and are respectively positioned below the walkway plates on both sides of the tank body, fixed by embedded parts and connectors. The drive motor 4012 is an AC servo motor or stepper motor with an output power of not less than 200 watts, providing ample power to drive the slag-cleaning actuator 402 to move along the track direction. The transmission device 4013 includes a reducer and a rack and pinion mechanism, used to convert the rotational motion output by the drive motor 4012 into linear motion along the track direction, achieving precise position control. The position sensors 4014 are located at both ends and the middle of the track beams 4011, employing photoelectric or magnetic induction position sensors, used to detect the real-time position of the slag-cleaning actuator 402 on the track and feed the position signal back to the control system module 3, achieving precise speed control and stop control.

[0041] In one specific embodiment, the track length of the mobile track system 401 is twelve meters, the track spacing is three meters, and the drive motor 4012 adopts an AC servo motor with a rated power of 400 watts and a rated speed of 3,000 revolutions per minute. With the help of a reducer with a reduction ratio of 1:20 and a gear rack mechanism with a module of 3, it can achieve motion control performance with a maximum moving speed of six meters per minute and a minimum positioning accuracy of ±5 millimeters.

[0042] The slag removal actuator 402 includes a mounting frame 4021, a traveling mechanism 4022, a retrieval mechanism 4023, and a water flow disturbance mechanism 4024. The mounting frame 4021 is mounted on the mobile track system 401 via the traveling mechanism 4022 and can move back and forth along the track direction. The traveling mechanism 4022 adopts a wheeled traveling structure, with the drive wheels meshing with the rack on the track beam 4011 to achieve reliable travel along the track direction.

[0043] The retrieval mechanism 4023 is located at the bottom of the mounting frame 4021 and includes a retrieval arm 4031, a drive cylinder 4032, and a retrieval claw 4033. The retrieval arm 4031 is a telescopic structure made of stainless steel or aluminum alloy, with an extension length of no less than two meters, enabling it to cover a wider water surface area for scum retrieval. The drive cylinder 4032 is a hydraulic or pneumatic cylinder, installed inside the mounting frame 4021, used to drive the horizontal extension and retraction of the retrieval arm 4031. In one specific embodiment, the drive cylinder 4032 is a pneumatic cylinder with a diameter of 80 mm and a stroke of 2000 mm, providing sufficient thrust to push the retrieval arm 4031 to complete the extension and retraction actions. The retrieval claw 4033 is located at the end of the retrieval arm 4031, employing a spring-reset structure, and consists of multiple openable and closable claw flaps for grabbing and collecting scum. After the retrieval claw 4033 grabs the scum, the claw flaps remain closed under the action of the spring, locking the scum inside the claw flaps; when the retrieval claw 4033 reaches the collection position, the claw flaps are released by the trigger mechanism, allowing the scum to fall into the collection trough below.

[0044] A water flow disturbance mechanism 4024 is located on both sides of the retrieval mechanism 4023, and includes several nozzles 4041 and a small submersible pump 4042. The nozzles 4041 are installed on the bottom side of the mounting frame 4021 to spray water and create disturbance. The small submersible pump 4042 is located inside the horizontal sedimentation tank 6, with an output power of not less than 500 watts and a flow rate of not less than 20 cubic meters per hour, to provide the water source for water flow disturbance. Before the retrieval operation, the water flow disturbance mechanism 4024 is activated to disturb the water flow in the scum accumulation area, causing the dispersed scum to converge towards the retrieval area, facilitating the retrieval claw 4033 to grab and collect it.

[0045] The scum collection and discharge assembly 403 includes a collection tank 4031, a conveying pipe 4032, a discharge pump 4033, and a scum collection box 4034. The collection tank 4031 is located on the inner wall of the horizontal sedimentation tank 6 and is connected to the collection position of the retrieval claw 4033 to receive the scum collected. The collection tank 4031 is made of stainless steel, has a volume of not less than 0.2 cubic meters, and is equipped with a filter baffle and a drain outlet to achieve scum-water separation. One end of the conveying pipe 4032 is connected to the collection tank 4031, and the other end is connected to the scum collection box 4034. It is made of PVC or stainless steel and has good corrosion resistance. The discharge pump 4033 is located on the conveying pipe 4032, is a screw pump or diaphragm pump, and has an output power of not less than 1.5 kilowatts. It is used to pump the scum mixture in the collection tank 4031 to the scum collection box 4034. The scum collection box 4034 is located at the edge of the horizontal flow sedimentation tank 6 or below the walkway, with a volume of not less than 0.5 cubic meters, and has dumping and cleaning functions. The scum collection box 4034 is equipped with a liquid level sensor 4035 and a scum quantity sensor 4036 to monitor the amount of scum collected and the wastewater level within the scum collection box 4034. When the amount of scum in the scum collection box 4034 reaches a preset full-load threshold or the liquid level reaches a preset overflow threshold, the system automatically triggers an alarm signal, prompting maintenance personnel to clean the scum collection box 4034 in a timely manner.

[0046] The cleaning effect verification unit 5 includes a second camera 501 and image analysis software 502. The second camera 501 is located on the side of the hyperspectral monitoring module 1 and is used to perform secondary imaging of the original monitoring field of view after the slag removal is completed. The second camera 501 of the cleaning effect verification unit 5 uses the same model of industrial digital camera as the camera imaging monitoring unit 201 to ensure consistency and comparability of image acquisition conditions. The image analysis software 502 is used to process and analyze the secondary imaging images acquired by the second camera 501, determine the residual slag within the monitoring field of view after cleaning, and send the analysis results to the control system module 3 as the basis for cleaning effect verification. The processing logic of the image analysis software 502 is as follows: when the proportion of residual slag area after cleaning is less than a preset cleaning qualification threshold, the cleaning effect is judged to be qualified, and the system returns to normal operation mode; when the proportion of residual slag area after cleaning is greater than or equal to the preset cleaning qualification threshold, the cleaning effect is judged to be unqualified, and the system repeats the slag removal execution procedure. In a specific embodiment, the cleaning qualification threshold is set to five percent, which ensures that the cleaning effect meets the expected standard.

[0047] The control method of the automatic scum removal system for hyperspectral water quality monitoring of the present invention includes the following steps: Step one: The hyperspectral monitoring module 1 continuously acquires spectral signals from the water surface of the horizontal sedimentation tank 6, obtains real-time water quality parameter data, and transmits the data to the control system module 3. After the hyperspectral water quality monitor 101 starts up, it first performs self-test and calibration operations, including light source intensity calibration, spectral response calibration, and camera parameter calibration. After the self-test and calibration are completed, the system enters normal operation mode and performs spectral signal acquisition according to the preset monitoring cycle. In a specific embodiment, the monitoring cycle is set to five minutes, that is, a complete spectral signal acquisition and data processing process is completed every five minutes.

[0048] Step two: The control system module 3 synchronously activates the camera image acquisition and monitoring unit 201 to acquire images of the monitoring field of view of the hyperspectral monitoring module 1. A data synchronization mechanism is set up between the hyperspectral monitoring module 1 and the scum identification and interference judgment module 2 to ensure that the image acquisition of the camera image acquisition and monitoring unit 201 and the spectral acquisition of the hyperspectral water quality monitor 101 are synchronized in time, so as to facilitate subsequent data correlation analysis.

[0049] Step 3: The scum identification algorithm unit 203 preprocesses, extracts features, and performs pattern matching on the images acquired by the camera to identify the presence of scum and determine its type and distribution location. The image preprocessing steps first denoise the original image using median filtering or Gaussian filtering to remove random noise; then, it enhances contrast using histogram equalization or adaptive contrast enhancement algorithms; finally, it sharpens edges using Laplacian or Sobel operators to enhance edge information. The feature extraction process extracts texture features from the scum image, including gray-level co-occurrence matrix features and local binary pattern features; color features, including HSV color space features; and morphological features, including contour features, area features, and compactness features. The pattern matching process compares the extracted feature vectors with the scum sample image database stored in the data storage unit 202, using a support vector machine or convolutional neural network classifier for classification and identification to determine whether the scum is algal aggregation, oil sludge floating, flocculent suspended matter accumulation, or a mixed type, and determines the distribution coordinates of the scum within the monitoring field of view.

[0050] Step four: The interference judgment algorithm unit 204 calculates the proportion of the occlusion area of ​​the floating scum within the monitoring field of view based on the recognition results of the floating scum recognition algorithm unit 203. The occlusion area proportion is calculated as follows: First, the identified floating scum area is binarized to generate a floating scum mask image; then, the number of white pixels in the floating scum mask image is counted, i.e., the number of pixels in the floating scum area; finally, the ratio of the number of pixels in the floating scum area to the total number of pixels in the monitoring field of view is calculated to obtain the occlusion area proportion. The interference judgment algorithm unit 204 compares the calculated occlusion area proportion with the preset interference judgment threshold in the data storage unit 202 to determine the current interference level. When the occlusion area proportion is less than 15% of the first preset threshold, it is judged as no interference; when the occlusion area proportion is greater than or equal to 15% of the first preset threshold and less than 30% of the second preset threshold, it is judged as a slight interference; when the occlusion area proportion is greater than or equal to 30% of the second preset threshold, it is judged as a severe interference.

[0051] Step 5: The decision logic unit 205 generates corresponding control commands based on the interference level. When the interference level is no interference, the system maintains normal operation mode and continues to execute the monitoring cycle from Step 1 to Step 4. When the interference level is mild interference, the system enters early warning mode, triggering an audible and visual alarm to alert maintenance personnel. Simultaneously, the monitoring cycle is shortened from five minutes to two minutes, increasing the monitoring frequency to closely track changes in scum. When the interference level is severe interference, the system immediately initiates the scum removal program, calling the automatic scum removal module 4 to clean the scum.

[0052] The slag removal procedure includes the following sub-steps: In sub-step 501, the control system module 3 sends a cleaning start command to the automatic cleaning execution module 4. The cleaning execution mechanism 402 starts from its initial position and moves along the moving track system 401 to the preset area to be cleaned. Upon receiving the start command, the drive motor 4012 begins operation, driving the cleaning execution mechanism 402 to move along the track direction via the transmission device 4013. The position sensor 4014 detects the position of the cleaning execution mechanism 402 in real time and feeds back the position signal to the control system module 3, achieving precise position control.

[0053] In sub-step 502, after the scum removal actuator 402 reaches the area to be cleaned, the water flow disturbance mechanism 4024 is activated. The small submersible pump 4042 starts operating, spraying pool water from the nozzle 4041 to form a directional water flow that disturbs the scum accumulation area. After the water flow disturbance continues for a certain period of time, the dispersed scum is pushed towards the retrieval area by the water flow, forming a relatively concentrated scum pile, which is convenient for the retrieval claw 4033 to grab. In a specific embodiment, the water flow disturbance lasts for thirty seconds.

[0054] In sub-step 503, the drive cylinder 4032 of the retrieval mechanism 4023 pushes the retrieval arm 4031 to extend horizontally, and the retrieval claw 4033 enters the scum accumulation area to perform a grabbing and collection action. The drive cylinder 4032 extends, driving the retrieval arm 4031 to extend towards the water surface, while the retrieval claw 4033 is in an open state; when the retrieval claw 4033 enters the scum accumulation area, the drive cylinder 4032 stops extending, and the retrieval claw 4033 closes under the action of the spring, grabbing the scum.

[0055] In sub-step 504, after the retrieval is completed, the drive cylinder 4032 retracts, and the retrieval arm 4031 drives the retrieval claw 4033 to reset, returning to above the collection tank 4031. When the slag removal actuator 402 moves above the collection tank 4031, the release mechanism of the retrieval claw 4033 is triggered, the claw flaps open, and the slag falls into the collection tank 4031.

[0056] In sub-step 505, the discharge pump 4033 is started, pumping the scum mixture in the collection tank 4031 to the scum collection box 4034 through the conveying pipe 4032. The discharge pump 4033 uses a positive pressure conveying method to transport the scum mixture from the collection tank 4031 to the scum collection box 4034, thereby achieving centralized collection of scum.

[0057] In sub-step 506, the sludge removal actuator 402 returns to its initial position along the moving track system 401 or moves to the next area to be cleaned, repeating sub-steps 502 to 505 until all areas to be cleaned have been treated. The maximum travel of the moving track system 401 covers the entire width of the water surface of the horizontal sedimentation tank 6, allowing the sludge removal actuator 402 to move back and forth to clean different areas. In one specific embodiment, after completing one cleaning operation, the sludge removal actuator 402 moves to the next adjacent area to continue cleaning until the entire monitoring field of view is covered.

[0058] In sub-step 507, the cleaning effect verification unit 5 is activated to perform secondary imaging acquisition on the original monitoring field of view. The second camera 501 performs imaging acquisition on the cleaned monitoring field of view to obtain an image of the cleaned water surface.

[0059] In sub-step 508, image analysis software 502 processes and analyzes the secondary imaging image to determine whether the proportion of residual scum area after cleaning is less than 5% of the qualified cleaning threshold. The image processing and analysis method is the same as the scum identification method in step three: extracting image features and performing pattern matching, counting the number of pixels in the residual scum area, and calculating the residual area proportion.

[0060] In sub-step 509, if the cleaning is satisfactory (i.e., the residual area percentage is less than 5%), the system returns to normal operation and continues the monitoring cycle of steps one through four. If the cleaning is unsatisfactory (i.e., the residual area percentage is greater than or equal to 5%), the system repeats the slag removal procedure for a second cleaning of the residual slag. The number of repeated cleanings is determined based on the actual situation, but shall not exceed three times. If the satisfactory standard is still not met after three cleanings, the system triggers a deep cleaning alarm, prompting maintenance personnel to intervene.

[0061] The execution timing of the entire control method is uniformly managed by the timing control logic block 306. Preset time delays and waiting confirmation logic are set between each stage of hyperspectral data acquisition, scum identification, interference judgment and scum removal to ensure the coordination of actions between modules and the stability of system operation.

[0062] The control method also includes step six, the safety protection monitoring step: During the entire system operation, the safety protection logic block 308 continuously monitors the operating status parameters of each module, including the operating current and temperature signals of the drive motor 4012, the position signal of the position sensor 4014, the air or hydraulic pressure signal of the drive cylinder 4032, and the communication status between modules. When any parameter is detected to exceed the preset safety range, the safety protection logic block 308 immediately executes protection actions, including stopping the operation of the drive motor 4012, locking the position of the retrieval mechanism 4023, cutting off the power supply to the discharge pump 4033, and triggering the corresponding alarm signal. At the same time, the control system module 3 records the time, type, and detailed parameter information of the fault, and displays the fault prompt to the maintenance personnel through the human-machine interface 305.

[0063] To verify the technical effects of the present invention, the following comparative experiments were conducted with respect to the embodiments and comparative examples: Example 1: The automatic scum removal system for hyperspectral water quality monitoring of the present invention was used for water quality monitoring and scum removal. The system parameters were set as follows: the monitoring cycle of the hyperspectral monitoring module 1 was set to five minutes; the first preset value for interference judgment threshold was 15%, the second preset value was 30%, and the qualified removal threshold was 5%. The system ran continuously for thirty days, during which a total of twelve scum interference events occurred, including seven minor interference events and five severe interference events. After detecting scum interference, the average response time of the system was five minutes and thirty seconds, the average time to complete the removal was forty-five minutes, the residual scum area after removal was less than 5%, and the water quality monitoring data integrity rate was 98.5%.

[0064] Comparative Example 1: Water quality monitoring and scum removal were conducted using traditional manual inspection and dredging methods. Manual inspections were conducted every two hours, and maintenance personnel were notified to clean up any scum interference detected. Under the same operating conditions, scum interference events occurred twelve times over thirty days, with an average response time of three hours and twenty minutes and an average cleanup time of two hours and fifteen minutes. Due to the delayed response, some monitoring data was missing, resulting in a water quality monitoring data integrity rate of 76.2%.

[0065] Example 2: Performance testing was conducted on the automatic slag removal execution module 4 of the present invention. The test included the operational stability of the moving track system 401, the retrieval efficiency of the slag removal execution mechanism 402, and the discharge capacity of the slag collection and discharge component 403. The test results showed that the moving track system 401 operated without failure for one hundred hours, demonstrating good operational stability; the slag removal execution mechanism 402 had a single retrieval cycle of twelve seconds and a retrieval success rate of 97%; the discharge pump 4033 of the slag collection and discharge component 403 had a flow rate of fifteen cubic meters per hour under rated operating conditions and a discharge efficiency of 94%.

[0066] Comparative Example 2: Using traditional manual slag removal methods, the cycle of a single slag removal operation is approximately three minutes, with a slag removal success rate of 75% and a discharge efficiency of 60%. The manual slag removal method in Comparative Example 2 is significantly inferior to the automatic slag removal module 4 of this invention in terms of both slag removal efficiency and discharge capacity.

[0067] Example 3: Adaptability tests were conducted on the hyperspectral water quality monitoring automatic scum removal system of the present invention under different water quality conditions. Test conditions included influent COD concentrations ranging from 100 mg / L to 500 mg / L, suspended solids concentrations ranging from 50 mg / L to 300 mg / L, and scum types including algal aggregation, oily floating, flocculent suspended solids accumulation, and mixed types. Test results showed that the system of the present invention maintained stable monitoring performance even under high-concentration influent conditions, with a scum identification accuracy of no less than 95%, a scum removal success rate of 92%, and good overall system reliability.

[0068] Comparative Example 3: Using traditional methods for monitoring and sludge removal under different water quality conditions, the frequency of manual inspections is difficult to adapt to changes in water quality, sludge identification mainly relies on manual judgment, with an accuracy rate of about 70%, and a sludge removal success rate of about 60%, resulting in low overall system reliability.

[0069] Example 4: Energy consumption of the automatic scum removal system for hyperspectral water quality monitoring of the present invention was tested and analyzed. Test results show that the overall system power consumption is approximately 3 kilowatts, of which the hyperspectral water quality monitor 101 consumes approximately 500 watts, the drive motor 4012 of the moving track system 401 consumes approximately 800 watts, the scum removal actuator 402 consumes approximately 900 watts, the discharge pump 4033 of the scum collection and discharge component 403 consumes approximately 600 watts, and other auxiliary equipment consumes approximately 200 watts. Compared with traditional manual inspection and manual dredging methods, the system of the present invention can reduce the workload of maintenance personnel while optimizing energy consumption through intelligent control, resulting in an overall energy consumption reduction of approximately 30%.

[0070] Comparative Example 4: Traditional manual inspection and manual retrieval methods require multiple maintenance personnel, resulting in high labor costs. Furthermore, work efficiency is greatly affected by personnel skills and physical strength, and long-term operating costs are significantly higher than the automatic slag removal system of this invention.

[0071] The comparative experiments of the above embodiments and comparative examples demonstrate that the automatic scum removal system and control method for hyperspectral water quality monitoring of the present invention have the advantages of high monitoring accuracy, accurate scum identification, fast response speed, high cleaning efficiency, stable and reliable operation, and low energy consumption. It can effectively solve the problems of scum interference and high manual maintenance costs in the prior art, and has significant technical effects and broad application prospects.

[0072] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An automatic scum removal system for hyperspectral water quality monitoring, characterized in that, include: The hyperspectral monitoring module (1) is fixedly installed above the horizontal sedimentation tank (6) to continuously collect water surface spectral signals and obtain water quality parameters; The scum identification and interference judgment module (2) is located inside the control system module (3) and is used to receive data transmitted by the hyperspectral monitoring module (1) and execute the scum identification and interference judgment algorithm to output a decision signal; The control system module (3) uses a programmable logic controller as the core control unit and is electrically connected to the hyperspectral monitoring module (1), the scum identification and interference judgment module (2), and the automatic scum removal execution module (4), respectively. The automatic slag removal execution module (4) is located inside the horizontal sedimentation tank (6) and includes a moving track system (401) extending along the length of the tank, a slag removal execution mechanism (402) installed on the moving track system (401), and a slag collection and discharge assembly (403) connected to the slag removal execution mechanism (402).

2. The automatic scum removal system for hyperspectral water quality monitoring according to claim 1, characterized in that, The hyperspectral monitoring module (1) includes a hyperspectral water quality monitor (101), a monitor mounting bracket (102), and a first communication interface (103). The hyperspectral water quality monitor (101) is fixedly installed above the horizontal sedimentation tank (6) at a preset position via the monitor mounting bracket (102). The preset position is located in the middle part of the tank body along its length and close to the outlet end. The first communication interface (103) is located on the housing of the hyperspectral water quality monitor (101) and is used to establish a data communication connection with the control system module (3).

3. The automatic scum removal system for hyperspectral water quality monitoring according to claim 2, characterized in that, The mounting bracket (102) of the monitoring instrument is made of stainless steel and includes a vertical column (1021), a horizontal beam (1022), and an adjusting connector (1023). The lower end of the vertical column (1021) is fixedly connected to the upper edge of the wall of the horizontal sedimentation tank (6). One end of the horizontal beam (1022) is fixedly connected to the upper end of the vertical column (1021), and the other end extends horizontally into the tank. The adjusting connector (1023) is located at the end of the horizontal beam (1022) and is used to adjust the installation angle and height of the hyperspectral water quality monitor (101) so that the central axis of the monitoring field of view forms a preset incident angle with the water surface.

4. The automatic scum removal system for hyperspectral water quality monitoring according to claim 1, characterized in that, The scum identification and interference judgment module (2) includes a camera image monitoring unit (201), a data storage unit (202), a scum identification algorithm unit (203), an interference judgment algorithm unit (204), and a decision logic unit (205). The camera image acquisition and monitoring unit (201) is located beside the hyperspectral water quality monitor (101) and is used to synchronously acquire images of the monitoring field of view of the hyperspectral water quality monitor (101); the data storage unit (202) is used to store the scum sample image database and the preset interference judgment threshold parameters; the scum identification algorithm unit (203) is used to extract features and match patterns from the images acquired by the camera image acquisition and monitoring unit (201) to determine whether there is scum within the monitoring field of view; the interference judgment algorithm unit (204) is used to calculate the scum occlusion area ratio based on the scum identification result and compare it with the preset interference judgment threshold to output the interference level signal; the decision logic unit (205) is used to generate a scum removal execution signal based on the interference level signal and send it to the control system module (3).

5. The automatic scum removal system for hyperspectral water quality monitoring according to claim 4, characterized in that, The camera image acquisition and monitoring unit (201) adopts an industrial digital camera, whose imaging chip is a complementary metal oxide semiconductor or charge-coupled device, with an effective pixel count of not less than 2 million pixels and a frame rate of not less than 30 frames per second.

6. The automatic scum removal system for hyperspectral water quality monitoring according to claim 4, characterized in that, The principle for setting the interference judgment threshold is as follows: when the proportion of the area covered by scum is less than the first preset threshold, it is judged to be an interference-free state, and the system maintains normal operation mode. When the proportion of the area covered by scum is greater than or equal to the first preset threshold and less than the second preset threshold, it is judged as a mild interference state, and the system starts the early warning mode; when the proportion of the area covered by scum is greater than or equal to the second preset threshold, it is judged as a severe interference state, and the system immediately starts the scum removal execution program.

7. The automatic scum removal system for hyperspectral water quality monitoring according to claim 1, characterized in that, The control system module (3) includes a programmable logic controller unit (301), a signal input module (302), a signal output module (303), a communication module (304), and a human-machine interface (305); the programmable logic controller unit (301) integrates a timing control logic block (306), an action sequence control logic block (307), and a safety protection logic block (308).

8. The automatic scum removal system for hyperspectral water quality monitoring according to claim 1, characterized in that, The automatic slag removal execution module (4) includes a moving track system (401), a slag removal execution mechanism (402), and a slag collection and discharge assembly (403); the moving track system (401) includes two parallel track beams (4011), a drive motor (4012), a transmission device (4013), and a position sensor (4014); the slag removal execution mechanism (402) includes an installation frame (4021), a walking mechanism (4022), a retrieval mechanism (4023), and a water flow disturbance mechanism (4024); the slag collection and discharge assembly (403) includes a collection tank (4031), a conveying pipe (4032), a discharge pump (4033), and a slag collection box (4034).

9. A control method for an automatic scum removal system for hyperspectral water quality monitoring, characterized in that, Includes the following steps: Step 1: The hyperspectral monitoring module (1) continuously collects spectral signals from the surface of the horizontal sedimentation tank (6), obtains real-time water quality parameter data, and transmits the data to the control system module (3). Step 2: The control system module (3) synchronously starts the camera image acquisition monitoring unit (201) to acquire images of the monitoring field of view of the hyperspectral monitoring module (1); Step 3: The scum identification algorithm unit (203) preprocesses, extracts features and matches patterns on the images captured by the camera, identifies whether scum exists in the images and determines the type and distribution location of the scum. Step 4: The interference judgment algorithm unit (204) calculates the proportion of the scum obstructing the monitoring field of view based on the scum identification results, and compares it with the preset interference judgment threshold to determine the current interference level. Step 5, the decision logic unit (205) generates corresponding control instructions according to the interference level: when the interference level is no interference, the system maintains normal operation mode and continues to execute the monitoring cycle from step 1 to step 4; when the interference level is mild interference, the system enters early warning mode; when the interference level is severe interference, the system immediately starts the slag removal execution program and calls the automatic slag removal execution module (4) to clean the floating slag. The slag removal procedure includes: the slag removal actuator (402) moves along the moving track system (401) to the area to be cleaned, the water flow disturbance mechanism (4024) disturbs the water flow in the slag accumulation area, the scooping mechanism (4023) performs a grabbing and collecting action, the discharge pump (4033) transports the collected slag to the slag collection box (4034), and the cleaning effect verification unit (5) verifies and evaluates the cleaning effect.

Images