SBR reactor sludge bulking early warning device and method based on computer vision

Through the early warning device of sludge expansion using computer vision technology in sewage treatment plants, sludge images are collected and processed in real time, solving the problem of difficulty and cost of identification in early warning and control of sludge expansion, and achieving efficient and accurate Sludge expansion monitoring and early warning have improved the operational stability and management efficiency of sewage treatment plants.

CN118908397AActive Publication Date: 2024-11-08BEIJING UNIV OF TECH
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410968464.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-18
Publication Date
2024-11-08
Estimated Expiration
2044-07-18

AI Technical Summary

Technical Problem

In the early warning and control of sludge expansion problems, existing sewage treatment plants have defects such as difficulty in identifying, slow response time, high cost and relying on manual detection, which makes sludge expansion difficult to prevent and inhibit in a timely manner.

Method used

Using an early warning device for sludge expansion based on computer vision technology, by installing a camera and an industrial control machine in the SBR reactor, the apparent and settlement images of the sludge are collected and processed in real time, and combined with data transmission and processing modules and alarm modules, the Real-time early warning and optimization control of sludge expansion reduces the cost of manual monitoring and misjudgment rate.

Benefits of technology

It realizes accurate identification and timely warning of the early characteristics of sludge expansion, reduces the difficulty of sludge expansion recovery, improves the monitoring density and operation stability of sewage treatment plants, reduces energy consumption and operation and maintenance costs, and is in line with 'smart water affairs' development trend.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118908397A_ABST
    Figure CN118908397A_ABST
Patent Text Reader

Abstract

The invention discloses an early warning device and method for sludge bulking of an SBR (Sequencing Batch Reactor) based on computer vision. According to the device, active sludge in each operation cycle of the SBR reactor is monitored in real time by adopting double visual angles, one mode is that sludge apparent images of the SBR reactor are collected, sludge bulking early-stage characteristics are extracted, and then the operation state of the reactor is judged by adopting a pre-constructed sludge bulking early-stage early-warning model; and the other method comprises the following steps of: acquiring a sludge sedimentation image in a sedimentation stage of the SBR reactor, automatically generating a sludge sedimentation curve and a sludge sedimentation rate curve, extracting information, further judging the sedimentation performance of the sludge in the SBR reactor, and alarming about sludge bulking. According to the method, information of the sludge apparent image and the sedimentation image is combined, the early-stage characteristics of sludge bulking are comprehensively analyzed, misjudgment is reduced, the recognition accuracy is improved, and the problems that the sludge bulking phenomenon of a sewage plant suddenly occurs, the follow-up control cost is high, and the recovery time is long are solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to an early warning device and method for sludge bulking in an SBR reactor of a sewage treatment plant, and belongs to the field of optimized operation of sewage treatment. Background Art

[0002] At present, most domestic sewage treatment plants use the activated sludge method to treat sewage. This biological treatment technology has attracted widespread attention for its advantages of low cost, stability and high efficiency. However, the incidence of sludge bulking is high when this method is used to treat sewage. Sludge bulking is one of the more serious abnormal phenomena in the biochemical treatment system. It has a high incidence rate, a wide range, and is difficult to suppress. In severe cases, it causes deterioration of effluent quality, deterioration of sedimentation performance, sludge loss, and even the collapse of the entire sewage treatment system. In the sewage treatment process, timely identification of the occurrence of sludge bulking is a prerequisite for suppressing sludge bulking. Therefore, it is of great significance to seek effective methods for early warning and suppression of sludge bulking.

[0003] At present, a variety of methods for preventing and controlling sludge bulking have been developed. The monitoring of conventional indicators such as dissolved oxygen, pH value, temperature, and effluent quality provides a basis for early warning and timely prevention of activated sludge bulking in sewage treatment plants, but manual detection is complicated, time-consuming, and has a certain lag. The sludge settling property is judged by measuring the sludge settling velocity (SV) and the sludge volume index (SVI), and the measurement process is often time-consuming and laborious. Moreover, SVI can only reflect the settling characteristics of activated sludge after 30 minutes, ignoring the process monitoring of sludge settling performance. The recognition method based on microscope image information can identify sludge bulking under certain circumstances, but this method is highly dependent on the operator and the equipment itself, the operation is complicated, and it does not have the ability of online recognition. The sludge bulking prediction method based on mechanism and data-driven model can achieve accurate description and detection of sludge bulking, but due to the complex nonlinearity of sludge bulking problem, it is highly dependent on data, has poor adaptability, and has a slow response time. Although there are numerous studies on early warning and control of sludge bulking, the lack of key information and the difficulty in capturing early characteristics limit the effectiveness of existing control methods.

[0004] With the advent of the information age, the field of artificial intelligence is developing rapidly. Image recognition technology, as an important branch, is gradually intersecting with the field of sewage treatment. Further promote the application of "smart water" in sewage treatment plants, realize intelligent control of sewage treatment plants, and improve the treatment capacity and management efficiency of sewage treatment plants. As an emerging technology, computer vision can realize automatic analysis and processing of images, improve work efficiency and reduce labor costs. It has been widely used in sewage treatment monitoring systems in recent years.

[0005] The development of an early warning device and method for sludge bulking based on computer vision technology can replace the daily inspections of operation and maintenance personnel and realize unmanned management. When the sludge is about to swell, the system will alarm, conduct a comprehensive analysis of image information and conventional data, and give operation and control suggestions to quickly and efficiently solve the sludge bulking problem, reduce the difficulty of sludge bulking recovery, and achieve "early detection and early intervention". At the same time, the use of low-cost cameras to realize process monitoring of sewage treatment plants is of great significance to improving the monitoring level and operation stability of sewage treatment plants. Summary of the invention

[0006] The present invention proposes an early warning device and method for sludge bulking in SBR reactors based on computer vision technology. The sludge online warning system includes a model warning of sludge surface image and information extraction of sludge settling curve. The comprehensive judgment of the two methods can fully reflect the settling performance of sludge and accurately identify the early characteristics of sludge bulking.

[0007] The objective of the present invention is achieved through the following technical solutions:

[0008] The present invention is based on a traditional activated sludge process and develops a novel early warning system for sludge bulking. The system realizes real-time data collection, processing and graphical presentation by installing hardware such as a display (1), an industrial computer (2), an alarm (3), a data transmission line (4), a top-view near-infrared / visible light binocular camera (5), a side-view near-infrared / visible light binocular camera (6), a telescopic bracket (7), a tripod (8), an LED fill light (9), a luminous scale (10), a pH probe (11), a dissolved oxygen probe (12), a COD probe (13), a photovoltaic panel (14), and a storage battery (15) in an SBR reactor, thereby ultimately realizing real-time early warning and optimized control of sludge bulking in a sewage treatment plant.

[0009] like Figure 1 Shown is a schematic diagram of the device. The device is divided into three modules according to its functions, including image acquisition module, data transmission and processing module, and data visualization and alarm module.

[0010] 1. Image acquisition module:

[0011] (1) In order to monitor in real time whether the sludge in the SBR reactor is about to expand, a near-infrared / visible light binocular camera (5) for collecting sludge surface images is installed on the top of the SBR reactor. An industrial computer (2) is used to issue instructions to call the near-infrared / visible light binocular camera (5) to collect sludge surface images in the aerobic, anoxic and sedimentation stages of the SBR reactor at a high frequency. It can be selected whether to turn on the LED fill light (9) according to the light conditions.

[0012] (2) In order to monitor the settling performance of the SBR reactor sludge in real time, a luminous scale (10) is installed inside the SBR reactor to measure the position of the mud-water interface, and a side-view near-infrared / visible light binocular camera (6) is installed on the side to collect sludge settling images. An industrial computer (2) is used to issue instructions to call the side-view near-infrared / visible light binocular camera (6) to collect sludge settling images during the sedimentation stage of the SBR reactor.

[0013] 2. Data transmission and processing module:

[0014] The data transmission and processing module comprises an industrial computer (2), a data transmission line (4), a pH probe (11), a dissolved oxygen probe (12) and a COD probe (13). The data transmission line (4) is used to transmit the sludge surface image collected by the top-view near-infrared / visible light binocular camera (5) and the sludge sedimentation image collected by the side-view near-infrared / visible light binocular camera (6) to the industrial computer (2). At the same time, the industrial computer (2) is connected to the alarm (3), the LED fill light (9), the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13) to realize the control of each device while obtaining real-time data. Among them, the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13) are used to obtain the reactor operation parameters as the input of the conventional detection parameters.

[0015] (1) Recognition of sludge surface image: The surface image of the sludge in the SBR reactor is collected using a near-infrared / visible light binocular camera (5) that looks down, and the image processing OpenCV module in the Python software pre-installed on the industrial computer (2) is used to pre-process the collected sludge surface image. First, the non-main area is cropped and the image features are enhanced. Computer vision technology is used to extract features from the sludge surface images in the aerobic, anoxic and sedimentation stages, including color, lines, contours or more abstract high-level semantic features, and the accuracy of model recognition is tested; secondly, the pre-built sludge bulking early warning model is called to recognize the surface image of the sludge in the reactor; finally, the recognition result is compared with the normal state to determine whether the SBR reactor is about to expand, and then a real-time alarm is issued for the system that is about to experience sludge bulking. In addition, the model can also realize the range prediction of SVI.

[0016] (2) Identification of sludge sedimentation images: A side-view near-infrared / visible light binocular camera (6) is used to collect sludge sedimentation images during the sedimentation stage of the SBR reactor. The image processing OpenCV module in the Python software pre-installed on the industrial computer (2) is used to perform bilateral filtering, histogram equalization, Otsu algorithm binarization and other image preprocessing on the collected sludge sedimentation images to complete image noise reduction and improve image quality. The image is then refined using the connected region labeling method to identify the mud-water interface and generate a sludge sedimentation curve and a sludge sedimentation rate curve in real time. Finally, the industrial computer (2) is used to extract information from the sedimentation curve to determine the sedimentation performance of the sludge in the SBR reactor and issue a real-time alarm to the system that is about to experience sludge bulking.

[0017] 3. Data visualization and alarm module:

[0018] The data visualization and alarm module includes an electronic display (1) and an alarm (3). The final processing result of the sludge data is presented on the display (1) in the form of a user interface, which displays the sludge swelling state, sludge settling performance and basic operating parameters of the SBR reactor in real time. According to the image characteristics of the early appearance image of sludge swelling and the settling curve, the alarm (3) sounds an alarm when the sludge is about to swell, reminding the operation and maintenance personnel to intervene in time. At the same time, the industrial control computer (2) combines the prior knowledge of sewage treatment to comprehensively analyze the data of the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13), make a preliminary judgment on the cause of the sludge swelling and give operation and control suggestions, so as to achieve early warning and control of sludge swelling.

[0019] In addition, in order to save energy and reduce costs, an external photovoltaic panel (14) and a storage battery (15) are used to supply energy to the entire system. The photovoltaic panel (14) is used to directly convert solar light energy into electrical energy, and the storage battery (15) is used to store and convert electrical energy to provide energy for the system.

[0020] Compared with the existing early warning methods, the present invention has the following advantages:

[0021] 1) This method combines the sludge surface image and sedimentation image information to conduct a comprehensive analysis of the early characteristics of sludge bulking, reduce misjudgment, and improve recognition accuracy. It is expected to solve the problems of sudden sludge bulking in sewage plants, high subsequent control costs, and long recovery time.

[0022] 2) The present invention can automatically generate a sludge settling curve and extract information from it, which makes up for the lack of key information in the settling process, can give an early warning of sludge swelling, reduce the cost of manpower monitoring and increase the monitoring density.

[0023] 3) This device realizes timely early warning and control of sludge bulking through automated intelligent analysis. It is easy to operate and simple to control. Compared with traditional manual inspections and routine testing, it has low costs, high recognition accuracy, strong real-time performance, and does not require manual supervision, which is in line with the development trend of "smart water services".

[0024] 4) The entire system is powered by a combination of an external photovoltaic panel (14) and a storage battery (15), thereby saving energy and reducing costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 It is a schematic diagram of the structure of the device of the present invention;

[0026] Figure 2 This is the working flow chart of the sludge bulking early warning device.

[0027] Figure 3 This is the processing flow chart of sludge surface image.

[0028] Figure 4 Flow chart for processing sludge sedimentation images.

[0029] Figure 5 Heat map of the prediction results of the sludge bulking early warning model.

[0030] Figure 6 Training and validation loss plots for the sludge bulking early warning model.

[0031] Figure 7 This is the training and validation accuracy diagram of the sludge bulking early warning model.

[0032] Figure 8 This is the sedimentation curve and sedimentation rate change diagram within 30 minutes. DETAILED DESCRIPTION

[0034] The operation of the device is divided into the following stages.

[0035] 1. Construction of early test platform for sludge bulking

[0036] Reference Figure 1The top-view near-infrared / visible light binocular camera (5) and the side-view near-infrared / visible light binocular camera (6) of the system are arranged on the top and side of the SBR reactor respectively. Among them, an LED fill light (9) is arranged on the top as a supplementary light source, and a built-in luminous ruler (10) is used to measure the position of the mud-water interface. The top-view near-infrared / visible light binocular camera (5) is fixed at the center position 50 cm above the liquid surface of the SBR reactor by a telescopic bracket (7) to collect the sludge surface image of the SBR reactor; similarly, the side-view near-infrared / visible light binocular camera (6) is installed on the side of the SBR reactor by a tripod (8), 100 cm away from the reactor wall. The built-in WTW3420 portable multi-parameter measuring instrument monitors the DO and pH values ​​in the reactor in real time, and the COD probe (13) is used to detect the content of organic matter. The overall operation of the reactor is controlled by the sewage treatment intelligent control device.

[0037] 2. Construction of the Early Intelligent Perception System for Sludge Bulking

[0038] 1. Collection of sludge image data

[0039] (1) Collection of sludge surface images:

[0040] Using the Python programming language integrated development environment PyCharm pre-installed on the industrial computer (2), a command is sent to turn on the LED fill light (9), and the OpenCV module is used to call the top-view near-infrared / visible light binocular camera (5) fixed on the top of the SBR reactor. The image acquisition frequency is 1 image per second, and the sludge surface images of the anoxic stage, aerobic stage, and sedimentation stage of each cycle of the SBR reactor are collected at a high frequency.

[0041] (2) Collection of sludge sedimentation images:

[0042] The image automatic acquisition program was written in advance using Python pre-installed on the industrial computer (2), and the side-view near-infrared / visible light binocular camera (6) was called to collect sludge sedimentation images in each cycle of the SBR reactor. The image acquisition frequency was set to 1 image / second, and the total acquisition was 30 minutes. The number of cycles was set to ensure that the total number of images collected was 1,800, and the images were named after the time of shooting. After 30 minutes, 1,800 sedimentation images were collected, the camera was disconnected, and the program automatically terminated.

[0043] 2. Data transmission and processing

[0044] The sludge surface image and sludge sedimentation image collected by the top-view near-infrared / visible light binocular camera (5) and the side-view near-infrared / visible light binocular camera (6) are transmitted in real time to the specified folder of the industrial control computer (2) through the data transmission line (4), so as to realize automatic image collection and storage. The image is named after the shooting time, and the data is stored in the set storage path. The image name "24-03-02_09-25-24" indicates that the image was collected at 9:25:24 on March 2, 2024, which is convenient for operation and maintenance personnel to call it later. At the same time, the online monitoring data of the process water quality variables are transmitted to the industrial control computer (2), and the early feature extraction and timely warning of sludge swelling are realized through comprehensive analysis.

[0045] (1) Sludge surface image processing process: refer to Figure 3 The collected sludge surface images were preprocessed using the OpenCV module responsible for image processing in the Python software pre-installed in the industrial computer (2). Each data set includes 4 types of surface images of different sludge bulking stages, each type of image has 1800 images, and is divided into training set and test set in a ratio of 4:1. The sludge bulking early warning model is established using the VGG16 deep learning network. Finally, the surface image of the sludge in the SBR reactor is input into the model, and the output is the sludge bulking stage category, so as to provide a real-time alarm for sludge bulking.

[0046] a. Dataset preprocessing: The size of the collected apparent image is 840*840. The [0:1000,0:1000] in the input layer is modified to [120:810,730:1440], where [120:810] represents the longitudinal size of the image and [730:1440] represents the lateral size of the image. The sludge apparent image is cropped to a similar size.

[0047] b. Dataset division: According to the SVI changes during the operation of the SBR reactor, there are 4 types of sludge bulking stages, namely "normal, about to expand, mild expansion and severe expansion", with 1,800 images in each category. The training set and the validation set are divided in a 4:1 ratio, with 1,440 images in each category used for model training and 360 images used to verify the model accuracy. In order to evaluate the model training effect, the average accuracy (Avg_acc) is used to measure the prediction accuracy of the model. The calculation formula is as follows:

[0048]

[0049] Where ns is the total number of sample categories (there are 4 sludge bulking categories, ns = 4); i is the sample category label i∈[1,4]; ni represents the total number of samples in category i; nii represents the total number of samples in category i with prediction result i, that is, the number of correctly predicted samples in each category.

[0050] c. Model construction: Based on the VGG16 deep learning network, adjust the network structure parameters, set the number of image classification categories, number of iterations, learning rate, batchsize and other indicators, name the model file and specify the model storage location after training, and perform model training and verification. The model will automatically extract features from the sludge surface image, including color, lines, contours, bubbles or more abstract high-level semantic features, and then establish an early warning model for sludge bulking.

[0051] d. Model application: Call the pre-built model to identify the apparent image of the sludge in the reactor, compare the identification result with the normal state, determine whether the SBR reactor is about to expand, and then issue a real-time alarm to the system that is about to experience sludge expansion. In addition, the model can also predict the range of SVI.

[0052] (2) Sludge sedimentation image processing process: refer to Figure 4 , the image of sludge sedimentation collected 30 minutes before the sedimentation stage was preprocessed using the OpenCV module responsible for image processing in the Python software pre-installed in the industrial computer (2). First, the irrelevant background was removed by image cropping and the image of the area where the reactor luminous scale was located in the image was obtained. The position of the reactor was fixed. When the position of the camera for collecting images remained unchanged, the pixel range of the target area in each batch of images was fixed, so the cropping position could remain unchanged after being set by the operator. Then, the image was subjected to bilateral filtering, histogram equalization, Otsu algorithm binarization and other fine processing to suppress noise and improve image quality; finally, the connected area was marked to compare the positional relationship between the mud-water interface and the luminous scale (10), the specific value of the mud-water interface was obtained, and the scale was automatically normalized to generate the sludge sedimentation curve and sludge sedimentation rate curve in real time.

[0053] a. Bilateral filtering: Calculate the spatial distance between the pixel and the surrounding pixels and the gray value similarity between the pixels. The spatial distance weight and the pixel value similarity weight are combined to obtain the final filter weight. For example, in the window centered on q, the weight of a point p in the bilateral filtering process is calculated as follows:

[0054]

[0055] Among them, W q is the weight sum of each pixel value in the filter window, which is used for weight normalization; G r Represents pixel value weight; G s Represents the spatial distance weight.

[0056] b. Histogram equalization: First, grayscale the image, then calculate the histogram of the grayscale image and count the frequency of each grayscale level. For each grayscale level in the histogram, calculate its cumulative distribution function (CDF) to get the proportion of pixels that are lower than or equal to the grayscale level. For a grayscale value i, the calculation formula of CDF is as follows:

[0057]

[0058] Among them, P(j) represents the frequency of pixels with gray value j appearing in the image.

[0059] According to the cumulative distribution function, each gray level is mapped to a new gray level, so that the dense gray distribution part in the gray distribution histogram of the original image is diffused into a uniform distribution in the global range. The gray value of each pixel in the image is replaced with the corresponding mapped gray value to obtain the equalized image.

[0060] c. Binarization using Otsu's algorithm: Calculate the grayscale histogram of the grayscale image and count the number of pixels at each grayscale level. For each possible threshold t, the image is divided into two categories: foreground (target) and background. Then calculate the mean of the pixels in the two categories:

[0061] σ 2 (t)=w1(t)·w2(t)·[m1(t)-m2(t)] 2

[0062] Among them, w1(t) and w2(t) represent the proportion of the number of pixels of the two categories to the total number of pixels; m1(t) and m2(t) represent the average grayscale value of the pixels of the two categories.

[0064] d. Connected region marking: Scan each pixel in the image from top left to bottom right, and divide it into background pixels with grayscale value 0, foreground pixels with grayscale value non-0, and marked pixels in foreground pixels that have been assigned identifiers. By marking the target pixels in the reactor binary image, each individual connected region forms an identified block, and the position and number of scale lines are recorded in the data structure.

[0065] e. Generation of sedimentation curve: After pre-processing and fine processing of the sludge sedimentation image of the reactor, the number of scale lines above the mud-water interface is obtained. According to the actual reactor volume and the interval between each scale line, the product of the number of scale lines and the unit range is subtracted from the total range to obtain the current mud-water interface position value, and the scale is automatically normalized. Finally, a sludge sedimentation curve is drawn with the sedimentation time as the x-axis and the mud-water interface position as the y-axis.

[0066] Finally, the industrial computer (2) extracts information from the sedimentation curve in the following two ways, thereby judging the sedimentation performance of the sludge in the SBR reactor and giving a real-time alarm to the system that is about to experience sludge bulking.

[0067] a. Automatically calculate the integral area of ​​the sludge settling curve and the change in the integral area through the setting program and save them as parameters in real time, and then calculate the incremental ratio of the integral area of ​​the sludge settling curve in adjacent operation cycles and compare it with the set value.

[0068] b. The industrial computer (2) automatically generates a sludge settling rate curve and extracts the maximum settling rate for comparison with the set value.

[0069] 3. Data visualization and online warning

[0070] The sludge online early warning system includes model early warning of sludge surface image and information extraction of sludge settling curve. The comprehensive judgment of the two methods can fully reflect the settling performance of sludge and accurately identify the early characteristics of sludge expansion. When one of the following two methods starts to alarm, it reminds the operation and maintenance personnel to intervene and take measures.

[0071] (1) Early warning of sludge surface image: The VGG16 model has a good effect on the prediction of the sludge surface image of the SBR reactor. The accuracy of the model is more than 99% after 10 iterations, and it can accurately identify the image features of the early stage of sludge bulking. That is, the sludge surface image is input in real time. When the model output is "warning", it indicates that the sludge is about to expand, and the system starts to alarm.

[0072] (2) Early warning of sludge settling images: The industrial computer (2) extracts information from the settling curve in the following two ways. It then determines the settling performance of the sludge in the SBR reactor and issues a real-time alarm for the system that is about to experience sludge bulking. That is, when the settling curve meets one of the following two conditions, the system starts to alarm.

[0073] a. In adjacent operation cycles, when the increment ratio of the integral area of ​​the sedimentation curve within 30 minutes is greater than 5 times, it means that the sludge sedimentation curve has changed significantly, the sludge is about to expand, and the system automatically alarms.

[0074] b. When the maximum sedimentation rate of the sedimentation curve within 30 minutes is less than 2 cm 3 ·s -1 When , it indicates that the sludge settling performance has deteriorated and the sludge is about to swell, the system will automatically alarm.

[0075] The final processing result of the sludge data is presented in real time on the display (1) in the form of a user interface, which visualizes the sludge bulking status, sludge settling performance and basic operating parameters. By comparing the identified image with the normal image, the model can successfully identify the current sludge bulking status. For example, in the normal operation stage, the user interface displays "normal, good sludge settling, SVI: 60-90"; when the sludge is about to swell, the system alarms and provides operation prompts based on the online monitoring data of the process water quality variables. The user interface displays "about to swell, SVI: 90-120, abnormal apparent image / sedimentation curve, low dissolved oxygen", reminding the operation and maintenance personnel to intervene in time; when the sludge has already swelled, the interface prompts can help the operation and maintenance personnel to eliminate the abnormality and provide relevant information for the subsequent sludge bulking recovery.

[0076] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

[0077] 3. Operation and maintenance of sludge bulking early warning and control system

[0078] In order to ensure the long-term stable operation of the early intelligent sensing system for sludge bulking, staff are required to conduct periodic inspections and maintenance of the device; in case of extreme weather (rainstorms, typhoons, hail, etc.), the equipment must be inspected and repaired. The sewage plant operation and maintenance personnel need to regularly check the integrity and operating status of the system equipment and fill in the operation and maintenance records truthfully. The photovoltaic panels (14) require maintenance personnel to clean the dust regularly to ensure the stability of the system's power supply. Check the working performance and operating status of the industrial computer (2) every month, and back up and clean up the data in a timely manner.

Claims

1. A computer vision-based SBR reactor sludge bulking early warning device, characterized in that include: (1) display, (2) industrial computer, (3) alarm, (4) data transmission line, (5) top-view near-infrared / visible light binocular camera, (6) side-view near-infrared / visible light binocular camera, (7) telescopic bracket, (8) tripod, (9) LED fill light, (10) luminous ruler, (11) pH probe, (12) dissolved oxygen probe, (13) COD probe, (14) photovoltaic panel, (15) battery; A dual-viewing angle is used to collect real-time images of the activated sludge in each operation cycle of the SBR reactor; a top-view near-infrared / visible light binocular camera (5) is fixed at the exact center position 50 cm above the liquid surface of the SBR reactor through a telescopic bracket (7) to collect the surface image of the sludge in the SBR reactor; a side-view near-infrared / visible light binocular camera (6) is installed on the side of the reactor through a tripod (8) at a distance of 100 cm from the reactor wall to obtain the sedimentation performance of the sludge inside the SBR reactor; an LED fill light (9) is used to provide auxiliary light when the lighting conditions are poor; a luminous ruler (10) is used to provide a reference for the height of the mud-water interface; the top-view near-infrared / visible light binocular camera (5) and the side-view near-infrared / visible light binocular camera (6) are both controlled by an industrial computer (2), and the collected images are transmitted to the industrial computer (2) in real time through a data transmission line (4).

2. A method for applying the device as claimed in claim 1, characterized in that: The device is divided into three modules according to its functions, including image acquisition module, data transmission and processing module, and data visualization and alarm module; (1) Image acquisition module In order to monitor the properties of activated sludge in an SBR reactor in real time, a top-view near-infrared / visible light binocular camera (5) for collecting sludge surface images and a side-view near-infrared / visible light binocular camera (6) for collecting sludge sedimentation images are respectively installed on the top and side of the SBR reactor; an industrial computer (2) is used to issue instructions to call the top-view near-infrared / visible light binocular camera (5) to collect sludge surface images in the aerobic, anoxic and sedimentation stages of the SBR reactor at a high frequency, and whether to turn on the LED fill light (9) can be selected according to light conditions; at the same time, the side-view near-infrared / visible light binocular camera (6) is called to collect sludge sedimentation images in the sedimentation stage of the SBR reactor; (2) Data transmission and processing module The data transmission and processing module comprises an industrial computer (2), a data transmission line (4), a pH probe (11), a dissolved oxygen probe (12) and a COD probe (13); the data transmission line (4) is used to transmit the sludge surface image collected by the top-view near-infrared / visible light binocular camera (5) and the sludge sedimentation image collected by the side-view near-infrared / visible light binocular camera (6) to the industrial computer (2); at the same time, the industrial computer (2) is connected to the alarm (3), the LED fill light (9), the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13) to realize the control of each device and data acquisition; wherein the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13) are used to obtain the reactor operation parameters as the input of the conventional detection parameters; a. Recognition of sludge surface image b. Use the OpenCV module for image processing in the Python software pre-installed in the industrial computer (2) to pre-process the collected sludge surface image; first, crop the non-main area and enhance the image features, use computer vision technology to extract features from the sludge surface images in the aerobic, anoxic and sedimentation stages, and test the accuracy of model recognition; second, call the pre-built sludge bulking early warning model to recognize the surface image of the sludge in the reactor; finally, compare the recognition result with the normal state to determine whether the SBR reactor is about to expand, and then issue a real-time alarm to the system that is about to experience sludge bulking, that is, input the sludge surface image in real time. When the model output is "warning", it indicates that the sludge is about to expand, and the system starts to alarm; the model is a VGG16 model; c. Recognition of sludge settling images The image processing OpenCV module in the Python software pre-installed on the industrial computer (2) is used to perform image filtering, histogram equalization, and Otsu algorithm binarization on the collected sludge sedimentation image to complete image noise reduction and improve image quality; then the image is refined by the connected region labeling method to identify the mud-water interface and generate the sludge sedimentation curve and sludge sedimentation rate curve in real time; finally, the sedimentation curve is extracted by the industrial computer (2) to determine the sedimentation performance of the sludge in the SBR reactor and to issue a real-time alarm to the system that is about to experience sludge bulking; The system starts to alarm when the settlement curve meets one of the following two conditions; A). When the increment ratio of the integral area of ​​the sedimentation curve within 30 minutes is greater than 5 times in adjacent operation cycles, it means that the sludge sedimentation curve changes significantly, the sludge is about to expand, and the system automatically alarms; B). When the maximum sedimentation rate of the sedimentation curve within 30 minutes is less than 2 cm 3 ·s -1 When the sludge settling performance deteriorates, the sludge is about to swell, and the system automatically alarms; (3) Data visualization and alarm module The data visualization and alarm module includes an alarm (3) and an electronic display (1); the final processing result of the sludge data is presented on the display (1) in the form of a user interface, and the sludge bulking state, sludge settling performance and basic operating parameters of the SBR reactor are displayed in real time; the alarm (3) sounds an alarm when the sludge is about to swell; at the same time, the industrial control computer (2) combines the prior knowledge of sewage treatment to comprehensively analyze the data of the pH probe (11), the dissolved oxygen probe (12) and the COD probe (13), makes a preliminary judgment on the cause of the sludge bulking and gives operation and control suggestions, thereby realizing early warning and control of sludge bulking.

Citation Information

Patent Citations

  • Measurement method for predicting sludge sedimentation ratio of 30-minute sludge sedimentation

    CN114705596A

  • Device and method for intelligently sensing operation state of sedimentation tank of sewage treatment plant based on computer vision technology

    CN116320281A

  • Method and system for detecting granulation degree of aerobic granular sludge

    CN116342505A

  • Sewage treatment sludge settleability detection device and method thereof

    CN117571711A

  • On-line sludge volume index meter using image analysis

    KR100708393B1