Organic biochemical reaction bacteria species feeding guidance method and system

Abstract

The application discloses a kind of organic biochemical reaction bacterial species feeding guidance method and system, the method includes: collecting historical water quality monitoring data and carrying out data cleaning to part of data;Establish mechanism water quality modeling rating model, the parameters of mechanism water quality modeling rating model are adjusted by the historical water quality monitoring data after cleaning;Establish expert database based on mechanism water quality modeling rating model, for simulating the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment bacterial species feeding quantity and water quality standard time;Through a variety of AI models, establish AI rating modeling model, train the AI rating modeling model in expert database;Through the sewage treatment bacterial species feeding quantity guidance of trained AI rating modeling model in sewage treatment field edge end.The application establishes expert database and fuses multiple-source heterogeneous model to carry out organic biochemical reaction bacterial species feeding guidance, can guide sewage treatment bacterial species feeding quantity in real time in field end, more with practical guiding function.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of water resource monitoring and management, and particularly relates to an organic biochemical reaction bacteria species feeding guidance method and system. BACKGROUND

[0002] In a sewage environment, dissolved organic matter can cause water quality problems such as color, odor, turbidity, and also plays an important role in the migration and transformation of water body micro-pollutants. Dissolved organic matter is a substance that is often concerned in sewage treatment, and the treatment of these substances is particularly important in the sewage treatment process.

[0003] The conventional method for treating dissolved organic matter in sewage is the biological aerated filter method, which is also recognized as the best method for treating dissolved organic matter. Before and after sewage treatment, various water quality indicators are usually monitored, including BOD water quality indicator detection principle and biological aerated filter method principle. The BOD and other comprehensive organic matter monitoring indicators can also guide the state of the biological aerated filter. These methods often have a lag, which is not conducive to rapid sewage treatment.

[0004] Some methods for predicting and treating related sewage indicators by intelligent algorithms have appeared in the prior art, but most of them cannot produce actual effects in specific applications, the input parameters are not determined, the model is easy to distort, the sewage treatment time cannot be effectively controlled, and the actual application conditions are basically not met. SUMMARY

[0005] Therefore, the application provides an organic biochemical reaction bacteria species feeding guidance method and system to solve the problem that the prior art cannot guide the organic biochemical reaction bacteria species feeding amount in real time.

[0006] In a first aspect, the application discloses an organic biochemical reaction bacteria species feeding guidance method, which comprises the following steps:

[0007] Collecting historical water quality monitoring data and performing data cleaning on part of the data;

[0008] Establishing a mechanism water quality modeling rating model, and adjusting the parameters of the mechanism water quality modeling rating model through the cleaned historical water quality monitoring data;

[0009] Establishing an expert database based on the mechanism water quality modeling rating model, which is used to simulate the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment bacteria species feeding amounts and water quality standard reaching times;

[0010] Training an AI rating modeling model through a plurality of AI models, taking different water quality situations, hydrodynamic situations and corresponding water quality standard reaching times in the expert database as inputs, and taking corresponding sewage treatment bacteria species feeding amounts as outputs to train the AI rating modeling model;

[0011] The sewage treatment bacteria dosage guidance is performed on the sewage treatment site edge end through the trained AI rating modeling model.

[0012] Preferably, the historical water quality monitoring data includes historical monitoring time series data of multiple water quality indicators and corresponding historical sewage treatment bacteria dosage data, corresponding hydrodynamic historical monitoring time series data, and corresponding water quality indicator historical degradation process data; the multiple water quality indicators include but are not limited to COD, dissolved oxygen, UV254, and BOD; and the hydrodynamic historical monitoring time series data includes but is not limited to water level, flow rate, and flow cross-section data.

[0013] Preferably, the data cleaning specifically includes cleaning the historical monitoring time series data of multiple water quality indicators and the hydrodynamic historical monitoring time series data through a data cleaning tool, which includes but is not limited to Kalman filtering, filtering algorithm, fuzzy theory, and principal component analysis.

[0014] Preferably, the parameter setting adjustment of the mechanism water quality modeling rating model based on the cleaned historical water quality monitoring data specifically includes:

[0015] The mechanism water quality modeling rating model includes but is not limited to a water quality complete mixing model, CE-QUAL-W2, EFDC, WASP, DELFT3D, or HEC-RAS.

[0016] The cleaned multiple water quality historical monitoring time series data, corresponding historical sewage treatment bacteria dosage data, and corresponding cleaned hydrodynamic historical monitoring time series data are input into the mechanism water quality modeling rating model to obtain water quality indicator simulation degradation process data output by the mechanism water quality modeling rating model.

[0017] The parameter setting of the mechanism water quality modeling rating model is adjusted through a water quality indicator historical degradation process data and water quality indicator simulation degradation process rating verification cycle until the model reaches a preset standard efficiency coefficient.

[0018] Preferably, the expert database based on the mechanism water quality modeling rating model is used to simulate the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment bacteria dosage, and water quality standard reaching time.

[0019] The solubility of multiple water quality indicators under different water quality situations, sewage treatment bacteria dosage under different sewage treatment bacteria dosage situations, and situation parameters under different hydrodynamic situations are respectively set.

[0020] Input the solubility of multiple water quality indicators under different water quality scenarios, the sewage treatment strain input amount under different sewage treatment strain input scenarios, and the situation parameters under different hydrodynamic scenarios into the mechanism water quality modeling calibrated model after adjusting the parameters, and output the water quality indicator degradation process data under the corresponding scenarios.

[0021] According to the water quality standard time under the corresponding scenario, the corresponding relationship between the different water quality scenarios, the hydrodynamic scenarios, the sewage treatment strain input amount, and the water quality standard time is obtained.

[0022] Preferably, the AI calibration modeling model is established by multiple AI models, and the different water quality scenarios, the hydrodynamic scenarios, and the corresponding water quality standard time in the expert library are input, and the corresponding sewage treatment strain input amount is output, and the AI calibration modeling model is trained, which specifically includes:

[0023] An AI calibration modeling model is obtained by multiple AI models, and the multiple AI models include but are not limited to support vector machine, K-nearest neighbor method, stochastic gradient descent, multivariate linear regression, multilayer perception, decision tree, back propagation neural network, and radial basis function network.

[0024] The different water quality scenarios, the hydrodynamic scenarios, and the corresponding water quality standard time in the expert library are input into each AI model, the sewage treatment strain input simulation quantity is calculated by the set average method, the sewage treatment strain input amount is verified in the calibration verification cycle process according to the sewage treatment strain input simulation quantity and the corresponding sewage treatment strain input amount in the expert library, and the parameter setting of the AI calibration modeling model is adjusted according to the corresponding sewage treatment strain input amount in the expert library.

[0025] Preferably, the sewage treatment strain input amount guidance is performed at the edge of the sewage treatment site by the trained AI calibration modeling model, which specifically includes:

[0026] The trained AI calibration modeling model is deployed on each sewage treatment site edge as an edge computing AI model.

[0027] Real-time monitoring time series data of multiple water quality indicators and hydrodynamic real-time monitoring time series data are obtained at the edge, and data cleaning is performed;

[0028] The expected water quality standard time is set;

[0029] The cleaned real-time monitoring time series data of multiple water quality indicators, the hydrodynamic real-time monitoring time series data, and the expected water quality standard time are input into the edge computing AI model, and the sewage treatment strain input amount real-time guidance value is output.

[0030] The second aspect of the present application discloses an organic biochemical reaction strain feeding guidance system, the system comprises:

[0031] The data collection module is used for collecting historical water quality monitoring data and performing data cleaning on part of the data.

[0032] The parameter adjustment module is used for establishing a mechanism water quality modeling calibration model, and adjusting parameters of the mechanism water quality modeling calibration model through the cleaned historical water quality monitoring data.

[0033] The expert library establishment module is used for establishing an expert library based on the mechanism water quality modeling calibration model, for simulating corresponding relationships between different water quality situations, hydrodynamic situations, sewage treatment strain feeding amounts and water quality standard reaching times.

[0034] The model construction module is used for establishing an AI calibration modeling model through multiple AI models, taking different water quality situations, hydrodynamic situations and corresponding water quality standard reaching times in the expert library as inputs and corresponding sewage treatment strain feeding amounts as outputs to train the AI calibration modeling model.

[0035] The real-time guidance module is used for guiding sewage treatment strain feeding amounts at a sewage treatment field edge end through the trained AI calibration modeling model.

[0036] The third aspect of the present application discloses an electronic device, comprising at least one processor, at least one memory, a communication interface and a bus; wherein the processor, the memory and the communication interface complete mutual communication through the bus; the memory stores program instructions executable by the processor, and the processor invokes the program instructions to implement the method according to the first aspect of the present application.

[0037] The fourth aspect of the present application discloses a computer readable storage medium, which stores computer instructions, and the computer instructions enable a computer to implement the method according to the first aspect of the present application.

[0038] The present application has the following beneficial effects relative to the prior art:

[0039] 1) The present application adds a data cleaning tool for water body monitoring data, has the ability to deeply explore meaningful information of data, improves the model fault tolerance, establishes a mechanism water quality modeling calibration model, and uses the mechanism water quality modeling calibration model for digital twinning to simulate corresponding relationships between different water quality situations, hydrodynamic situations, sewage treatment strain feeding amounts and water quality standard reaching times, so that when historical observation data is insufficient, various situations can be simulated to simulate mirror twinning, a situation expert library of multi-source heterogeneous data is established, and rich data support is provided for organic biochemical reaction strain feeding guidance.

[0040] 2) The present application uses multiple AI models to establish an AI calibration modeling model and is used for strain dosage guidance, which can solve the problem of weak generalization performance of a single model in different situations. Finally, a stable, reliable, practical and intelligent decision-making operation scheme is obtained through the set average method.

[0041] 3) The present application deploys the trained AI calibration modeling model to each sewage treatment field edge, has the function of decentralized edge computing, and the AI model parameters can be input into the on-site chip for on-site distributed computing and decision-making, which can reduce the pressure of the central server, improve the response speed of each edge of the sewage treatment, reduce the lag problem caused by the conventional water quality determination and strain dosage calculation, and improve the actual sewage treatment effect.

[0042] 4) The present application fully considers the influence of different water quality situations and different hydrodynamic situations on the strain dosage of sewage treatment, and also considers the water quality standard time requirement in practical application, and the water quality standard time control is included in the strain dosage guidance, which is more proactive operation. Users can specifically guide important processing processes in the sewage treatment link, including water quality standard setting and reaction time setting, etc., to achieve the actual guidance effect of sewage treatment. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0044] Figure 1 The present application is an organic biochemical reaction strain dosage guidance method total flow chart;

[0045] Figure 2 The present application is a mechanism water quality modeling calibration model establishment step flow chart;

[0046] Figure 3 The present application is an expert database establishment flow chart;

[0047] Figure 4 The present application is an AI calibration modeling model establishment step flow chart;

[0048] Figure 5 The present application is an AI calibration modeling model architecture diagram;

[0049] Figure 6 The present application is an edge computing application process flow chart. DETAILED DESCRIPTION

[0050] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of protection of the present application.

[0051] Please refer to Figure 1 The present application provides a method for guiding the release of organic biochemical reaction bacteria, which comprises the following steps:

[0052] Step 1, a historical water quality monitoring data collection step, for collecting historical water quality monitoring data and performing data cleaning on part of the data.

[0053] The historical water quality monitoring data includes historical monitoring time series data of a plurality of water quality indicators and corresponding historical sewage treatment bacteria release data, corresponding hydrodynamic historical monitoring time series data, and corresponding historical degradation process data of the water quality indicators; the plurality of water quality indicators include but are not limited to COD, dissolved oxygen, UV254 and BOD; the hydrodynamic historical monitoring time series data include but are not limited to water level, flow rate and flow cross section data.

[0054] The specific steps of step 1 are as follows:

[0055] Step 1.1, collecting historical monitoring time series data of a plurality of water quality indicators, including but not limited to COD, dissolved oxygen, UV254, BOD, etc.

[0056] Step 1.2, inputting the historical monitoring time series data of a plurality of water quality indicators collected in step 1.1 into a data cleaning tool to correct unreasonable data, wherein the data cleaning tool includes but is not limited to Kalman filtering, filtering algorithm, fuzzy theory, principal component analysis, etc.

[0057] Step 1.3, collecting historical sewage treatment bacteria release data corresponding to the historical monitoring time series data of a plurality of water quality indicators.

[0058] Step 1.4, collecting hydrodynamic historical monitoring time series data corresponding to the historical monitoring time series data of a plurality of water quality indicators, including but not limited to water level, flow rate, flow cross section data, etc.

[0059] Step 1.5, inputting the hydrodynamic historical monitoring time series data collected in step 1.4 into the data cleaning tool to correct unreasonable data.

[0060] Step 1.6, collecting historical degradation process data of the water quality indicators corresponding to the historical sewage treatment bacteria release data.

[0061] The collected historical water quality monitoring data can be used in the mechanism water quality modeling rate model establishment process of step 2.

[0062] Step 2, mechanism water quality modeling rate model establishment step, is used to establish a mechanism water quality modeling rate model, and parameters of the mechanism water quality modeling rate model are adjusted by the cleaned historical water quality monitoring data.

[0063] As shown in Figure 2 , the specific steps of step 2 are as follows:

[0064] Step 2.1, the present application uses 0D, 1D, 2D, 3D water quality models including but not limited to water quality complete mixing model, CE-QUAL-W2, EFDC, WASP, DELFT3D, HEC-RAS, etc. to establish a mechanism water quality modeling rate model. The achievements of step 1.2, step 1.3, step 1.5 are input into the mechanism water quality modeling rate model, and the simulated degradation process data of the water quality index are output.

[0065] Step 2.2, a rate verification cycle process is carried out by the historical degradation process data of the water quality index of step 1.6 and the simulated degradation process data of the water quality index of step 2.1, and the parameter setting of the mechanism water quality modeling rate model is adjusted until the model reaches a standard efficiency coefficient of 0.5 or more.

[0066] The mechanism water quality modeling rate model that completes the rate verification can be used in the expert library establishment process of step 3.

[0067] Step 3, expert library establishment step, is used to establish an expert library based on the mechanism water quality modeling rate model, and the expert library is used to simulate the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment strain dosages and water quality compliance times.

[0068] As shown in Figure 3 , the specific steps of step 3 are as follows:

[0069] Step 3.1, a plurality of water quality situations are established, and different water quality index solubilities are set, including but not limited to COD, dissolved oxygen, UV254, BOD, etc.

[0070] Step 3.2, different sewage treatment strain dosing situations are established, and different strain dosages are set.

[0071] Step 3.3, different hydrodynamic situations are established, and different situation parameters are set, including but not limited to water level, flow rate, flow cross section data, etc.

[0072] Step 3.4, the achievements of step 3.1, step 3.2, step 3.3 are input into the mechanism water quality modeling rate model that completes the rate verification in step 2.

[0073] Step 3.5: Output the degradation process data of water quality indicators under the corresponding scenario from Step 3.4.

[0074] Step 3.6: Analyze the results of Step 3.5 to determine the water quality compliance time under the corresponding scenario.

[0075] Step 3.7: From Steps 3.1, 3.2, 3.3 and 3.6, we can obtain the correspondence between different water quality scenarios, different hydrodynamic scenarios, and different wastewater treatment microbial inoculation scenarios and the time for water quality to meet standards, thus completing the establishment of the expert database.

[0076] The established expert database can be used in the AI ​​modeling process in step 4.

[0077] This invention utilizes a mechanistic water quality modeling calibration model to create digital twins, simulating the correspondence between different water quality scenarios, hydrodynamic scenarios, wastewater treatment microbial dosage and water quality compliance time. When historical observation data is insufficient, various scenarios can be set to simulate mirror twin situations, establishing a scenario expert database of multi-source heterogeneous data, providing rich data support for guiding the dosage of organic biochemical reaction microbial strains.

[0078] Step 4: AI calibration model establishment step, used to establish an AI calibration model through multiple AI models, using different water quality scenarios, hydrodynamic scenarios and corresponding water quality compliance times in the expert database as inputs and the corresponding amount of sewage treatment bacteria as outputs to train the AI ​​calibration model.

[0079] like Figure 4 As shown, the specific steps of step 4 are as follows:

[0080] Step 4.1: Obtain data on different water quality compliance times, various water quality scenarios, and different hydrodynamic scenarios from the expert database completed in Step 3.

[0081] The results of steps 4.2 and 4.1 are input into the AI ​​calibration model. The architecture of the AI ​​model is as follows: Figure 5 As shown, the various AI models include, but are not limited to, support vector machines, K-nearest neighbors, stochastic gradient descent, multivariate linear regression, multilayer perceptron, decision trees, backpropagation neural networks, and radial basis function networks. The results of step 4.1 are input into each AI model, and the output results of each AI model are used to calculate the simulated amount of wastewater treatment bacteria to be added using the set averaging method.

[0082] Step 4.3: Obtain the wastewater treatment microbial inoculation scenarios corresponding to different water quality compliance times, multiple water quality scenarios, and different hydrodynamic scenarios from the expert database in Step 3.

[0083] Step 4.4, the rate verification cycle process is carried out through the sewage treatment bacteria release scenario of step 4.3 and the sewage treatment bacteria release simulation quantity of step 4.2, the parameter settings of the AI rate modeling model are adjusted through the corresponding sewage treatment bacteria release quantity in the expert library, such as the influence weight coefficient of each AI model on the sewage treatment bacteria release quantity, the parameters of the AI model itself, etc. Until the model reaches the available standard efficiency coefficient of 0.5 or more.

[0084] The AI rate modeling model completed in the rate verification is used in the step 5 edge computing application process.

[0085] The present application uses multiple AI models to establish an AI rate modeling model and for bacteria release quantity guidance, which can solve the problem of weak generalization performance of a single model in different situations. The parameters of the AI rate modeling model are adjusted using the data in the expert library as training data, and finally a stable, reliable, practical and intelligent decision-making operation scheme is obtained through the ensemble average method.

[0086] Step 5, edge computing application step, is used for sewage treatment bacteria release quantity guidance through the trained AI rate modeling model at the edge of the sewage treatment site.

[0087] The trained AI rate modeling model is deployed at each sewage treatment site edge as an edge computing AI model, and the edge computing application process step is as follows Figure 6 , the specific steps are as follows:

[0088] Step 5.1, collect multiple water quality real-time monitoring time series data, including but not limited to COD, dissolved oxygen, UV254, BOD, etc.

[0089] Step 5.2, input the multiple water quality real-time monitoring time series data collected in step 5.1 into the data cleaning tool to correct unreasonable data.

[0090] Step 5.3, set the expected water quality compliance time.

[0091] Step 5.4, collect water power real-time monitoring time series data, including but not limited to water level, flow rate, flow cross section data, etc.

[0092] Step 5.5, input the water power real-time monitoring time series data collected in step 5.4 into the data cleaning tool to correct unreasonable data.

[0093] Step 5.6, input the results of steps 5.2, 5.3 and 5.5 into the edge computing AI model.

[0094] Step 5.7, the sewage treatment bacteria release quantity real-time guidance value can be obtained from step 5.6.

[0095] The application fully considers the influence of different water quality situations and different hydrodynamic situations on the sewage treatment strain dosage, and simultaneously considers the water quality standard time requirement in actual application, and the water quality standard time control is incorporated into the strain dosage guidance, which is more proactive operation, and the user can specifically guide the important treatment process in the sewage treatment link, including the setting of water quality standard and reaction time, etc., to achieve the actual guidance role of sewage treatment. The trained AI rate modeling model is used for edge calculation at the edge of the sewage treatment site, the data calculation center server pressure is reduced through decentralization, and at the same time, the sewage treatment strain dosage is conveniently guided in real time at the site end, and the sewage treatment response speed is improved.

[0096] Corresponding to the method embodiment, the application also proposes an organic biochemical reaction strain dosage guidance system based on multi-source heterogeneous model fusion, which comprises:

[0097] A data collection module is configured to collect historical water quality monitoring data and perform data cleaning.

[0098] A parameter adjustment module is configured to establish a mechanism water quality modeling rate model, and adjust the parameters of the mechanism water quality modeling rate model through the cleaned historical water quality monitoring data.

[0099] An expert library establishment module is configured to establish an expert library based on the mechanism water quality modeling rate model, and simulate the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment strain dosages and water quality standard times.

[0100] A model construction module is configured to establish an AI rate modeling model through a plurality of AI models, and train the AI rate modeling model by taking different water quality situations, hydrodynamic situations and corresponding water quality standard times in the expert library as inputs and corresponding sewage treatment strain dosages as outputs.

[0101] A real-time guidance module is configured to guide the sewage treatment strain dosage at the edge of the sewage treatment site through the trained AI rate modeling model.

[0102] The above method embodiment and system embodiment are one-to-one corresponding, and the system embodiment can be referred to the method embodiment for a brief description.

[0103] The application also discloses an electronic device, which comprises at least one processor, at least one memory, a communication interface and a bus; wherein the processor, the memory and the communication interface complete mutual communication through the bus; the memory stores program instructions executable by the processor, and the processor calls the program instructions to implement the method of the application.

[0104] The application further discloses a computer readable storage medium which stores computer instructions, and the computer instructions make the computer realize all or part of steps of the method.

[0105] The system embodiments described above are only illustrative, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be distributed to multiple network units. A person of ordinary skill in the art can select part or all of the modules to achieve the purpose of the embodiments according to actual needs without creative labor.

[0106] The above only describes the preferred embodiments of the application and is not intended to limit the application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the application shall be included in the protection scope of the application.

Claims

1. An organic biochemical reaction strain inoculation guidance method, characterized by, The method comprises: collecting historical water quality monitoring data and performing data cleaning on part of the data; establishing a mechanism water quality modeling calibration model, and adjusting parameters of the mechanism water quality modeling calibration model through the cleaned historical water quality monitoring data; establishing an expert library based on the mechanism water quality modeling calibration model, for simulating the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment bacterial strain dosages and water quality compliance times, specifically comprising: setting the solubility of multiple water quality indicators under different water quality situations, the sewage treatment bacterial strain dosage under different sewage treatment bacterial strain dosages, and the situation parameters under different hydrodynamic situations; inputting the solubility of multiple water quality indicators under different water quality situations, the sewage treatment bacterial strain dosage under different sewage treatment bacterial strain dosages, and the situation parameters under different hydrodynamic situations into the mechanism water quality modeling calibration model with adjusted parameters, and outputting water quality indicator degradation process data under the corresponding situation respectively; analyzing the water quality compliance time under the corresponding situation according to the water quality indicator degradation process data under the corresponding situation, to obtain the corresponding relationship between different water quality situations, hydrodynamic situations, sewage treatment bacterial strain dosages and water quality compliance times; training the AI calibration modeling model through multiple AI models, taking different water quality situations, hydrodynamic situations and corresponding water quality compliance times in the expert library as inputs, and taking corresponding sewage treatment bacterial strain dosages as outputs; guiding the sewage treatment bacterial strain dosage at the edge of the sewage treatment site through the trained AI calibration modeling model, specifically comprising: deploying the trained AI calibration modeling model at each sewage treatment site edge as an edge computing AI model; obtaining real-time monitoring time series data of multiple water quality indicators and water dynamic real-time monitoring time series data at the edge and performing data cleaning; setting an expected water quality compliance time; inputting the cleaned real-time monitoring time series data of multiple water quality indicators, water dynamic real-time monitoring time series data and expected water quality compliance time into the edge computing AI model, and outputting a real-time guidance value of the sewage treatment bacterial strain dosage.

2. The organic biochemical reaction strain inoculation guidance method according to claim 1, characterized by, The historical water quality monitoring data comprises historical monitoring time series data of multiple water quality indicators and corresponding historical sewage treatment bacterial strain dosage data, corresponding water dynamic historical monitoring time series data and corresponding water quality indicator historical degradation process data; the multiple water quality indicators include but are not limited to COD, dissolved oxygen, UV254 and BOD; the water dynamic historical monitoring time series data includes but is not limited to water level, flow rate and flow cross section data.

3. The organic biochemical reaction strain inoculation guidance method according to claim 2, characterized by, The data cleaning on part of the data specifically comprises: cleaning the historical monitoring time series data of multiple water quality indicators and the water dynamic historical monitoring time series data through a data cleaning tool, the data cleaning tool including but not limited to Kalman filtering, filtering algorithm, fuzzy theory and principal component analysis.

4. The organic biochemical reaction strain inoculation guidance method according to claim 2, characterized by, The adjustment of the parameters of the mechanism water quality modeling calibration model through the cleaned historical water quality monitoring data specifically comprises: establishing a mechanism water quality modeling calibration model, which includes but is not limited to a water quality complete mixing model, CE-QUAL-W2, EFDC, WASP, DELFT3D or HEC-RAS; inputting the cleaned multiple water quality historical monitoring time series data, corresponding historical sewage treatment strain release data and corresponding cleaned hydrodynamic historical monitoring time series data into the mechanism water quality modeling calibration model to obtain water quality index simulation degradation process data output by the mechanism water quality modeling calibration model; adjusting parameter settings of the mechanism water quality modeling calibration model through water quality index historical degradation process data and water quality index simulation degradation process calibration verification cycles until the model reaches a preset standard efficiency coefficient.

5. The organic biochemical reaction seed inoculation guidance method according to claim 1, characterized by, The AI calibration modeling model is established by multiple AI models, with different water quality situations, hydrodynamic situations and corresponding water quality compliance times in the expert library as inputs and corresponding sewage treatment strain release amounts as outputs, and the AI calibration modeling model is trained, which specifically includes: obtaining an AI calibration modeling model composed of multiple AI models, which include but are not limited to support vector machines, K-nearest neighbor method, stochastic gradient descent, multivariate linear regression, multilayer perception, decision tree, back propagation neural network and radial basis function network; inputting different water quality situations, hydrodynamic situations and corresponding water quality compliance times in the expert library into each AI model respectively, calculating a sewage treatment strain release simulation amount by set average method from output results of each AI model, and performing calibration verification cycles according to the sewage treatment strain release simulation amount and corresponding sewage treatment strain release amounts in the expert library to adjust parameter settings of the AI calibration modeling model.

6. An organic biochemical reaction seed inoculation guidance system, characterized by, The system includes: a data collection module for collecting historical water quality monitoring data and cleaning part of the data; a parameter adjustment module for establishing a mechanism water quality modeling calibration model and adjusting parameters of the mechanism water quality modeling calibration model through cleaned historical water quality monitoring data; an expert library establishment module for establishing an expert library based on the mechanism water quality modeling calibration model to simulate corresponding relationships between different water quality situations, hydrodynamic situations, sewage treatment strain release amounts and water quality compliance times, which specifically includes: setting solubilities of multiple water quality indexes under different water quality situations, sewage treatment strain release amounts under different sewage treatment strain release situations and situation parameters under different hydrodynamic situations respectively; inputting the solubilities of the multiple water quality indexes under the different water quality situations, the sewage treatment strain release amounts under the different sewage treatment strain release situations and the situation parameters under the different hydrodynamic situations into the mechanism water quality modeling calibration model with adjusted parameters and outputting water quality index degradation process data under corresponding situations respectively; analyzing water quality compliance times under corresponding situations according to the water quality index degradation process data under the corresponding situations to obtain corresponding relationships between different water quality situations, hydrodynamic situations, sewage treatment strain release amounts and water quality compliance times; and The model construction module is configured to establish an AI calibration modeling model by using a plurality of AI models, and train the AI calibration modeling model by taking different water quality scenarios, hydrodynamic scenarios, and corresponding water quality compliance times in an expert library as inputs and corresponding sewage treatment strain dosages as outputs. The real-time guidance module is configured to guide the sewage treatment strain dosage by using the trained AI calibration modeling model at a sewage treatment site edge, and the guidance of the sewage treatment strain dosage by using the trained AI calibration modeling model at the sewage treatment site edge specifically includes: deploying the trained AI calibration modeling model at each sewage treatment site edge as an edge computing AI model; acquiring real-time monitoring time series data of a plurality of water quality indexes and hydrodynamic real-time monitoring time series data at the edge and performing data cleaning; setting an expected water quality compliance time; inputting the cleaned real-time monitoring time series data of the plurality of water quality indexes, the hydrodynamic real-time monitoring time series data, and the expected water quality compliance time into the edge computing AI model, and outputting a real-time guidance value of the sewage treatment strain dosage.

7. An electronic device, comprising: The system comprises: at least one processor, at least one memory, a communication interface, and a bus; wherein the processor, the memory, and the communication interface communicate with each other through the bus; the memory stores program instructions executable by the processor, and the processor invokes the program instructions to implement the method of any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions, and the computer instructions enable the computer to implement the method of any one of claims 1-5.

Images