CN116797557A - Device for intelligent sensing of anaerobic ammonia oxidation sludge activity - Google Patents
Device for intelligent sensing of anaerobic ammonia oxidation sludge activity Download PDFInfo
- Publication number
- CN116797557A CN116797557A CN202310634829.6A CN202310634829A CN116797557A CN 116797557 A CN116797557 A CN 116797557A CN 202310634829 A CN202310634829 A CN 202310634829A CN 116797557 A CN116797557 A CN 116797557A
- Authority
- CN
- China
- Prior art keywords
- sludge
- machine learning
- learning model
- image
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000010802 sludge Substances 0.000 title claims abstract description 82
- QGZKDVFQNNGYKY-UHFFFAOYSA-N Ammonia Chemical compound N QGZKDVFQNNGYKY-UHFFFAOYSA-N 0.000 title claims abstract description 46
- 230000000694 effects Effects 0.000 title claims abstract description 41
- 229910021529 ammonia Inorganic materials 0.000 title claims abstract description 23
- 230000003647 oxidation Effects 0.000 title claims abstract description 22
- 238000007254 oxidation reaction Methods 0.000 title claims abstract description 22
- 238000010801 machine learning Methods 0.000 claims abstract description 44
- 238000005259 measurement Methods 0.000 claims abstract description 13
- 238000004364 calculation method Methods 0.000 claims abstract description 11
- 238000010191 image analysis Methods 0.000 claims abstract description 10
- 238000012937 correction Methods 0.000 claims abstract description 8
- 230000008447 perception Effects 0.000 claims abstract description 7
- 239000000463 material Substances 0.000 claims abstract description 6
- 230000006872 improvement Effects 0.000 claims abstract description 3
- 238000012549 training Methods 0.000 claims description 21
- 238000005286 illumination Methods 0.000 claims description 13
- 230000002572 peristaltic effect Effects 0.000 claims description 10
- 238000007637 random forest analysis Methods 0.000 claims description 10
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 239000007788 liquid Substances 0.000 claims description 5
- 230000000737 periodic effect Effects 0.000 claims description 4
- 230000006870 function Effects 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 230000008878 coupling Effects 0.000 abstract description 2
- 238000010168 coupling process Methods 0.000 abstract description 2
- 238000005859 coupling reaction Methods 0.000 abstract description 2
- 238000000034 method Methods 0.000 description 26
- 230000008569 process Effects 0.000 description 14
- 238000004422 calculation algorithm Methods 0.000 description 4
- 229910052799 carbon Inorganic materials 0.000 description 4
- 238000003066 decision tree Methods 0.000 description 4
- 238000011065 in-situ storage Methods 0.000 description 4
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000005265 energy consumption Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000007774 longterm Effects 0.000 description 3
- IJGRMHOSHXDMSA-UHFFFAOYSA-N nitrogen Substances N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 239000010865 sewage Substances 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000001651 autotrophic effect Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000012417 linear regression Methods 0.000 description 2
- 229910052757 nitrogen Inorganic materials 0.000 description 2
- QGZKDVFQNNGYKY-UHFFFAOYSA-O Ammonium Chemical compound [NH4+] QGZKDVFQNNGYKY-UHFFFAOYSA-O 0.000 description 1
- IOVCWXUNBOPUCH-UHFFFAOYSA-M Nitrite anion Chemical compound [O-]N=O IOVCWXUNBOPUCH-UHFFFAOYSA-M 0.000 description 1
- MMDJDBSEMBIJBB-UHFFFAOYSA-N [O-][N+]([O-])=O.[O-][N+]([O-])=O.[O-][N+]([O-])=O.[NH6+3] Chemical compound [O-][N+]([O-])=O.[O-][N+]([O-])=O.[O-][N+]([O-])=O.[NH6+3] MMDJDBSEMBIJBB-UHFFFAOYSA-N 0.000 description 1
- 238000005273 aeration Methods 0.000 description 1
- XKMRRTOUMJRJIA-UHFFFAOYSA-N ammonia nh3 Chemical compound N.N XKMRRTOUMJRJIA-UHFFFAOYSA-N 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010170 biological method Methods 0.000 description 1
- 239000006227 byproduct Substances 0.000 description 1
- 230000006652 catabolic pathway Effects 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000007789 gas Substances 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001590 oxidative effect Effects 0.000 description 1
- 231100000614 poison Toxicity 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 239000003440 toxic substance Substances 0.000 description 1
Classifications
-
- C—CHEMISTRY; METALLURGY
- C02—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F—TREATMENT OF WATER, WASTE WATER, SEWAGE, OR SLUDGE
- C02F3/00—Biological treatment of water, waste water, or sewage
- C02F3/28—Anaerobic digestion processes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Microbiology (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Environmental & Geological Engineering (AREA)
- Data Mining & Analysis (AREA)
- Organic Chemistry (AREA)
- Chemical & Material Sciences (AREA)
- Water Supply & Treatment (AREA)
- Hydrology & Water Resources (AREA)
- Artificial Intelligence (AREA)
- Quality & Reliability (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Biodiversity & Conservation Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a device for intelligent perception of anaerobic ammonia oxidation sludge activity, which comprises a sludge acquisition unit, an image analysis unit and an accuracy improvement unit; the sludge collection unit is used for collecting sludge; the image acquisition unit is used for acquiring an image of the sludge; the image analysis unit comprises embedded end equipment and a machine learning model I, and is used for analyzing and outputting a sludge activity result; the accuracy improving unit comprises a calculation power server, a machine learning model II and an accurate activity measurement input interface, the sludge collecting unit provides collecting materials, the image analyzing unit outputs the collecting materials to the machine learning model I for use, and the machine learning model II in the accuracy improving unit carries out self-supervision correction and deviation correction through periodically inputting actual accurate sludge activity values. According to the invention, through the coupling of machine vision and machine learning, the anaerobic ammonia oxidation sludge activity result is intelligently output in real time; meanwhile, continuous iteration prediction precision is achieved based on supervised learning, end deployment is achieved, and stable operation is achieved.
Description
Technical Field
The invention belongs to the technical field of intelligent water treatment, and particularly relates to a device for intelligent perception of anaerobic ammonia oxidation sludge activity.
Background
In recent years, with the real demands of environmental ecology, green development and the like, environmental protection sewage treatment has become a main constraint bottleneck for the high-speed development of various industries. Common sewage treatment methods include chemical, physical, biological, and the like. Among them, biological methods are paid attention to because of few byproducts and the principle of reaction follows natural degradation pathways. In a high-nitrogen low-carbon scene of sewage treatment, the traditional nitrification/denitrification approach needs to complete the conversion from ammonia nitrogen to nitrate nitrogen through aeration, and an organic carbon source needs to be additionally added to complete the denitrification process, so that the energy consumption and the cost of the treatment process are high, and the searching of related technologies for energy conservation and consumption reduction is urgent in the field.
Anaerobic ammonia oxidation is a newer biological denitrification technique that reacts ammonia with nitrite and generates nitrogen primarily through the anaerobic ammonia oxidizing flora. In the process, the flora is strictly anaerobic, so that the energy consumption is saved by more than 60% compared with the whole-process nitrification. Meanwhile, as the flora performs strict autotrophic reaction, an external organic carbon source is not needed, and the carbon source adding cost can be saved by 65%. In addition, the apparent yield coefficient of the sludge in the autotrophic reaction is greatly reduced compared with that in the heterotrophic reaction, so that the yield of the sludge in the whole reaction process is greatly reduced, and the sludge disposal cost is reduced by more than 85%. The advantages make the anaerobic ammonia oxidation technology a star technology in the field of water treatment, and related research and industrialization practical development are rapid.
Anaerobic ammonia oxidation flora activity is the key to successful operation of the anaerobic ammonia oxidation process, and because the flora growth rate is slow, the growth environment is strictly anaerobic and is easily inhibited by external factors (such as toxic substances), so that stable operation of the anaerobic ammonia oxidation process has a plurality of challenges. Therefore, the method realizes quick perception of the activity change of anaerobic ammonia oxidation flora, and combines the optimization and adjustment of the operation parameters of the process in combination with the water inlet load, the water quality characteristics and the like, thereby being the key for realizing the long-term stable operation of the process. In the prior art, the collection of anaerobic ammonium oxidation sludge and the measurement of gas production over a period of time is a common activity measurement method. However, since a certain time is required for the whole measurement period and the measurement process is not in situ in the culture environment, there is a time lag between the measurement activity (generally referred to as specific activity) and the actual sludge sampling time activity. In the long-term operation process, the accumulated difference is easy to cause operation parameter adjustment distortion, and finally causes process instability or breakdown. In conclusion, the rapid in-situ sensing of anaerobic ammonia oxidation sludge activity is one of the core technical problems to be solved in realizing long-term stable operation of the process.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention aims to provide a device for intelligent perception of anaerobic ammonia oxidation sludge activity. The anaerobic ammonia oxidation sludge image is obtained in situ through the device, machine learning model analysis is carried out, and the sludge activity is accurately predicted, so that a perception layer support is provided for automatic intelligent digital stable operation of the anaerobic ammonia oxidation process.
The aim of the invention can be achieved by the following technical scheme:
the device for intelligent perception of anaerobic ammonia oxidation sludge activity comprises a sludge acquisition unit, an image analysis unit and an accuracy improvement unit; the sludge collecting unit is used for collecting sludge; the image acquisition unit is used for acquiring images of the sludge; the image analysis unit comprises embedded end equipment and a machine learning model I, and is used for analyzing and outputting a sludge activity result; the accuracy improving unit comprises a calculation server, a machine learning model II and an accurate activity measurement input interface, the sludge collecting unit provides collecting materials, the image analyzing unit outputs the collecting materials to the machine learning model I for use, and the machine learning model II in the accuracy improving unit carries out self supervision correction and deviation correction through periodically inputting actual measurement accurate sludge activity values.
The sludge collecting unit comprises a sludge flowing groove, a peristaltic pump, an electromagnetic check valve and an ultrasonic liquid level meter; the peristaltic pump sucks the sludge into the sludge and flows through the groove, after the suction quantity triggers the ultrasonic liquid level meter to set a threshold value, the electromagnetic check valves at the two ends of the inlet and the outlet are automatically closed, and meanwhile, the peristaltic pump is also closed.
The image acquisition unit comprises a CCD camera, an illumination sensor, a strip light source and a light source driver; after sludge is sucked into the sludge flowing through the tank, the illumination sensor detects the illumination of the periphery of the sludge flowing through the tank, whether the strip-shaped light source is started or not is determined according to a set illumination threshold, the intensity of the strip-shaped light source is realized by adjusting the PWM duty ratio driven by the light source, and after the set illumination is reached, the CCD camera is started and captures a sludge image.
The machine learning model II uses image preprocessing and color space conversion (RGB space conversion into HSV space) as independent variables, sludge actual activity as an objective function, and training is carried out through polynomial, random forest and XGBOOST 3 mathematical relations respectively.
And 3 mathematical relations obtained by using the machine learning model II are used as a primary learner, and training results are output after linear integration by using Stacking.
And the primary learner result of the machine learning II is periodically replaced and then is output to the machine learning model I, wherein the periodic replacement time is 45-60 d/time.
The beneficial effects of the invention are as follows:
(1) By coupling machine vision and machine learning, the anaerobic ammonia oxidation sludge activity result is intelligently output in real time, so that time lag under the measurement of the traditional method is avoided;
(2) The accuracy improving unit continuously and iteratively improves the accuracy of the machine learning I and II models based on supervised learning through periodic accurate activity detection.
(3) The in-situ detection device has high automation and intelligent degree, can realize end deployment, and can strongly support the digital intelligent operation of the anaerobic ammonia oxidation process.
Drawings
FIG. 1 is a flow chart diagram of an anaerobic ammonia oxidation sludge activity intelligent sensing device and a using method thereof.
Fig. 2 is a machine learning model I and II training flowchart.
Detailed Description
The invention is further illustrated in the following figures and examples.
As shown in FIG. 1, the device for intelligent sensing of anaerobic ammonia oxidation sludge activity comprises a sludge acquisition unit, an image analysis unit and an accuracy analysis unit.
The sludge collecting unit comprises a sludge flowing groove, a peristaltic pump, an electromagnetic check valve and an ultrasonic liquid level meter; the peristaltic pump sucks the sludge into the sludge and flows through the groove, after the suction quantity triggers the ultrasonic liquid level meter to set a threshold value, the electromagnetic check valves at the two ends of the inlet and the outlet are automatically closed, and meanwhile, the high-flow peristaltic pump is also closed.
In this example, the sludge flow tank has a diameter of 110mm and a length of 80mm. The suction amount of the sludge is 75+/-1% of the volume of the whole sludge flowing through the tank, after the set threshold is reached, the electromagnetic check valves at the two ends of the inlet and the outlet are automatically closed, and meanwhile, the large-flow peristaltic pump is also closed.
The image acquisition unit comprises a CCD camera, an illumination sensor, a strip light source and a light source driver; the working flow is that when the sludge is sucked into the sludge flowing through the groove, the illumination sensor detects the illuminance of the position + -12 mm around the sludge flowing through the groove, the illuminance range is set to be 2000-2500lx, meanwhile, whether the strip-shaped light source is started or not is determined according to the set illuminance threshold, and the intensity of the strip-shaped light source is realized by adjusting the PWM duty ratio of the light source driver. After the set illumination is reached, the CCD camera is started and shoots the image of the sludge. The CCD camera pixel requirement is 800w or more.
The image analysis unit comprises embedded end equipment and a machine learning model I; the image shot by the workflow is calculated through a machine learning model I and embedded end equipment, and a sludge activity result is directly output. The embedded end equipment of the image analysis unit can use the embedded equipment such as RK3588, RK3399, STM32-H7 and the like, requires more than 1TOPS, and obtains a sludge activity prediction result by operating a machine learning model I after the linux is deployed.
The accuracy improving unit comprises a calculation force server, a machine learning model II and a precise activity measurement input interface. The accuracy enhancing unit includes a GPU computing force server which may use Jetson TX2, jetson Xavier NX or NVIDIA A100, requiring computing force of 5 TOPS or more. And (3) training by operating the machine learning model II after the linux is deployed and periodically providing the training result to the primary learner of the machine learning model I, wherein the periodic replacement time is 40-60 d/time. In addition, the sludge activity measurement result is periodically input by manpower for supervised learning correction.
The specific implementation method of the machine learning model II is as follows: and processing the image acquired by the image acquisition unit by using Gaussian filtering of a 3x3 Gaussian kernel to remove noise in the image. And detecting a highlight region by using a python language and an OpenCV module, and then processing the highlight region by using an image restoration and image enhancement method to remove reflection caused by shooting conditions in an image. And detecting and distinguishing the foreground and the background by using an OpenCV module, and removing irrelevant backgrounds in the photo by using an image segmentation algorithm. The OpenCV is used for converting the image which is denoised, light-reflected and background-removed from the original RGB image into the HSV image. And traversing the pixel values of the full image of the HSV image, and carrying out weighted average on the pixel values of the points according to the morphological characteristics to obtain the pixel average value of H, S, V three channels. And the data is divided into a training set and a prediction set.
After the above processing is completed, training is performed by using polynomial, random forest and XGBOOST 3 mathematical relations respectively:
(1) The polynomial calculation method comprises the following steps: based on PyTorch and Sklearn, taking H, S, V value of each image as independent variable, carrying out linear and polynomial regression on the activity of a sample corresponding to the image as dependent variable, and determining polynomial degree by comparing a prediction result on a prediction set with an actual result R party, MSE and the like and adjusting regression parameters.
(2) The random forest calculation method comprises the following steps: based on the PyTorch and Sklearn, a random forest is established, a cross 5-fold verification calculation is used for representing training results, parameters such as an R party, MSE and the like are used as the representation, and after multiple times of training, the super parameters of the random forest, such as N_identifiers (the number of decision trees in the random forest), max_depth (the maximum depth of the decision trees) and the like, are adjusted gradually. And determining the number and the maximum depth of decision trees under the best regression result of the random forest.
(3) The XGBOOST calculation method comprises the following steps: integrating the random forest decision tree by using an XGBOOST algorithm, using cross 5-fold verification calculation to represent a training result, using parameters such as an R party, MSE and the like as the representation, synthesizing the influence of the XGBOOST super-parameters on the training result, and obtaining the optimal super-parameters after multiple adjustment.
The specific implementation method of the machine learning model I is as follows: the method comprises the steps of integrating 3 mathematical relations in a machine learning model II by using a Stacking algorithm, using the algorithm as a primary learner, using only a simple linear regression as a secondary learner to prevent overfitting, training by using a training set to obtain a plurality of primary learners, and using the predicted value of the trained primary learner as the training input value of the secondary learner to train the secondary learner. The training results are characterized by using cross 5-fold verification, and the super parameters of the primary learner are adjusted through the final results. After multiple times of training, taking a model under the optimal R-side and MAE values as a machine learning model I at the current stage, and outputting a sludge activity result in real time.
In actual training, after photographing a sludge image, taking a H, S, V value of the sludge image as a characteristic, taking the actually measured sludge activity as a label, respectively fitting mathematical relations between the characteristic and the label on a training set in a machine learning model II by using three calculation methods of polynomial, random forest and XGBOOST to obtain three models, and introducing a prediction set into the models to predict and evaluate a prediction result. If the result is insufficient to meet the requirement, the machine learning model II automatically adjusts the parameters of the calculation method until the estimation of the prediction result is sufficient to meet the actual requirement, and the finally trained model is provided to the machine learning model I as a primary learner.
In the machine learning model I, H, S, V values of the sludge image are input into a primary learner to obtain a result serving as a characteristic, and the actually measured sludge activity is used as a label. Linear regression is used as a secondary learner to fit the mathematical relationship between the features and the labels. And parameters are adjusted by adopting a method similar to that in the machine learning model II, so that the machine learning model I which can be put into use is finally obtained.
In actual measurement, the sludge image shot by the CCD camera is guided into an image analysis unit, is denoised through Gaussian filtering, and is converted into an HSV image after reflection and background removal by an OpenCV module. And reading pixel values of the HSV image, and importing H, S, V values read out of the image into a machine learning model I to obtain a sludge activity prediction result.
In supervised learning, sludge images and activity data different from the pre-training period are input by correction, i.e., periodically. Training and learning are performed by using the same method in the machine learning model II so as to correct errors possibly caused by interference and the like in the use process of the model.
The embodiments in the foregoing description may be further combined or replaced, and the embodiments are merely illustrative of the preferred embodiments of the present invention and are not intended to limit the spirit and scope of the present invention, and various changes and modifications made by those skilled in the art to which the present invention pertains without departing from the spirit of the present invention. The scope of the invention is given by the appended claims and any equivalents thereof.
Claims (6)
1. The device for intelligent perception of anaerobic ammonia oxidation sludge activity is characterized by comprising a sludge acquisition unit, an image analysis unit and an accuracy improvement unit; the sludge collecting unit is used for collecting sludge; the image acquisition unit is used for acquiring images of the sludge; the image analysis unit comprises embedded end equipment and a machine learning model I, and is used for analyzing and outputting a sludge activity result; the accuracy improving unit comprises a calculation server, a machine learning model II and an accurate activity measurement input interface, the sludge collecting unit provides collecting materials, the image analyzing unit outputs the collecting materials to the machine learning model I for use, and the machine learning model II in the accuracy improving unit carries out self supervision correction and deviation correction through periodically inputting actual measurement accurate sludge activity values.
2. The apparatus of claim 1, wherein the sludge collection unit comprises a sludge flow trough, a peristaltic pump, an electromagnetic check valve, and an ultrasonic level gauge; the peristaltic pump sucks the sludge into the sludge and flows through the groove, after the suction quantity triggers the ultrasonic liquid level meter to set a threshold value, the electromagnetic check valves at the two ends of the inlet and the outlet are automatically closed, and meanwhile, the peristaltic pump is also closed.
3. The device according to claim 1, wherein the image acquisition unit comprises a CCD camera, an illumination sensor, a strip light source and a light source driver; after sludge is sucked into the sludge flowing through the tank, the illumination sensor detects the illumination of the periphery of the sludge flowing through the tank, whether the strip-shaped light source is started or not is determined according to a set illumination threshold, the intensity of the strip-shaped light source is realized by adjusting the PWM duty ratio driven by the light source, and after the set illumination is reached, the CCD camera is started and captures a sludge image.
4. The apparatus of claim 1, wherein the machine learning model II is configured to use image preprocessing and color space conversion to independent variables, and the actual sludge activity is an objective function, and is trained by polynomial, random forest, XGBOOST 3 mathematical relations, respectively.
5. The apparatus of claim 1, wherein the machine learning model I uses 3 mathematical relationships obtained by the machine learning model II as a primary learner and uses Stacking to output training results based on linear integration.
6. The apparatus of claim 1, wherein the primary learner result of the machine learning II is periodically replaced and output to the machine learning model I, and the periodic replacement time is 45-60 d/time.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310634829.6A CN116797557A (en) | 2023-05-31 | 2023-05-31 | Device for intelligent sensing of anaerobic ammonia oxidation sludge activity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310634829.6A CN116797557A (en) | 2023-05-31 | 2023-05-31 | Device for intelligent sensing of anaerobic ammonia oxidation sludge activity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116797557A true CN116797557A (en) | 2023-09-22 |
Family
ID=88047342
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310634829.6A Pending CN116797557A (en) | 2023-05-31 | 2023-05-31 | Device for intelligent sensing of anaerobic ammonia oxidation sludge activity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116797557A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106769893A (en) * | 2016-11-22 | 2017-05-31 | 西安建筑科技大学 | A kind of characterizing method of the sludge microbe COMMUNITY CHARACTERISTICS based on color space |
| CN110826611A (en) * | 2019-10-30 | 2020-02-21 | 华南理工大学 | Stacking sewage treatment fault diagnosis method based on weighted integration of multiple meta-classifiers |
| EP3842693A1 (en) * | 2018-08-23 | 2021-06-30 | Ebara Environmental Plant Co., Ltd. | Information processing device, information processing program, and information processing method |
| CN114459970A (en) * | 2022-01-26 | 2022-05-10 | 海若斯（北京）环境科技有限公司 | Device, method and system for automatically measuring sludge settlement ratio |
| CN115019095A (en) * | 2022-06-06 | 2022-09-06 | 哈尔滨工业大学（深圳） | Active sludge state online monitoring method based on convolutional neural network |
| CN115273078A (en) * | 2022-09-30 | 2022-11-01 | 南通炜秀环境技术服务有限公司 | Sewage treatment method and system based on image data |
-
2023
- 2023-05-31 CN CN202310634829.6A patent/CN116797557A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106769893A (en) * | 2016-11-22 | 2017-05-31 | 西安建筑科技大学 | A kind of characterizing method of the sludge microbe COMMUNITY CHARACTERISTICS based on color space |
| EP3842693A1 (en) * | 2018-08-23 | 2021-06-30 | Ebara Environmental Plant Co., Ltd. | Information processing device, information processing program, and information processing method |
| CN110826611A (en) * | 2019-10-30 | 2020-02-21 | 华南理工大学 | Stacking sewage treatment fault diagnosis method based on weighted integration of multiple meta-classifiers |
| CN114459970A (en) * | 2022-01-26 | 2022-05-10 | 海若斯（北京）环境科技有限公司 | Device, method and system for automatically measuring sludge settlement ratio |
| CN115019095A (en) * | 2022-06-06 | 2022-09-06 | 哈尔滨工业大学（深圳） | Active sludge state online monitoring method based on convolutional neural network |
| CN115273078A (en) * | 2022-09-30 | 2022-11-01 | 南通炜秀环境技术服务有限公司 | Sewage treatment method and system based on image data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110543878B (en) | Pointer instrument reading identification method based on neural network | |
| CN111127429B (en) | Water conservancy system pipe thread defect detection method based on self-training deep neural network | |
| CN109919012A (en) | A kind of indicative microorganism image-recognizing method of sewage treatment based on convolutional neural networks | |
| CN111950812B (en) | Method and device for automatically identifying and predicting rainfall | |
| CN111192236B (en) | SVI measurement method based on activated sludge phase-contrast microscopic image | |
| CN111562242A (en) | Method for quickly identifying source of overflowing sewage in rainy days of urban drainage system | |
| US10788473B2 (en) | Computing system for detecting total phosphorus in effluent using data driven A2/O process | |
| CN113313109A (en) | Semi-quantitative analysis method of fluorescence immunochromatographic test paper | |
| CN101976343B (en) | Technology and device for identifying diatom in lake sediments | |
| CN113902914B (en) | Automatic water meter reading identification method based on microcontroller and convolutional neural network | |
| CN116797557A (en) | Device for intelligent sensing of anaerobic ammonia oxidation sludge activity | |
| CN117171660A (en) | Microorganism repair state monitoring method, system and medium based on support vector machine | |
| CN114663783B (en) | Remote sensing identification method for water body pollution of river entering sea based on machine learning | |
| CN108845075B (en) | Compost maturity real-time prediction method based on deep learning network | |
| CN114693493B (en) | IoT-based polluted river water ecological restoration system | |
| CN116343205A (en) | Automatic labeling method for fluorescence-bright field microscopic image of planktonic algae cells | |
| CN112931309B (en) | Method and system for monitoring fish proliferation and releasing direction | |
| CN104463898A (en) | Image identification counting system for logistics platform | |
| CN110097552A (en) | A kind of automatic division method of mouse prefrontal lobe neuron two-photon fluorescence image | |
| CN114049370A (en) | Intelligent mobile terminal colored solution concentration detection method based on deep learning method | |
| Liping et al. | Prediction and Realization of DO in sewage Treatment based on Machine Vision and BP Neural Network | |
| CN114149076A (en) | Intelligent debugging system of anaerobic ammonia oxidation sewage treatment system | |
| TW202226057A (en) | Water pollution level automatic identification model establishment system and water pollution level identification system and method wherein the establishment system includes a sample input end and an identification processing end | |
| CN111027852B (en) | Method for analyzing wind power generation benefits based on big data processing and grey theory | |
| CN117237773B (en) | Sewage purification analysis method and system based on multiple modes |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |