Sewage disposal machine control method

Abstract

This application provides a method for controlling a trash rack cleaning machine. It simultaneously collects multi-source data, including liquid level difference data before and after the trash rack, flow rate through the rack, and image data of floating debris in front of the rack. After time alignment and comprehensive judgment of the multi-source data, the method controls the start and stop of the trash rack cleaning machine based on the liquid level difference, flow rate through the rack, and floating debris image recognition results. Implementing this application enables accurate start and stop of the trash rack cleaning machine, improving the reliability of its start and stop control.

Description

Technical Field

[0001] This application relates to the field of water body cleaning technology, and in particular to a method for controlling a cleaning machine. Background Technology

[0002] Hydraulic cleaning machines are installed at the trash racks at the inlet of hydropower stations or pumping stations to remove floating debris trapped in front of the racks, preventing it from entering the hydropower station or pumping station. During operation, the cleaning machine needs to be started and stopped according to the degree of blockage in front of the trash racks to promptly remove floating debris and restore the flow capacity of the trash racks.

[0003] Existing screen cleaning machines typically use liquid level difference control for start and stop. That is, a liquid level difference sensor collects the liquid level difference before and after the screen. When the liquid level difference is greater than the start threshold, the screen cleaning machine starts; when the liquid level difference is less than the stop threshold, the machine stops.

[0004] However, in actual operating conditions, the liquid level difference is easily affected by factors unrelated to blockage, such as fluctuations in upstream water flow and downstream gate regulation. The liquid level changes caused by these disturbances are difficult to distinguish from those caused by the accumulation of floating debris. Furthermore, when floating debris has just begun to accumulate and the flow area has not yet significantly decreased, the liquid level difference has not yet reached the activation threshold, causing the trash rack to fail to respond promptly and resulting in a delayed start. Therefore, relying solely on the liquid level difference is insufficient to stably and accurately reflect the degree of blockage in the trash rack, and the reliability of the trash rack's start-stop control is inadequate. Summary of the Invention

[0005] This application provides a method for controlling a cleaning machine, which can accurately start and stop the cleaning machine and improve the reliability of the start and stop control of the cleaning machine.

[0006] To achieve the above objectives, the embodiments of this application adopt the following technical solutions: Firstly, a method for controlling a cleaning machine is provided, applied to a control system. The control system includes a multimodal sensing unit, an edge intelligent control unit, and a variable frequency drive unit. The edge intelligent control unit includes a data preprocessing module and a multi-source fusion decision module. The method includes: The multimodal sensing unit collects data on the liquid level difference before and after the trash rack, the flow rate through the rack, and images of floating debris in front of the rack. The data preprocessing module preprocesses the liquid level difference data, the flow rate through the rack, and the floating debris image data. The multi-source fusion decision module identifies the floating debris coverage rate based on the floating debris image data. When one or more of the liquid level difference, the flow rate through the rack, and the floating debris coverage rate meet the corresponding start-up conditions, the module outputs a start command for the trash rack cleaning machine to the variable frequency drive unit. When all of the liquid level difference, the flow rate through the rack, and the floating debris coverage rate meet the corresponding stop-up conditions, the module outputs a stop command for the trash rack cleaning machine to the variable frequency drive unit. The variable frequency drive unit controls the drive motor to start according to the start command and controls the drive motor to stop according to the stop command. The start-up conditions include a liquid level difference greater than or equal to a first liquid level difference threshold, a flow rate ratio less than a first ratio threshold, and a floating debris coverage rate greater than or equal to a first coverage rate threshold; the shutdown conditions include a liquid level difference less than or equal to a second liquid level difference threshold, a flow rate ratio greater than or equal to a second ratio threshold, and a floating debris coverage rate less than or equal to a second coverage rate threshold; the flow rate ratio is the ratio of the flow rate data to the reference flow rate, the reference flow rate is the flow rate through the screen when the screen is operating normally, the first liquid level difference threshold is greater than the second liquid level difference threshold, the first ratio threshold is less than the second ratio threshold, and the first coverage rate threshold is greater than the second coverage rate threshold.

[0007] By implementing the embodiments of this application, the liquid level difference, flow rate through the screen, and images of floating objects in front of the screen are collected by the multimodal sensing unit. After the fusion decision module performs time alignment on the multi-source data, the start-stop judgment is made based on the liquid level difference, flow rate through the screen ratio, and floating object coverage. This can more stably and accurately reflect the true degree of blockage of the trash rack and improve the reliability of the start-stop control of the trash rack.

[0008] In some embodiments, the edge intelligent control unit further includes a load adaptive adjustment module; the method further includes: a multimodal sensing unit acquiring operating condition data of the drive motor of the cleaning machine; after the cleaning machine starts, the load adaptive adjustment module identifies the density and mass of floating objects based on floating object image data, and determines the operating frequency of the variable frequency drive unit based on the floating object density; the operating frequency is used by the variable frequency drive unit to control the speed of the drive motor; when the mass of floating objects is greater than or equal to a mass threshold, the load adaptive adjustment module reduces the speed of the drive motor according to the speed reduction ratio and increases the output torque of the drive motor according to the torque increase ratio, and outputs control commands corresponding to the output torque and speed to the variable frequency drive unit; wherein, the speed reduction ratio and the torque increase ratio are preset; the variable frequency drive unit adjusts the output torque and speed of the drive motor according to the control commands.

[0009] In some embodiments, determining the operating frequency of the variable frequency drive unit based on the floating object density includes: determining the operating frequency as a first frequency when the floating object density is less than a first density value; determining the operating frequency as a third frequency when the floating object density is greater than a second density value; and determining the operating frequency as a second frequency when the floating object density is greater than or equal to the first density value and less than or equal to the second density value; wherein the first density value is less than the second density value, the first frequency is lower than the second frequency, and the second frequency is lower than the third frequency.

[0010] In some embodiments, the operating condition data includes the current data of the drive motor, and the edge intelligent control unit further includes a fault early warning module; the method further includes: during the operation of the cleaning machine, the fault early warning module compares the operating condition data with a pre-established normal parameter range, determines the abnormality level based on the degree of deviation of the actual data from the normal range, and executes the protection action corresponding to the abnormality level.

[0011] In some embodiments, preprocessing of the liquid level difference data, the flow rate through the grid, and the floating object image data includes: filtering and denoising the liquid level difference data and the flow rate through the grid, and removing outliers; performing histogram equalization and dehazing on the floating object image data; performing time synchronization on the liquid level difference data, the flow rate through the grid, and the floating object image data using a network time protocol; performing time alignment on the liquid level difference data, the flow rate through the grid, and the floating object image data; and performing normalization on the liquid level difference data, the flow rate through the grid, and the floating object image data.

[0012] In some embodiments, the fusion decision module further includes a time-series prediction model, which is a long short-term memory neural network; the method further includes: the fusion decision module inputs the liquid level difference data and the flow rate data through the screen for a first preset duration into the time-series prediction model to obtain the blockage probability of the screen within a second preset duration in the future; the fusion decision module outputs a screen cleaning machine start command when the blockage probability is greater than or equal to a preset blockage probability threshold.

[0013] In some embodiments, the multimodal sensing unit includes a liquid level difference sensor, a flow sensor, and an industrial vision camera. The multimodal sensing unit also includes a motor condition sensor, which is used to collect operating data of the drive motor, including current, temperature, torque, and speed. The motor condition sensor includes a current sensor, a temperature sensor, a torque sensor, and a speed sensor, which are used to collect the current, temperature, torque, and speed of the drive motor, respectively. The industrial vision camera is equipped with an automatic fill light and a wiper device. The automatic fill light is used to supplement illumination under low light conditions, and the wiper device is used to clean the lens of the industrial vision camera.

[0014] Secondly, an electronic device is provided, which has the function of implementing the cleaning machine control method of the first aspect described above. This function can be implemented by hardware or by hardware executing corresponding software. The hardware or software includes one or more modules corresponding to the above-described function.

[0015] Thirdly, a computer-readable storage medium is provided that stores instructions which, when executed on a computer, enable the computer to perform the cleaning machine control method of any of the first aspects described above.

[0016] Fourthly, a computer program product containing instructions is provided, which, when run on a computer, enables the computer to execute the cleaning machine control method of any one of the first aspects described above.

[0017] The technical effects of any of the design methods in the second to fourth aspects can be found in the technical effects of different design methods in the first aspect, and will not be repeated here. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the structure of a cleaning machine control system provided in an embodiment of this application; Figure 2 This is an interaction diagram between a multimodal sensing unit and an edge intelligent control unit provided in an embodiment of this application; Figure 3 This is a schematic diagram of a multi-source data preprocessing process provided in an embodiment of this application; Figure 4 This is a schematic diagram of a start / stop control process provided in an embodiment of this application; Figure 5 This is a schematic flowchart of motor parameter control provided in an embodiment of this application; Figure 6 This is a schematic diagram of a fault early warning process provided in an embodiment of this application; Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0019] The following description, in conjunction with the accompanying drawings, illustrates exemplary embodiments of this application, including various details to aid understanding. These embodiments should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope of this application. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0020] The trash rack cleaner includes a drive motor and a cleaning mechanism, which are connected via a transmission mechanism. The cleaner is installed at the trash rack at the inlet of a hydropower station or pumping station. The trash rack is used to intercept floating debris such as branches, weeds, and plastics carried in the water to prevent them from entering the hydropower station or pumping station. When floating debris is intercepted by the trash rack and accumulates in front of it, the flow area of ​​the trash rack decreases, and its flow capacity declines. At this time, the drive motor of the cleaner rotates, driving the cleaning mechanism along the trash rack through the transmission mechanism to remove the accumulated floating debris and restore the flow capacity of the trash rack.

[0021] During the control of the trash rack cleaning machine, it is necessary to determine the degree of blockage in front of the trash rack to decide whether to initiate the cleaning operation. Existing control methods typically use a level difference sensor to collect the level difference before and after the trash rack, and judge the degree of blockage based on this difference. The more floating debris accumulates in front of the trash rack, the smaller the flow area, and the greater the level difference. However, the level difference is easily affected by various factors in actual operating conditions. For example, fluctuations in upstream water flow can cause changes in the water level before the rack, and adjustments to downstream gates can cause changes in the water level after the rack. Furthermore, when floating debris begins to accumulate, the flow area has not yet decreased significantly, and the level difference has not yet reached the activation value, preventing the trash rack cleaning machine from responding promptly. Therefore, under conditions of fluctuating water flow and continuous accumulation of floating debris, relying solely on the level difference is insufficient to stably and accurately reflect the degree of blockage in the trash rack, leading to unreliable start-stop control of the trash rack and a tendency for false starts or missed starts.

[0022] Based on this, this application proposes a control method that simultaneously collects multi-source data such as liquid level difference data before and after the trash rack, flow rate through the rack, and floating object image data in front of the rack. After time alignment and comprehensive judgment of the multi-source data, the method controls the start and stop of the trash rack based on the data of liquid level difference, flow rate through the rack, and floating object image recognition results, thereby achieving accurate start and stop of the trash rack.

[0023] See Figure 1 , Figure 1 This is a schematic diagram of the structure of a cleaning machine control system provided in an embodiment of this application. For example... Figure 1 As shown, the system includes a multimodal sensing unit, an edge intelligent control unit, a variable frequency drive unit, and a remote operation and maintenance unit. The multimodal sensing unit communicates with the edge intelligent control unit, uploading collected liquid level difference data, grid flow data, floating object image data, and motor operating condition data to the edge intelligent control unit. The edge intelligent control unit communicates with both the variable frequency drive unit and the remote operation and maintenance unit. The edge intelligent control unit sends control commands to the variable frequency drive unit, which adjusts the speed and output torque of the cleaning machine's drive motor according to the control commands, thus driving the cleaning machine's drive motor and cleaning mechanism. The remote operation and maintenance unit and the edge intelligent control unit perform bidirectional interaction for status uploading and remote control.

[0024] The multimodal sensing unit includes a liquid level difference sensor, a flow sensor, an industrial vision camera, and a motor condition sensor.

[0025] Liquid level difference sensors and flow sensors are used to collect water flow status data. Specifically, the liquid level difference sensor is installed before and after the trash rack to collect liquid level height data before and after the rack, and calculates the liquid level difference based on the liquid level heights before and after the rack. The liquid level difference refers to the difference between the water level before and after the trash rack. Generally speaking, the larger the liquid level difference, the more severe the blockage of the trash rack. The flow sensor is installed in the pipe before or after the rack to collect flow velocity data and flow area data of the water flowing through the pipe, and calculates the flow rate through the rack based on the flow velocity data and flow area data. The flow rate through the rack refers to the actual flow rate of fluid passing through the trash rack. When the trash rack is blocked, the flow rate through the rack decreases.

[0026] Industrial vision cameras are used to acquire image data of floating objects. Specifically, they are mounted above trash racks to capture images of the water surface within a designated area in front of the racks, thus obtaining image data of floating objects. The industrial vision cameras are equipped with automatic supplemental lighting and wipers. The automatic supplemental lighting is used to provide additional illumination at night or in low-light conditions. The wipers are used to clean the camera lens in rainy or foggy weather.

[0027] Motor condition sensors are used to collect operating parameter data of the drive motor of the cleaning machine. Specifically, these sensors are installed at various points on the drive motor to continuously collect current, temperature, torque, and speed data during machine operation. Motor condition sensors can include current sensors, temperature sensors, torque sensors, and speed sensors. The current sensor collects the motor current. The temperature sensor collects the motor temperature. The torque sensor collects the motor output torque. The speed sensor collects the motor speed. Each sensor transmits the collected condition data to the edge intelligent control unit.

[0028] Each sensor continuously collects data according to its set sampling frequency and transmits the collected data to the edge intelligent control unit through its corresponding data channel. The liquid level difference sensor outputs liquid level difference data, the flow sensor outputs flow rate data, the industrial vision camera outputs floating object image data, and the motor condition sensor group outputs motor condition data including current, temperature, torque, and speed.

[0029] See Figure 2 , Figure 2 This is an interaction diagram between a multimodal sensing unit and an edge intelligent control unit provided in an embodiment of this application. For example... Figure 2As shown, a level difference sensor collects level monitoring data before and after the trash rack. A flow sensor collects channel flow monitoring data to obtain the flow rate through the rack. An industrial vision camera captures images of floating objects in front of the rack. A motor operating condition sensor group collects current, temperature, torque, and speed monitoring data. These various data types are input to the edge intelligent control unit for further processing in the form of level difference data, flow rate through the rack, floating object image data, and motor operating condition data, respectively.

[0030] The edge intelligent control unit includes a data preprocessing module, a fusion decision module, a load regulation module, and a fault early warning module (such as...). Figure 1 (As shown). The data preprocessing module performs filtering and noise reduction, outlier removal, time synchronization, and data alignment on the raw data uploaded by the multimodal sensing unit. The fusion decision module determines the blockage status of the trash rack based on the preprocessed multi-source data and decides whether to start or stop the trash rack cleaning machine based on the judgment result. The load adjustment module dynamically adjusts the operating frequency, output torque, and speed of the variable frequency drive unit after the trash rack cleaning machine starts, based on the load size reflected by the density and volume of the floating debris. This ensures that the variable frequency drive unit drives the trash rack cleaning machine motor with the adjusted parameters, matching the cleaning machine speed and cleaning capacity to the current floating debris. The fault early warning module monitors the operating parameters of the trash rack cleaning machine motor in real time during operation. It compares the monitored data with a pre-established normal parameter range, determines the abnormality level based on the degree of deviation of the actual data from the normal range, and executes the corresponding graded protection action.

[0031] In low-light, rainy, foggy, or nighttime conditions, the image quality of floating objects captured by industrial vision cameras is low, directly affecting the recognition performance of subsequent image recognition models. To address this issue, the data preprocessing module performs image enhancement processing. This includes histogram equalization and dehazing. Histogram equalization redistributes the grayscale values ​​of image pixels to expand the dynamic range and improve contrast. Dehazing estimates and removes fog components from the image to restore clarity, making the edges and textures of the floating objects more distinct.

[0032] Liquid level difference data, flow rate data, floating object image data, and motor operating condition data originate from different sensors, each with its own independent time reference. That is, each sensor records data timestamps based on its own local clock. Without time synchronization, data collected by different sensors at the same physical moment will have different timestamps, resulting in time discrepancies and leading to incorrect association of data from different times during subsequent comprehensive analysis. For the aforementioned multi-source data, the data preprocessing module first performs time synchronization. Time synchronization involves calibrating the timestamps of each sensor's data using NTP (Network Time Protocol), ensuring that all sensor timestamps are based on the same time starting point, eliminating time discrepancies introduced by the asynchronous local clocks of each sensor. After time synchronization, the timestamps of all sensor data are marked under a unified time reference.

[0033] However, different sensors have different sampling frequencies and sampling times. Even if the timestamp reference is consistent, the positions of data points from different sensors at the same physical moment are not the same in their respective data sequences. To address this issue, the data preprocessing module further performs data alignment processing. Data alignment processing refers to extracting data points corresponding to the same physical moment from the synchronized data sequences of each sensor, using a unified time axis as a reference, to obtain a set of multi-source data that are matched in time. For example, the liquid level difference value collected by the liquid level difference sensor at time T, the flow rate value collected by the flow sensor at time T, the image frame collected by the industrial vision camera at time T, and the current and temperature values ​​collected by the motor condition sensor at time T are added to the same set of data, so that subsequent comprehensive judgments can be analyzed based on the complete dataset at the same time.

[0034] See Figure 3 , Figure 3 This is a schematic diagram of a multi-source data preprocessing process provided in an embodiment of this application. For example... Figure 3 As shown, the liquid level difference data, flow rate data, floating object image data, and motor operating condition data need to be filtered for noise reduction and outlier removal first. Then, all types of data are synchronized in time via NTP. The time-synchronized data enters the multi-source time alignment stage, and alignment is completed based on a unified time axis. After alignment, the data is normalized and then output to the fusion decision module.

[0035] The above-mentioned filtering, denoising, time synchronization, time alignment, and normalization processes are as follows: Figure 3 As shown. Figure 3This is a flowchart of multi-source data processing provided in an embodiment of this application. Liquid level difference data, flow rate data, floating object image data, and motor operating condition data need to be filtered for noise reduction and outlier removal first. Then, all types of data are synchronized in time via NTP. The synchronized data is then time-aligned using a unified time axis as a reference. After normalization, the aligned data is output to the fusion decision module.

[0036] It should be noted that after the trash rack is started and in operation, the moving parts of the cleaning mechanism continuously move underwater, agitating the water in front of the trash rack. This agitation causes periodic fluctuations in the water surface in front of the rack, which are related to the movement of the cleaning mechanism. These fluctuations may be superimposed on the liquid level difference data and the flow rate through the rack, causing periodic changes in the liquid level difference and the flow rate through the rack during the operation of the cleaning machine that are unrelated to clogging. In this case, the fluctuations caused by the movement of the cleaning mechanism may be misinterpreted as changes in the degree of clogging.

[0037] To address the above situation, the data preprocessing module performs operational interference compensation on the liquid level difference data and grid flow rate data based on the operating parameters of the cleaning mechanism during the operation of the cleaning machine. The data preprocessing module acquires the current operating frequency and drive motor speed data output by the variable frequency drive unit; the operating frequency and speed reflect the movement rate of the cleaning mechanism. Using the movement cycle corresponding to the operating frequency as a benchmark, the data preprocessing module separates the fluctuation component with the same frequency as this movement cycle from the liquid level difference data and grid flow rate data. This fluctuation component is treated as an interference component caused by the movement of the cleaning mechanism and removed. The liquid level difference data and grid flow rate data after removing the interference component are output to the fusion decision module and the load adjustment module. Through operational interference compensation, the liquid level difference and grid flow rate data used by the fusion decision module and the load adjustment module during the operation of the cleaning machine reflect the true degree of blockage of the screen, no longer including fluctuations introduced by the movement of the cleaning mechanism itself, thus avoiding misjudgments and incorrect parameter adjustments during the operation of the cleaning machine.

[0038] Before the screen cleaner can be started, any one of three conditions—liquid level difference, grid flow ratio, and floating debris coverage—must meet the start-up requirements. The fusion decision module then issues a start-up command to the screen cleaner. In other words, if any one of these three conditions is met, the fusion decision module outputs a start-up command to the screen cleaner.

[0039] The activation condition corresponding to the liquid level difference is: the liquid level difference is greater than or equal to the first liquid level difference value, which is preset.

[0040] The activation condition corresponding to the screen flow rate is: the screen flow rate ratio is less than the first ratio threshold. The screen flow rate ratio refers to the ratio of the actual screen flow rate to the reference flow rate, where the reference flow rate can be the screen flow rate measured under normal operating conditions. When the screen flow rate ratio is less than the first ratio threshold, it indicates that the flow capacity of the screen has been significantly reduced due to blockage, and the screen cleaning machine can be started at this time.

[0041] The activation condition for the floating debris coverage rate is: the floating debris coverage rate is greater than or equal to the first coverage rate value. Floating debris coverage rate refers to the proportion of the water surface area covered by floating debris to the total area of ​​the monitored area, as monitored by the industrial vision camera. The floating debris coverage rate is output by the image recognition model after analyzing the floating debris image. Taking a first coverage rate value of 15% as an example, when the floating debris coverage rate in front of the screen reaches 15% or higher, it indicates a significant accumulation of floating debris in front of the screen, at which point the cleaning machine can be activated.

[0042] The cleaning machine can be started if any one of the three data points mentioned above meets the start-up condition, but all three data points must meet the shutdown condition to shut down the cleaning machine. In other words, the shutdown conditions corresponding to the liquid level difference, the flow rate through the grid, and the floating debris coverage rate must all be satisfied simultaneously before the fusion decision module outputs a shutdown command to the cleaning machine. This ensures that the cleaning machine completes the cleaning operation and avoids premature shutdown.

[0043] The closing condition corresponding to the liquid level difference is: the liquid level difference is less than or equal to the second liquid level difference value, and the second liquid level difference value is lower than the first liquid level difference value.

[0044] The shutdown condition corresponding to the overflow rate is: the overflow rate ratio is greater than or equal to the second ratio threshold. The second ratio threshold is greater than the first ratio threshold. Taking a second ratio threshold of 95% as an example, when the overflow rate has recovered to 95% or more of the baseline flow rate, it indicates that the overflow capacity of the trash rack has been basically restored, and the trash rack can be shut down at this time.

[0045] The shutdown condition corresponding to the floating debris coverage rate is: the floating debris coverage rate is less than or equal to the second coverage rate value. The second coverage rate value is less than the first coverage rate value. Taking a second coverage rate value of 3% as an example, when the floating debris coverage rate drops to 3% or below, it indicates that the floating debris on the water surface in front of the screen has been basically cleared, and the cleaning machine can be turned off at this time.

[0046] The process of determining the above start-up and stop conditions is as follows: Figure 4 As shown. Figure 4This is a schematic flowchart of a start-stop control method provided in an embodiment of this application. In the start-up determination, three conditions are used: the liquid level difference is greater than or equal to a first liquid level difference value, the flow rate ratio through the grid is less than a first ratio threshold, and the floating debris coverage is greater than or equal to a first coverage value. When any one of these conditions is met, the fusion decision module issues a start command for the cleaning machine. In the stop-down determination, three conditions are used: the liquid level difference is less than or equal to a second liquid level difference value, the flow rate ratio through the grid is greater than or equal to a second ratio threshold, and the floating debris coverage is less than or equal to a second coverage value. Only when all three conditions are met does the fusion decision module issue a stop command for the cleaning machine.

[0047] In some embodiments, after outputting the start command for the cleaning machine, the fusion decision module first performs a pre-check operation. The pre-check operation includes: a motor condition sensor collecting static current and temperature data of the drive motor; a fault warning module determining whether the static current and temperature data are within the normal start-up range; and a level difference sensor and a flow sensor collecting current level difference and grid flow rate data; a data preprocessing module determining whether the current data meets the start-up conditions. If the static current and temperature are within the normal start-up range, and the current level difference and grid flow rate data meet the start-up conditions, the fusion decision module sends a start command to the variable frequency drive unit, which then controls the drive motor to start. If the static current or temperature is not within the normal start-up range, or the current level difference and grid flow rate data do not meet the start-up conditions, the fusion decision module may delay for a preset time before sending the start command.

[0048] During the operation of the cleaning machine, an industrial vision camera continuously acquires images of floating objects in front of the screen. After the data preprocessing module enhances the acquired images, an image recognition model performs real-time analysis of the enhanced images. The image recognition model uses a target detection algorithm to detect and classify floating objects in the images, outputting the type, coverage, density, and individual size of the floating objects in each frame. Floating object density refers to the concentration of floating objects per unit area of ​​water surface. Individual size refers to the volume of each floating object. The recognition results are transmitted to the load regulation module in real-time at a set frequency.

[0049] Specifically, an object detection algorithm is used to generate multiple candidate regions on each frame of the image. Features are extracted from each candidate region to determine whether it contains floating objects and the type of the objects. For each detected floating object, the following processing can be performed: the pixel ratio of the bounding box of the floating object in the water surface area is counted to calculate the floating object coverage; the number of targets detected per unit area is counted to determine the floating object density; and the volume of each target is estimated as a single volume based on the pixel size of the bounding box and the pre-calibrated conversion relationship between pixels and actual physical size.

[0050] It should be noted that the monitoring area corresponding to the water surface image captured by the industrial vision camera has a certain size, and the distribution of floating objects within this monitoring area is usually not uniform. When floating objects are concentrated in a localized area of ​​the monitoring area, if the image recognition module only outputs the global floating object coverage rate for the entire monitoring area, this global coverage rate, as the average of the coverage rates of each localized area, will be lower than the actual coverage rate of the localized area where the accumulation occurred. In this case, using the global coverage rate as the basis for start / stop judgments and load adjustments may result in severe localized accumulation being masked by the global average.

[0051] To address the above situation, the image recognition module divides the monitoring area of ​​the water surface in front of the screen into multiple sub-regions, identifies floating objects in each sub-region, and outputs the sub-region coverage, floating object density, and floating object quantity for each sub-region. When determining the operating frequency based on floating object density, the load adjustment module uses the density value of the sub-region with the highest floating object density among the multiple sub-regions as the basis for determining the operating frequency of the variable frequency drive unit. If there are sub-regions with density values ​​exceeding a second density value, the load adjustment module determines the operating frequency according to the density level above the second density value, even if the global average density is lower than that density level. In this way, the cleaning machine operates at a frequency matching the density of locally high-density accumulation areas, avoiding insufficient cleaning of those areas.

[0052] In its startup decision-making process, the fusion decision-making module sets the startup condition corresponding to the floating object coverage rate as follows: There must be a sub-region among multiple sub-regions whose coverage rate is greater than or equal to a first coverage rate value. In other words, the cleaning machine is triggered as soon as the coverage rate of any sub-region exceeds the first coverage value, without waiting for the global coverage rate to exceed it. By using sub-region coverage rate as the startup criterion, the fusion decision-making module can trigger cleaning even when local accumulation occurs before the global average reaches the startup condition, avoiding the omission of local accumulation due to the global average. By dividing the monitoring area in front of the grid into multiple sub-regions and identifying them separately, load adjustment and start / stop decisions are based on sub-region data, preventing severe local blockages from being masked by the global average, improving the ability to identify local accumulation and the targeted nature of cleaning.

[0053] The load adjustment module adjusts the frequency based on the density of floating debris. When the density of floating debris is less than a first density value, the load adjustment module sends a control command for the first frequency to the variable frequency drive unit, controlling the variable frequency drive unit to drive the motor at the first frequency. Taking a first density value of 10% and a first frequency of 30Hz as an example, when the density of floating debris is less than 10%, it indicates that there are fewer floating debris in front of the grid and the load on the cleaning machine is lighter. The variable frequency drive unit drives the motor at a frequency of 30Hz to complete the light cleaning operation with lower speed and power consumption.

[0054] When the density of the floating object is greater than or equal to the first density value and less than or equal to the second density value, the load regulating module sends a control command for the second frequency to the variable frequency drive unit, controlling the variable frequency drive unit to drive the motor at the second frequency. For example, the second frequency is 40Hz.

[0055] When the density of floating debris exceeds the second density value, the load regulation module sends a control command for the variable frequency drive unit at a third frequency, controlling the variable frequency drive unit to drive the motor at the third frequency. Taking a third frequency of 50Hz as an example, when the density of floating debris exceeds 30%, it indicates a large accumulation of floating debris in front of the grid and a heavy load on the cleaning machine. The variable frequency drive unit drives the motor at a frequency of 50Hz to cope with the high-density floating debris conditions with higher speed and greater cleaning capacity.

[0056] In some embodiments, after the image recognition model outputs the floating object density and volume data for the first time, the load adjustment module does not immediately perform parameter adjustment, but first confirms the continuity of the data. The load adjustment module acquires the recognition results of N consecutive frames of images. If at least M frames in the N frames have floating object densities at the same density level, the load adjustment module confirms that the density level is valid, and then determines the corresponding operating frequency based on the density level. If the number of frames at the same density level in the N frames is less than M, it indicates that the distribution of floating objects in front of the grid is uneven or the image recognition results are unstable. The load adjustment module operates at a frequency lower than the frequency corresponding to the current density level, and confirms again in the next recognition cycle until the density level stabilizes, after which it switches to the corresponding frequency.

[0057] Based on the basic operating frequency determined by the density of floating objects, the load adjustment module also determines the presence of large-volume floating objects based on the individual volume data output by the image recognition model. Large-volume floating objects refer to those with a single volume greater than or equal to the specified volume value. For example, with a volume value set to 0.5 cubic meters, if any single floating object reaches a volume of 0.5 cubic meters or more, the load adjustment module determines the presence of a large-volume floating object and adjusts the torque and rotational speed accordingly.

[0058] Upon detecting large floating objects, the load regulation module, in addition to issuing control commands for the basic operating frequency, also issues control commands to perform superimposed adjustments to torque and speed. Specifically, the load regulation module sends torque control commands and speed control commands to the variable frequency drive unit. The torque control command increases torque according to a set torque increase ratio, while the speed control command decreases speed according to a set speed decrease ratio. For example, with a torque increase ratio of 15% and a speed decrease ratio of 10%, the variable frequency drive unit increases the motor's output torque to 115% of the current value while simultaneously reducing the motor speed to 90% of the current value, allowing the cleaning machine to pass through water areas containing large floating objects with stronger traction and lower speed.

[0059] When the load regulation module detects that the amount of floating object has decreased below its value, it sends a recovery command to the variable frequency drive unit. The recovery command is used to restore the motor's torque and speed to the basic operating parameters corresponding to the current density, ensuring that the system returns to normal operating mode after the special operating condition disappears, avoiding unnecessary increases in energy consumption and efficiency losses.

[0060] See Figure 5 , Figure 5 This is a schematic flowchart illustrating motor parameter control according to an embodiment of this application. Figure 5 As shown, after the cleaning machine starts, the image recognition model identifies the density and volume of floating debris in real time. The load adjustment module first selects the corresponding basic operating frequency based on the density of the floating debris: 30Hz when the density is less than 10%, 40Hz when the density is between 10% and 30%, and 50Hz when the density is greater than 30%. After determining the basic operating frequency, it further determines whether there are large-volume floating debris with a volume greater than or equal to 0.5 cubic meters. If large-volume floating debris is present, the control command to increase the torque according to the set torque increase ratio and the control command to decrease the speed according to the set speed decrease ratio are executed, for example, a superimposed adjustment of increasing the motor torque by 15% and decreasing the speed by 10%. If there are no large-volume floating debris, the basic operating parameters are maintained. The cleaning machine continues to operate in the above-mentioned frequency conversion control mode, that is, until a stop command is received.

[0061] The fault early warning module monitors the current, temperature, torque, and speed data of the drive motor in real time during the operation of the cleaning machine. It compares the real-time data with the normal parameter range, determines the anomaly level based on the degree of deviation, and executes corresponding protective actions. Anomaly levels include Level 1, Level 2, and Level 3 warnings.

[0062] See Figure 6 , Figure 6 This is a schematic diagram of a fault early warning process provided in an embodiment of this application. For example... Figure 6 As shown, the fault early warning module continuously monitors the motor's current, temperature, torque, and speed data throughout the entire operation of the cleaning machine, and performs real-time classification judgments on these parameters. If the parameters deviate from the normal range by 10% to 20%, a Level 1 warning is issued and an anomaly log is recorded. If the parameters deviate from the normal range by 20% to 30%, a Level 2 warning is issued and the motor speed is reduced to 30Hz. If the parameters deviate from the normal range by more than 30%, a Level 3 warning is issued and the machine is immediately shut down for protection.

[0063] In some embodiments of this application, the fusion decision module further includes a time-series prediction model, which is used to predict the clogging trend of the trash rack. The time-series prediction model refers to a model that analyzes and predicts time-series data based on a recurrent neural network. The time-series prediction model can be implemented using a long short-term memory (LSTM) neural network. Based on historical time-series data of liquid level difference and flow rate through the rack, the LSTM neural network learns the patterns of change in liquid level difference and flow rate over time, and outputs the probability of the trash rack becoming clogged within a future set time period.

[0064] The input data for the time series prediction model are liquid level difference time series data and grid flow rate time series data for a first preset duration. Taking a first preset duration of 1 hour as an example, the fusion decision module inputs the liquid level difference time series data and grid flow rate time series data for the most recent 1 hour into the trained time series prediction model.

[0065] The time-series prediction model extracts features and models time dependencies from the input time-series data, then outputs the probability of blockage within a second preset time period. Taking a second preset time period of 30 minutes as an example, the model output value is a probability value, representing the likelihood of the trash rack becoming blocked within the next 30 minutes.

[0066] The time-series prediction model is pre-trained based on historical operational data. During the model training phase, sample data can include time-series data of sample liquid level differences and sample flow rates through the screen. The labels corresponding to the sample data can be used to indicate whether blockage occurs within the corresponding time period and the degree of blockage. In addition, the time-series prediction model can continuously optimize itself by collecting new data through incremental learning during actual operation to adapt to changes in the trash rack under different seasons and water flow conditions.

[0067] The edge intelligent control unit connects to the remote operation and maintenance unit via industrial Ethernet or wireless communication networks, uploading the system's operating status and monitoring data to the remote operation and maintenance unit. The remote operation and maintenance unit is used to display the operating status of the cleaning machine control system, configure parameters, query historical data, and perform remote control operations. The remote operation and maintenance unit includes a local touchscreen and a remote terminal. The local touchscreen is for on-site operators. The remote terminal is for remote maintenance personnel.

[0068] A local touchscreen is installed at the control site of the screen cleaning machine and connected to the edge intelligent control unit via a wired connection. The touchscreen displays real-time system operating status information, including monitoring data such as liquid level difference, flow rate through the screen, and floating debris coverage; operating data such as motor current, temperature, torque, and speed; as well as the current operating mode, warning messages, and alarm prompts. On-site operators can use the touchscreen to view system status, query historical data, configure operating parameters, and switch control modes.

[0069] The remote terminal includes web-based and mobile applications, connecting to the edge intelligent control unit via industrial Ethernet or wireless communication networks. The remote terminal is used to view equipment operating status, real-time monitoring data, historical data curves, and alarm information. The edge intelligent control unit periodically uploads system operating data and alarm information to the remote terminal, allowing remote maintenance personnel to view the system's operational status.

[0070] In some embodiments, remote maintenance personnel modify the operating parameters of the edge intelligent control unit via a remote terminal. These operating parameters include a first liquid level difference, a second liquid level difference, a first ratio threshold, a second ratio threshold, a first coverage value, a second coverage value, a first density value, a second density value, a volume value, and the deviation range for each warning level. The remote terminal sends the modified parameters to the edge intelligent control unit in the form of a parameter configuration command via an industrial Ethernet or wireless communication network. Upon receiving the parameter configuration command, the edge intelligent control unit verifies the command, including the permission level of the command's source and the valid range of the parameter values. If the verification passes, the edge intelligent control unit writes the new parameters to its local memory.

[0071] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Figure 7 As shown, the electronic device includes: a memory 701, a transceiver 702, and at least one processor 703.

[0072] The transceiver 702 is used to interact with other devices to send and receive data.

[0073] The memory 701 is used to store computer program code, which includes computer instructions. These computer instructions run in the aforementioned electronic device to implement the method shown in the above-described method embodiments. For example, the memory may include high-speed random access memory (RAM), and may also include non-volatile memory (NVM), such as at least one disk storage device, and may also be a USB flash drive, portable hard drive, read-only memory, disk, or optical disc, etc.

[0074] Processor 703 can be a general-purpose processor, including a Central Processing Unit (CPU), a network processor (NP), etc.; it can also be a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components. Processor 703 can also be other general-purpose processors. The general-purpose processor can be a microprocessor or any conventional processor.

[0075] The memory 701, transceiver 702, and processor 703 are communicatively connected. For example, the memory 701 and transceiver 702 can be connected to the processor 703 via a system bus to complete communication between them. The system bus can be a peripheral component interconnect (PCI) bus, an extended industry standard architecture (EISA) bus, an industry standard architecture (ISA) bus, etc. The system bus can be divided into address bus, data bus, control bus, etc. For ease of representation, only one thick line is used in the figure, but this does not mean that there is only one bus or one type of bus.

[0076] Optionally, the memory 701 can be either independent or integrated with the processor 703. When the memory 701 is configured independently, it is connected to the processor 703 via a system bus.

[0077] This application also provides a chip for executing instructions, which is used to execute the information synchronization method described in the above embodiments.

[0078] This application also provides a computer-readable storage medium storing computer instructions. When these computer instructions are executed by a processor, they are used to implement the technical solution of the information synchronization method described in the above embodiments. Specifically, when the computer instructions are executed by a processor, the electronic device can execute the technical solution of the information synchronization method described in the above embodiments.

[0079] This application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium, and when the at least one processor executes the computer program, it can implement the technical solution of the information synchronization method in the above embodiments.

[0080] The aforementioned computer-readable storage media can be implemented from any type of volatile or non-volatile storage device or a combination thereof, such as Static Random-Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The computer-readable storage media can be any available medium accessible to a general-purpose or special-purpose computer.

[0081] An exemplary computer-readable storage medium is coupled to a processor, enabling the processor to read information from and write information to the storage medium. Of course, the computer-readable storage medium can also be a component of the processor. The processor and the computer-readable storage medium can reside in an application-specific integrated circuit (ASIC). Alternatively, the processor and the computer-readable storage medium can exist as discrete components in an electronic control unit or main control device; this application does not limit this.

[0082] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or modules, and may be electrical, mechanical, or other forms.

[0083] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to implement the solution of this embodiment according to actual needs.

[0084] Furthermore, the functional modules in the various embodiments of this application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The unit composed of the above modules can be implemented in hardware or in the form of hardware plus software functional units.

[0085] The integrated modules described above, implemented as software functional modules, can be stored in a computer-readable storage medium. These software functional modules, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods of the various embodiments of this application.

[0086] It should be understood that the steps of the method disclosed in the embodiments of this application can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules in the processor.

[0087] Those skilled in the art will understand that all or part of the steps of the above-described method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments; and the aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A method for controlling a cleaning machine, characterized in that, The method is applied to a control system, which includes a multimodal sensing unit, an edge intelligent control unit, and a variable frequency drive unit. The edge intelligent control unit includes a data preprocessing module and a multi-source fusion decision module. The multimodal sensing unit collects data on the liquid level difference before and after the trash rack, the flow rate through the rack, and the image data of floating objects in front of the trash rack. The data preprocessing module preprocesses the liquid level difference data, the grid flow rate data, and the floating object image data. The multi-source fusion decision module identifies the floating object coverage rate based on the floating object image data. When one or more of the liquid level difference, the grid flow ratio, and the floating object coverage rate meet the corresponding start-up conditions, it outputs a cleaning machine start-up command to the variable frequency drive unit. When the liquid level difference, the grid flow ratio, and the floating object coverage rate all meet the corresponding shutdown conditions, it outputs a cleaning machine shutdown command to the variable frequency drive unit. The variable frequency drive unit controls the drive motor to start according to the cleaning machine start command, and controls the drive motor to shut down according to the cleaning machine stop command; The activation conditions include the liquid level difference being greater than or equal to a first liquid level difference threshold, the grid flow rate ratio being less than a first ratio threshold, and the floating object coverage being greater than or equal to a first coverage threshold; the deactivation conditions include the liquid level difference being less than or equal to a second liquid level difference threshold, the grid flow rate ratio being greater than or equal to a second ratio threshold, and the floating object coverage being less than or equal to a second coverage threshold. The flow rate ratio is the ratio of the flow rate data to the reference flow rate, where the reference flow rate is the flow rate through the screen when the screen is operating normally. The first liquid level difference threshold is greater than the second liquid level difference threshold, the first ratio threshold is less than the second ratio threshold, and the first coverage threshold is greater than the second coverage threshold.

2. The method according to claim 1, characterized in that, The edge intelligent control unit further includes a load adaptive adjustment module; the method further includes: The multimodal sensing unit acquires the operating condition data of the drive motor of the cleaning machine; After the cleaning machine is started, the load adaptive adjustment module identifies the density and quantity of floating objects based on the floating object image data, and determines the operating frequency of the variable frequency drive unit based on the floating object density; the operating frequency is used by the variable frequency drive unit to control the speed of the drive motor. When the amount of the floating object is greater than or equal to the volume threshold, the load adaptive adjustment module reduces the speed of the drive motor according to the speed reduction ratio and increases the output torque of the drive motor according to the torque increase ratio, and outputs control commands corresponding to the output torque and the speed to the variable frequency drive unit; wherein, the speed reduction ratio and the matrix increase ratio are preset; The variable frequency drive unit adjusts the output torque and speed of the drive motor according to the control command.

3. The method according to claim 2, characterized in that, Determining the operating frequency of the variable frequency drive unit based on the density of the floating object includes: If the density of the floating object is less than a first density value, the operating frequency is determined to be the first frequency. If the density of the floating object is greater than the second density value, the operating frequency is determined to be the third frequency; If the density of the floating object is greater than or equal to the first density value and less than or equal to the second density value, the operating frequency is determined to be the second frequency. Wherein, the first density value is less than the second density value, the first frequency is lower than the second frequency, and the second frequency is lower than the third frequency.

4. The method according to claim 2, characterized in that, The operating data includes the current data of the drive motor, and the edge intelligent control unit further includes a fault early warning module; the method further includes: During the operation of the cleaning machine, the fault early warning module compares the operating data with a pre-established normal parameter range, determines the abnormality level based on the degree of deviation of the actual data from the normal range, and executes the protection action corresponding to the abnormality level.

5. The method according to claim 1, characterized in that, The preprocessing of the liquid level difference data, the flow rate data, and the floating object image data includes: The liquid level difference data and the grid flow rate data are subjected to filtering and noise reduction processing and outlier removal processing; The floating object image data is subjected to histogram equalization and dehazing. The liquid level difference data, the grating flow rate data, and the floating object image data are synchronized using the Network Time Protocol. Time alignment processing is performed on the liquid level difference data, the grid flow rate data, and the floating object image data; The liquid level difference data, the flow rate through the grid, and the floating object image data are normalized.

6. The method according to claim 1, characterized in that, The fusion decision module further includes a time-series prediction model, which is a long short-term memory neural network; the method further includes: The fusion decision module inputs the liquid level difference data and the flow rate data through the screen for a first preset time period into the time-series prediction model to obtain the probability of the screen becoming blocked within a second preset time period in the future. When the blockage probability is greater than or equal to a preset blockage probability threshold, the fusion decision module outputs a start command for the cleaning machine.

7. The method according to claim 1, characterized in that, The multimodal sensing unit includes a liquid level difference sensor, a flow sensor, and an industrial vision camera. The multimodal sensing unit also includes a motor operating condition sensor, which is used to collect operating condition data of the drive motor, including current, temperature, torque, and speed. The motor operating condition sensor includes a current sensor, a temperature sensor, a torque sensor, and a speed sensor, which are used to collect the current, temperature, torque, and speed of the drive motor, respectively. The industrial vision camera is equipped with an automatic fill light and a wiper device. The automatic fill light is used to supplement illumination in low-light conditions, and the wiper device is used to clean the lens of the industrial vision camera.

8. An electronic device, characterized in that, The electronic device includes a memory and a processor, both of which are coupled to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the electronic device to perform the cleaning machine control method as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when run on a computer, causes the computer to perform the cleaning machine control method as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, It includes a computer program, which, when executed by a processor, implements the cleaning machine control method as described in any one of claims 1 to 7.

Images