Sewage treatment detection method and system with diagnosis function

Abstract

The invention provides a sewage treatment detection method and system with a diagnosis function, and the method comprises the steps: firstly receiving multi-source detection data which is uploaded by each treatment unit in a sewage treatment process and comprises sewage pollutant component data, treatment equipment operation parameter data and flora activity data in a reaction tank; performing cross-unit association mapping processing on the multi-source detection data to generate a substance transformation association network among processing units, constructing a dynamic balance model of the sewage treatment system based on the network, inputting the multi-source detection data acquired in real time into the model, analyzing abnormal association nodes deviating from a balance state and corresponding abnormal influence factors, and determining the abnormal association nodes and the corresponding abnormal influence factors; and finally, generating a diagnosis report containing an abnormal processing path and an equipment regulation and control instruction according to an analysis result, thereby accurately positioning a problem source of the sewage treatment system, timely adjusting equipment parameters to recover system balance, and improving the processing efficiency and stability.

Description

Technical Field

[0001] The present invention relates to the technical field of sewage treatment, and in particular to a sewage treatment detection method and system with diagnostic functions. Background Art

[0002] In the field of wastewater treatment, ensuring the stable and efficient operation of wastewater treatment systems is crucial for safeguarding environmental quality and the sustainable use of water resources. Currently, wastewater treatment processes typically incorporate monitoring devices within each treatment unit to capture relevant data. However, existing wastewater treatment monitoring methods have numerous limitations.

[0003] On the one hand, the test data of each treatment unit is often collected and analyzed independently, lacking in-depth exploration of the data correlations between different treatment units. For example, they only focus on the removal of pollutants in a single reaction tank, while ignoring the impact of the operating parameters of the equipment in the preceding treatment unit on the activity of the reaction tank's microbial community and the efficiency of pollutant treatment. As a result, they cannot fully understand the complex processes of material transformation and energy flow in the entire sewage treatment system.

[0004] On the other hand, when anomalies occur in the sewage treatment system, existing detection methods struggle to quickly and accurately pinpoint the root cause. Due to a lack of modeling and analysis of the system's overall dynamic equilibrium, operations and maintenance personnel are typically forced to rely on experience to individually troubleshoot each treatment unit. This is not only inefficient but also prone to overlooking critical issues, preventing the timely implementation of effective treatment measures. This impacts the normal operation and effectiveness of the sewage treatment system and may even lead to excessive sewage discharge, causing serious environmental pollution. Summary of the Invention

[0005] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides a sewage treatment detection method with a diagnostic function, the method comprising: Receive multi-source detection data uploaded by each treatment unit in the sewage treatment process, the multi-source detection data including the pollutant composition data of the sewage, the operating parameter data of the treatment equipment and the bacterial activity data in the reaction tank; Performing cross-unit correlation mapping processing on the multi-source detection data to generate a material transformation correlation network between the processing units, wherein the material transformation correlation network is used to reflect the mutual influence relationship between pollutants, equipment operating parameters and bacterial activity in different processing units; Constructing a dynamic equilibrium model of the sewage treatment system based on the material transformation association network, wherein the dynamic equilibrium model is used to reflect the parameter matching relationship of each treatment unit of the sewage treatment system under a stable operation state; Inputting multi-source detection data collected in real time into the dynamic balance model, parsing out abnormal associated nodes that deviate from the equilibrium state and corresponding abnormal influencing factors, wherein the abnormal associated nodes are nodes in the processing unit where parameter imbalance occurs, and the abnormal influencing factors are pollutants or equipment operating parameters that cause parameter imbalance; A diagnosis report including an abnormality processing path and an equipment control instruction are generated according to the abnormality associated node and the abnormality influencing factor, and the equipment control instruction is used to adjust the equipment operating parameters of the corresponding processing unit to restore the system balance.

[0006] On the other hand, an embodiment of the present invention also provides a sewage treatment detection system with diagnostic function, including a processor and a machine-readable storage medium, the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0007] Based on the above aspects, the embodiment of the present invention receives multi-source detection data uploaded by each treatment unit in the sewage treatment process, performs cross-unit correlation mapping processing on the multi-source detection data, and generates a material transformation association network between each treatment unit, which deeply reveals the complex mutual influence relationship between pollutants, equipment operating parameters and bacterial activity in different treatment units. The dynamic balance model of the sewage treatment system constructed based on the material transformation association network can accurately reflect the parameter matching relationship of each treatment unit under the stable operation state of the system. The multi-source detection data collected in real time is input into the dynamic balance model, and the abnormal correlation nodes and corresponding abnormal influencing factors that deviate from the equilibrium state can be quickly analyzed, thereby realizing the accurate positioning of the root cause of the problem. The diagnostic report and equipment control instructions containing the abnormal processing path generated according to the analysis results can guide the operation and maintenance personnel to take effective treatment measures in time, adjust the equipment operating parameters of the corresponding treatment unit to restore the system balance, significantly improve the operation stability and treatment efficiency of the sewage treatment system, and reduce the risk of excessive sewage discharge. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a schematic diagram of the execution flow of the sewage treatment detection method with diagnostic function provided by an embodiment of the present invention.

[0009] Figure 2 Schematic diagram of exemplary hardware and software components of a sewage treatment detection system with diagnostic functions provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1FIG1 is a flow chart of a sewage treatment detection method with a diagnostic function provided by an embodiment of the present invention. The sewage treatment detection method with a diagnostic function is introduced in detail below.

[0011] This example uses a municipal sewage treatment plant as an example. The plant consists of a grid treatment unit, an equalization tank, a primary sedimentation tank, an A / O biological reactor, a secondary sedimentation tank, a deep filtration unit, and a UV disinfection unit. Each treatment unit is equipped with a sensor group and a data transmission module for real-time data collection and upload.

[0012] Step S110: receiving multi-source detection data uploaded by each treatment unit in the sewage treatment process, wherein the multi-source detection data includes pollutant component data of the sewage, operating parameter data of the treatment equipment, and bacterial activity data in the reaction tank.

[0013] The central control system of a municipal sewage treatment plant connects to the data transmission modules of each treatment unit via industrial Ethernet. Sensors in the screen treatment units collect pollutant composition data, such as the suspended solids content of wastewater after it passes through the screens and the amount of material intercepted by the screens. They also collect operating parameters, such as the screens' operating frequency and control parameters for bar spacing. These data are then uploaded to the central control system via the data transmission module.

[0014] The sensor group of the equalization tank treatment unit collects data related to the pollutant components of the sewage in the pool, such as the pH value, water temperature, water volume, as well as operating parameter data such as the operating speed of the agitator and the liquid level control parameters of the equalization tank, which are also uploaded through the data transmission module.

[0015] The data uploaded by the primary sedimentation tank treatment unit includes pollutant composition data such as turbidity and suspended matter concentration of the pool effluent, as well as operating parameter data such as the scraper operation cycle and sludge pump flow rate.

[0016] As the core processing unit, the A / O bioreactor uploads richer data. Pollutant composition data includes chemical oxygen demand, ammonia nitrogen, total nitrogen, and total phosphorus concentrations in both the aerobic and anoxic zones. Operating parameter data includes aeration fan pressure and aeration volume in the aerobic zone, agitator power in the anoxic zone, and return pump flow rate. Bacterial activity data, collected through biosensors within the tank, includes information on various bacterial species, including nitrifying bacteria and aerobic heterotrophic bacteria in the aerobic zone and denitrifying bacteria in the anoxic zone, as well as activity-related data such as cell concentration and respiration rate for each bacterial species.

[0017] The pollutant composition data uploaded by the secondary sedimentation tank treatment unit include the suspended solids concentration and transparency of the effluent, and the operating parameter data include the scraper operating speed, sludge return ratio, etc.

[0018] The deep filtration treatment unit uploads pollutant composition data such as turbidity and COD concentration of filtered water, as well as operating parameter data such as the backwash cycle and backwash water volume of the filter tank.

[0019] The pollutant composition data uploaded by the ultraviolet disinfection treatment unit is the number of fecal coliform bacteria in the effluent, and the operating parameter data includes the power of the ultraviolet lamp, water flow rate, etc.

[0020] The central control system's receiving module parses the data uploaded by each processing unit and stores it in a categorized manner according to the processing unit number and data type. The data is stored in a structured data table, where each data record contains the data collection time, processing unit identifier, data type identifier, and a data value set.

[0021] Step S120: performing cross-unit association mapping processing on the multi-source detection data to generate a material transformation association network between the processing units. The material transformation association network is used to reflect the mutual influence relationship between pollutants, equipment operating parameters and bacterial activity in different processing units.

[0022] Step S121: extracting characteristic pollutant indicators from the pollutant component data of each processing unit from the multi-source detection data, wherein the characteristic pollutant indicators are the types and concentrations of pollutants that change significantly during the processing process.

[0023] For the screen treatment unit, analyze the pollutant composition data for its influent and effluent, and calculate the removal rate of each pollutant. The removal rate is calculated as (influent concentration - effluent concentration) / influent concentration. Pollutants with removal rates exceeding a set percentage, such as suspended solids, are screened and identified as characteristic pollutant indicators for the screen treatment unit. The concentration data for this pollutant at both the influent and effluent levels are also recorded.

[0024] The treatment unit of the equalization tank mainly plays the role of regulating water quality and quantity. Its characteristic pollutant indicators are pH value and water temperature. Because these two indicators will change significantly in the equalization tank to adapt to the subsequent treatment process, it is necessary to extract pH value and water temperature data at different times.

[0025] The primary sedimentation tank treatment unit focuses on removing settleable suspended solids and some organic matter. By comparing the inlet and outlet data, suspended solids and five-day biochemical oxygen demand are used as characteristic pollutant indicators to extract corresponding concentration data.

[0026] In the A / O bioreactor treatment unit, ammonia nitrogen in the aerobic zone is converted to nitrate nitrogen through nitrification, while nitrate nitrogen in the anoxic zone is converted to nitrogen gas through denitrification. Simultaneously, chemical oxygen demand (COD) is degraded in both the aerobic and anoxic zones. Therefore, ammonia nitrogen, total nitrogen, and COD were identified as characteristic pollutant indicators for this treatment unit, and concentration data were collected for the aerobic zone influent, aerobic zone effluent, anoxic zone influent, and anoxic zone effluent.

[0027] The characteristic pollutant indicator of the secondary sedimentation tank treatment unit is suspended solids, and its effluent concentration data is extracted.

[0028] The deep filtration treatment unit further removes fine suspended matter and some dissolved organic matter, uses turbidity and chemical oxygen demand as characteristic pollutant indicators, and extracts concentration data before and after filtration.

[0029] The characteristic pollutant indicator of the UV disinfection treatment unit is the number of fecal coliform bacteria, and the quantitative data before and after disinfection are extracted.

[0030] Step S122: extracting key operating parameters from the operating parameter data of each processing unit, wherein the key operating parameters are equipment operating parameters that have a direct impact on the pollutant treatment effect.

[0031] Step S1221: Acquire all operating parameters and corresponding parameter values ​​in the operating parameter data of each processing unit.

[0032] The operating parameters of the grid processing unit include the operating frequency of the grid machine, the control parameters corresponding to the grid bar spacing, the motor current, etc. The specific values ​​of these parameters at different times are extracted from the uploaded data.

[0033] The operating parameters of the regulating tank treatment unit include the operating speed of the agitator, the liquid level control parameters of the regulating tank, the opening of the water inlet valve, the opening of the water outlet valve, etc. The parameter values ​​of each parameter are extracted.

[0034] The operating parameters of the primary sedimentation tank treatment unit include the operating cycle of the scraper, the flow rate of the sludge pump, the sludge discharge frequency of the sludge hopper, etc., and the corresponding parameter values ​​are obtained.

[0035] The operating parameters of the A / O biological reactor treatment unit include the wind pressure and aeration volume of the aeration fan in the aerobic zone, the power of the agitator in the anoxic zone, the flow rate of the return pump, as well as the sludge age control parameters, dissolved oxygen control values, etc. The specific values ​​of these parameters are extracted.

[0036] The operating parameters of the secondary sedimentation tank treatment unit include the scraper operating speed, sludge return ratio, residual sludge discharge, etc., and the values ​​of each parameter are extracted.

[0037] The operating parameters of the deep filtration treatment unit include the backwash cycle of the filter tank, the backwash water volume, the filtration flow rate, etc., and the corresponding parameter values ​​are obtained.

[0038] The operating parameters of the ultraviolet disinfection treatment unit include the power of the ultraviolet lamp, water flow rate, lamp cleanliness parameters, etc., and the parameter values ​​of each parameter are extracted.

[0039] Step S1222: collecting pollutant treatment effect data corresponding to different values ​​of each operating parameter, wherein the pollutant treatment effect data includes pollutant removal rate and pollutant concentration after treatment.

[0040] For the grid treatment unit's grid machine operating frequency parameter, data on suspended solids removal rates and post-treatment suspended solids concentrations were collected at different operating frequencies. At the same time, data on pollutant treatment effectiveness at different parameter values ​​for the control parameter corresponding to the grid bar spacing was collected.

[0041] The stirrer operating speed parameters of the regulating pool treatment unit are adjusted to collect treatment effect data such as the stability of the pH value at different speeds (reflected by the fluctuation range of the pH value) and the uniformity of the water temperature (reflected by the difference in water temperature at different monitoring points).

[0042] The scraper operation cycle parameters of the primary sedimentation tank treatment unit are used to collect suspended solids removal rate and suspended solids concentration data under different cycles; for the sludge pump flow parameters, pollutant treatment effect data under different flow rates are collected.

[0043] The aeration rate parameters of the aerobic zone of the A / O biological reactor treatment unit are used to collect data on the ammonia nitrogen removal rate and ammonia nitrogen concentration after treatment under different aeration rates; the power parameters of the agitator in the anoxic zone are used to collect data on the total nitrogen removal rate under different powers; the flow rate parameters of the reflux pump are used to collect data on the total nitrogen treatment effect under different flow rates.

[0044] The operating parameters of other treatment units are collected in a similar manner to collect corresponding pollutant treatment effect data.

[0045] Step S1223: Calculate the correlation coefficient between each operating parameter and the pollutant treatment effect data, where the correlation coefficient is used to represent the degree of linear correlation between the two.

[0046] Taking the grid machine operating frequency and suspended solids removal rate of the grid treatment unit as an example, the parameter values ​​of different operating frequencies and the corresponding suspended solids removal rate data were combined into two data series. The correlation coefficient was calculated by dividing the covariance of the two data sets by the product of the two data standard deviations.

[0047] For the agitator operating speed and pH value fluctuation range of the regulating tank treatment unit, the data sequence is also composed and the correlation coefficient is calculated to reflect the degree of linear correlation between the two.

[0048] Using the same method, the correlation coefficient between each operating parameter in each treatment unit and the corresponding pollutant treatment effect data is calculated.

[0049] Step S1224: Screen out operating parameters whose absolute values ​​of correlation coefficients exceed a preset correlation threshold as candidate key operating parameters.

[0050] A preset correlation threshold is set. For the screen treatment unit, if the absolute value of the correlation coefficient between the screen machine operating frequency and the suspended matter removal rate exceeds the threshold, the screen machine operating frequency will be determined as a candidate key operating parameter; if the absolute value of the correlation coefficient between the control parameter corresponding to the bar spacing and the pollutant treatment effect does not exceed the threshold, it will not be included in the candidate range.

[0051] The operating parameters of each processing unit are screened one by one to obtain all candidate key operating parameters.

[0052] Step S1225: Perform sensitivity analysis on the candidate key operating parameters, calculate the change in pollutant treatment effect when the parameter value changes by a unit, and obtain a sensitivity coefficient.

[0053] Taking the aeration volume of the aerobic zone of the A / O biological reactor treatment unit as an example, while other parameters remain unchanged, the parameter value of the aeration volume is increased by one unit, and the change in the ammonia nitrogen removal rate at this time is recorded. This change is the sensitivity coefficient of the parameter under the current value.

[0054] Repeat the above operation in different parameter value intervals to obtain the sensitivity coefficient of the candidate key operating parameter in different intervals, and take the average value as the final sensitivity coefficient.

[0055] In the same way, the sensitivity coefficients of all candidate key operating parameters are calculated.

[0056] Step S1226: Sort the sensitivity coefficients in descending order, and select the operating parameters with the highest order as key operating parameters. The key operating parameters are equipment operating parameters that have a direct impact on the pollutant treatment effect.

[0057] The sensitivity coefficients of the candidate key operating parameters of the A / O biological reactor treatment unit are ranked. If the sensitivity coefficient of the aeration rate in the aerobic zone is ranked high, it is determined as the key operating parameter of the treatment unit.

[0058] Other processing units also follow this method, selecting the top-ranked candidate key operating parameters as the final key operating parameters. For example, the screen processing unit selects the screen machine operating frequency, the regulating tank processing unit selects the agitator operating speed, and the primary sedimentation tank processing unit selects the scraper operating cycle.

[0059] Step S123: extracting the dominant bacterial index from the bacterial activity data in the reaction tank, wherein the dominant bacterial index is the type and activity of the bacterial group that plays a leading role in the pollutant degradation process.

[0060] Step S1231: Acquire all bacterial species and corresponding activity values ​​in the bacterial activity data in the reaction pool, wherein the activity value includes the bacterial population number and metabolic rate.

[0061] The aerobic and anoxic zones of the A / O biological reactor treatment unit are both equipped with biosensors. Through gene sequencing and metabolic monitoring technology, the presence of nitrifying bacteria (including ammonia-oxidizing bacteria and nitrite-oxidizing bacteria), aerobic heterotrophic bacteria and other bacterial species in the aerobic zone is determined; the presence of denitrifying bacteria and other bacterial species in the anoxic zone is determined.

[0062] The number of each bacterial group, such as the number of cells per milliliter of mixed solution, and metabolic rates, such as the ammonia oxidation rate of ammonia-oxidizing bacteria and the nitrate reduction rate of denitrifying bacteria, are extracted from the bacterial activity data.

[0063] Step S1232: analyzing the degradation correlation between each bacterial species and characteristic pollutant indicators, wherein the degradation correlation is calculated by the ratio of the degradation rate of the characteristic pollutant when the bacterial species exists to the degradation rate when the bacterial species does not exist.

[0064] Ammonia nitrogen is the characteristic pollutant for ammonia-oxidizing bacteria in the aerobic zone. Under experimental conditions, the degradation rate of ammonia nitrogen in the aerobic zone of the A / O bioreactor was measured in the presence of ammonia-oxidizing bacteria and in the absence of ammonia-oxidizing bacteria through sterilization. The ratio of these two degradation rates was used as the correlation between ammonia-oxidizing bacteria and ammonia nitrogen degradation.

[0065] For denitrifying bacteria, the corresponding characteristic pollutant is total nitrogen. Similarly, under experimental conditions, the degradation rate of total nitrogen is measured in the presence and absence of denitrifying bacteria, and the ratio of the two is calculated to obtain the correlation between denitrifying bacteria and total nitrogen degradation.

[0066] Using the same method, the degradation correlation between aerobic heterotrophic bacteria and characteristic pollutant indicators such as chemical oxygen demand was calculated.

[0067] Step S1233: Screen out bacterial species whose degradation correlation exceeds a preset correlation threshold as candidate dominant bacterial species, and calculate the proportion of the activity value of each candidate dominant bacterial species in the total activity value of all bacterial species to obtain the activity proportion.

[0068] A preset correlation threshold is set. If the correlation between ammonia oxidizing bacteria and ammonia nitrogen degradation exceeds the threshold, the ammonia oxidizing bacteria will be included in the candidate dominant bacterial group; if the correlation between denitrifying bacteria and total nitrogen degradation exceeds the threshold, they will also be included in the candidate dominant bacterial group.

[0069] Calculate the sum of the activity values ​​of the candidate dominant bacterial groups. For example, add the number and metabolic rate of ammonia-oxidizing bacteria to the corresponding activity values ​​of the other candidate dominant bacterial groups to obtain the total activity value. Then calculate the ratio of the activity value of each candidate dominant bacterial group to the total activity value, i.e., the activity percentage. For example, the ratio of the number of ammonia-oxidizing bacteria to the total number of all candidate dominant bacterial groups, and the ratio of their metabolic rate to the total metabolic rate, can be combined to obtain the activity percentage of ammonia-oxidizing bacteria.

[0070] Step S1234: determining the candidate dominant bacterial group whose activity ratio exceeds the preset activity ratio threshold as a dominant bacterial group type, and extracting the activity value corresponding to the dominant bacterial group type as dominant bacterial group activity data.

[0071] A preset activity ratio threshold is set. If the activity ratio of ammonia oxidizing bacteria exceeds the threshold, it will be determined as the dominant bacterial species, and its activity values ​​such as number and metabolic rate will be extracted as the dominant bacterial activity data.

[0072] Denitrifying bacteria and aerobic heterotrophic bacteria with a high correlation with chemical oxygen demand degradation and an activity ratio exceeding the threshold were identified as dominant bacterial species, and the corresponding activity values ​​were extracted.

[0073] Step S1235: combining the dominant bacterial species and the corresponding dominant bacterial activity data to obtain the dominant bacterial index in the bacterial activity data in the reaction tank, wherein the dominant bacterial index is the bacterial species and activity that play a leading role in the pollutant degradation process.

[0074] The determined dominant bacterial species, such as ammonia oxidizing bacteria, nitrite oxidizing bacteria, denitrifying bacteria, and specific aerobic heterotrophic bacteria, are combined with their corresponding activity values ​​(quantity, metabolic rate) to form the dominant bacterial index of the A / O biological reactor treatment unit.

[0075] These dominant bacterial community indicators can clearly reflect the status of the bacterial communities that play a dominant role in the pollutant degradation process, such as ammonia-oxidizing bacteria dominating the initial oxidation of ammonia nitrogen, and denitrifying bacteria dominating the reduction of nitrate nitrogen.

[0076] Step S124: establishing a processing unit association matrix, wherein rows of the processing unit association matrix represent preceding processing units, columns represent succeeding processing units, and matrix elements represent the degree of influence of the output substance of the preceding processing unit on the input substance of the succeeding processing unit.

[0077] Based on the treatment units of the municipal sewage treatment plant, a treatment unit association matrix was constructed. The rows of the matrix are screen treatment unit, regulating tank treatment unit, primary sedimentation tank treatment unit, A / O biological reactor treatment unit, secondary sedimentation tank treatment unit, and deep filtration treatment unit; the columns are regulating tank treatment unit, primary sedimentation tank treatment unit, A / O biological reactor treatment unit, secondary sedimentation tank treatment unit, deep filtration treatment unit, and ultraviolet disinfection treatment unit.

[0078] The values ​​of the matrix elements are determined by analyzing the degree to which the output of the preceding treatment unit affects the input of the subsequent treatment unit. For example, the output of the grid treatment unit enters the regulating tank treatment unit, and its influence on the input of the regulating tank treatment unit is reflected by the efficiency of the grid treatment unit in removing suspended solids. The higher the removal efficiency, the greater the influence, and the larger the value of the corresponding element in the matrix.

[0079] Similarly, the degree of influence of the equalization tank treatment unit on the primary sedimentation tank treatment unit is determined by the stability of its regulation of water quality and water quantity; the degree of influence of the primary sedimentation tank treatment unit on the A / O biological reactor treatment unit is determined by its effect in removing suspended solids and organic matter, and so on, to determine the value of each element in the treatment unit association matrix.

[0080] Step S125: The characteristic pollutant index, the key operating parameters and the dominant bacterial community index are used as nodes, input into the processing unit association matrix, and the association strength value between the nodes in different processing units is calculated. The association strength value is a quantitative value of the degree of mutual influence between two nodes.

[0081] The characteristic pollutant indicators of each treatment unit (such as suspended solids in the screen treatment unit, ammonia nitrogen in the A / O biological reactor treatment unit, etc.), key operating parameters (such as screen machine operation frequency, aeration volume in the aerobic zone, etc.), and dominant bacterial community indicators (such as ammonia oxidizing bacteria, denitrifying bacteria, etc.) are used as independent nodes.

[0082] These nodes are input into the processing unit association matrix. For two nodes in different processing units, such as the suspended solids node of the grid processing unit and the chemical oxygen demand node of the A / O biological reactor processing unit, the association strength value is calculated by analyzing the material conversion relationship and data correlation between the two.

[0083] The calculation of the association strength value comprehensively considers the influence degree of the corresponding processing unit in the processing unit association matrix and the correlation coefficient between the parameters represented by the two nodes, and is obtained through weighted combination to quantify the degree of mutual influence between the two nodes.

[0084] Step S126: Based on the association strength value, each node is connected according to the size of the association strength value to form a preliminary material transformation association network, and the nodes of the preliminary material transformation association network are optimized, and nodes with association strength values ​​lower than the preset threshold are merged, and nodes and connection relationships with association strength values ​​higher than the preset threshold are retained.

[0085] A preset threshold is set and nodes with correlation strength values ​​above the threshold are connected with line segments to form a preliminary material conversion correlation network. For example, the suspended solids node of the screen treatment unit has a high correlation strength value with the suspended solids node of the primary sedimentation tank treatment unit, so the two are connected; the aeration volume node of the aerobic zone of the A / O biological reactor treatment unit has a high correlation strength value with the ammonia oxidizing bacteria node, so the two are also connected.

[0086] For nodes whose association strength values ​​are lower than the preset threshold, such as the turbidity node of the deep filtration treatment unit and the fecal coliform count node of the ultraviolet disinfection treatment unit, these nodes are merged to simplify the network structure because their mutual influence is relatively low.

[0087] The nodes and connection relationships with association strength values ​​higher than the preset threshold are retained to form the basic structure of the optimized material transformation association network.

[0088] Step S127: Add the processing unit identifier corresponding to each node in the optimized material conversion association network to generate a final material conversion association network between each processing unit. The material conversion association network is used to reflect the mutual influence relationship between pollutants, equipment operating parameters and microbial activity in different processing units.

[0089] In the optimized material conversion network, each node is labeled with a corresponding treatment unit. For example, the suspended solids node is labeled "Screen Treatment Unit - Suspended Solids" or "Primary Sedimentation Tank Treatment Unit - Suspended Solids"; the aerobic zone aeration rate node is labeled "A / O Bioreactor Treatment Unit - Aerobic Zone Aeration Rate"; and the ammonia oxidizing bacteria node is labeled "A / O Bioreactor Treatment Unit - Ammonia Oxidizing Bacteria."

[0090] By adding treatment unit identifiers, the resulting material transformation association network can clearly reflect the mutual influence between pollutants, equipment operating parameters, and bacterial activity in different treatment units. For example, the connection between the "A / O biological reactor treatment unit-aeration volume in the aerobic zone" node and the "A / O biological reactor treatment unit-ammonia-oxidizing bacteria" node reflects the impact of aerobic zone aeration volume on the activity of ammonia-oxidizing bacteria, which in turn affects the degradation of ammonia nitrogen.

[0091] Step S130: constructing a dynamic equilibrium model of the sewage treatment system based on the material conversion association network, wherein the dynamic equilibrium model is used to reflect the parameter matching relationship of each treatment unit of the sewage treatment system under a stable operation state.

[0092] Step S131: collecting multiple sets of material conversion association networks and corresponding parameter data sets of each treatment unit in the sewage treatment system during stable operation, wherein the parameter data sets include stable values ​​of pollutant composition data, operation parameter data and bacterial activity data.

[0093] We selected a municipal sewage treatment plant for multiple consecutive stable operation periods, each lasting a certain length of time. Within each stable operation period, we recorded the state of the material transformation network at regular intervals, including the connection relationships and association strength values ​​of each node.

[0094] At the same time, parameter data of each treatment unit in the corresponding time period are collected, such as the stable value of suspended solids concentration and the stable value of screen machine operation frequency of the screen treatment unit; the stable value of ammonia nitrogen concentration, the stable value of aerobic zone aeration volume, and the stable value of ammonia oxidizing bacteria number of the A / O biological reactor treatment unit, etc., to form a parameter data set.

[0095] The material transformation association network within each stable operation period is associated with the corresponding parameter data set and stored to form multiple groups of training samples.

[0096] Step S132: extracting node features from each group of the substance conversion association network to obtain a feature vector of each node, wherein the feature vector includes the pollutant concentration, operating parameter value, and bacterial activity value corresponding to the node.

[0097] For each node in each group of material conversion association networks, such as the "grid treatment unit-suspended matter" node, the stable value of the suspended matter concentration of the node in the corresponding stable operation period is extracted; for the "A / O biological reactor treatment unit-aeration volume of the aerobic zone" node, the stable value of the aerobic zone aeration volume is extracted; for the "A / O biological reactor treatment unit-ammonia oxidizing bacteria" node, the stable values ​​of the number and metabolic rate of ammonia oxidizing bacteria are extracted.

[0098] These extracted values ​​are arranged in a pre-set order to form a feature vector for each node. For example, the feature vector for the "A / O Bioreactor Treatment Unit - Ammonia Oxidizing Bacteria" node can be composed of the stable value of the ammonia oxidizing bacteria population and the stable value of the metabolic rate, a two-dimensional numerical combination.

[0099] Step S133: aligning the node feature vectors within the same stable operation period with the parameter data set, and establishing a mapping relationship between the node features and the stable parameters.

[0100] During the same stable operation period, the eigenvector of each node is matched to the corresponding parameter value in the parameter dataset. For example, the eigenvector (stable value of ammonia nitrogen concentration) of the "A / O Bioreactor Treatment Unit - Ammonia Nitrogen" node is associated with the stable value of ammonia nitrogen in the A / O Bioreactor Treatment Unit in the parameter dataset; the eigenvector of the "A / O Bioreactor Treatment Unit - Aerobic Zone Aeration Rate" node is associated with the stable value of aerobic zone aeration rate in the parameter dataset.

[0101] Through the above-mentioned association alignment, the correspondence between each value in the node feature vector and the specific parameter in the parameter dataset is clarified, thereby establishing a mapping relationship between node features and stable parameters, ensuring the correspondence between input and output during model training.

[0102] Step S134: A graph neural network algorithm is used to train the node feature vectors and mapping relationships after association alignment to construct an initial dynamic equilibrium model. The initial dynamic equilibrium model can output a corresponding parameter matching relationship based on the input material conversion association network.

[0103] Step S1341: Input the node feature vector after association alignment into the input layer of the graph neural network, and perform dimension conversion on the node feature vector so that the dimension of the node feature vector matches the dimension of the hidden layer of the graph neural network.

[0104] The input layer of a graph neural network receives feature vectors from each node, which may have different dimensions. For example, some nodes may have two-dimensional feature vectors, while others may have three-dimensional feature vectors. The input layer transforms these feature vectors using a matrix transformation, converting them to vectors with the same dimensions as the hidden layer.

[0105] For example, if the hidden layer dimension is a specific value, the input layer will expand the two-dimensional feature vector and compress the three-dimensional feature vector to ensure that all vectors entering the hidden layer have the same dimension, which is convenient for subsequent processing.

[0106] Step S1342: In the hidden layer of the graph neural network, the connection relationship between each node is described by the adjacency matrix, and the node feature vectors are aggregated based on the connection relationship to obtain an aggregated feature vector containing neighbor node information.

[0107] In the hidden layer, an adjacency matrix is ​​constructed based on the structure of the material transformation network. The elements in the adjacency matrix indicate whether there is a connection between nodes and the strength of the connection. For each node, the hidden layer collects the feature vectors of all its neighboring nodes.

[0108] For example, the neighboring nodes of the "A / O Bioreactor Treatment Unit - Ammonia Oxidizing Bacteria" node might include the "A / O Bioreactor Treatment Unit - Aerobic Zone Aeration Rate" node and the "A / O Bioreactor Treatment Unit - Ammonia Nitrogen" node. The hidden layer performs a weighted combination of the feature vectors of these neighboring nodes with the node's own feature vector based on the strength of the association to generate an aggregated feature vector. This aggregated feature vector contains comprehensive information about the node and its neighbors.

[0109] Step S1343: Perform nonlinear activation processing on the aggregated feature vector to generate an activated feature vector, and input the activated feature vector into the output layer of the graph neural network. The activated feature vector is processed through the fully connected layer to output the parameter matching relationship prediction value of each processing unit. The parameter matching relationship prediction value includes the pollutant concentration matching range, the operating parameter matching interval and the bacterial community activity matching interval.

[0110] A nonlinear activation function is applied to the aggregated feature vector to enhance its expressiveness, generating an activated feature vector. This activated feature vector is then fed into the output layer. The fully connected layer multiplies the activated feature vector by a preset weight matrix and adds a bias term to produce the parameter matching relationship prediction.

[0111] For example, for the A / O biological reactor treatment unit, the output parameter matching relationship prediction value may include the ammonia nitrogen concentration matching range, the aeration volume matching interval of the aerobic zone, the ammonia oxidizing bacteria activity matching interval, etc. These intervals represent the reasonable matching range between the parameters under stable operating conditions.

[0112] Step S1344: Calculate the loss value of the parameter matching relationship prediction value and the actual parameter matching relationship, and use the back propagation algorithm to adjust the weight parameters of the graph neural network to minimize the loss value.

[0113] The parameter matching relationship predictions obtained by the output layer are compared with the actual parameter matching relationship after association alignment. The difference between the two is calculated using the loss function to obtain the loss value. The loss function can comprehensively consider the degree of deviation in each parameter matching interval.

[0114] Using the backpropagation algorithm, starting from the output layer, the gradient of the loss value with respect to the weight parameters of each layer is calculated layer by layer. The weight parameters are adjusted according to the gradient direction to reduce the loss value. This process is repeated until the loss value reaches a minimum level.

[0115] Step S1345: Set a training iteration number threshold. When the training iteration number reaches the threshold, stop training and determine the current graph neural network model as the initial dynamic balance model.

[0116] During model training, a threshold for the number of training iterations is set. Each iteration uses a set of training samples for forward computation and backpropagation adjustments. When the number of iterations reaches the threshold, training stops, regardless of whether the loss value has reached the minimum, and the graph neural network model at that point is saved as the initial dynamic equilibrium model.

[0117] Step S135: Selecting a material transformation association network and parameter data set during a portion of stable operation time as a validation set, inputting the initial dynamic equilibrium model, and obtaining parameter matching relationship prediction results.

[0118] From the collected stable operation data, a subset of material transformation networks and their corresponding parameter datasets that were not used in training are randomly selected as the validation set. The material transformation networks in the validation set are input into the initial dynamic equilibrium model, which then outputs the predicted parameter matching results for each processing unit.

[0119] Step S136: Calculate the deviation value between the parameter matching relationship prediction result and the actual parameter matching relationship in the validation set. If the deviation value exceeds the preset deviation threshold, adjust the inter-layer connection weight of the graph neural network algorithm.

[0120] Compare the parameter matching relationship prediction results of the validation set with the actual parameter matching relationship and calculate the deviation between the two. The deviation value can be calculated by comparing the overlap of each parameter matching interval, the center value deviation, etc.

[0121] If the deviation exceeds the preset deviation threshold, the initial dynamic equilibrium model's prediction accuracy is insufficient, and the connection weights between the graph neural network layers need to be adjusted. This adjustment is similar to backpropagation during training, but only the validation set data is used for fine-tuning.

[0122] Step S137: Repeat the model training and parameter adjustment process until the deviation value is less than or equal to the preset deviation threshold, and obtain the final dynamic balance model of the sewage treatment system. The dynamic balance model is used to reflect the parameter matching relationship of each treatment unit of the sewage treatment system under stable operation.

[0123] Apply the adjusted weight parameters to the graph neural network, retrain using the training set, and validate again using the validation set. Repeat this process, adjusting the weight parameters until the deviation value of the validation set is less than or equal to the preset deviation threshold.

[0124] The graph neural network model obtained at this time is the final dynamic equilibrium model, which can accurately reflect the parameter matching relationship between the pollutant concentration, operating parameters, and bacterial activity of each treatment unit in the sewage treatment system under stable operation.

[0125] Step S140: Input the multi-source detection data collected in real time into the dynamic balance model, and analyze the abnormal related nodes that deviate from the equilibrium state and the corresponding abnormal influencing factors. The abnormal related nodes are nodes in the processing unit where parameter imbalance occurs, and the abnormal influencing factors are pollutants or equipment operating parameters that cause parameter imbalance.

[0126] Step S141: performing node mapping processing on the multi-source detection data collected in real time, mapping the data to each node of the material conversion association network, and generating a real-time material conversion association network.

[0127] The central control system receives multi-source detection data uploaded by each processing unit in real time, such as the suspended solids concentration and screen machine operation frequency uploaded by the screen processing unit in real time; the ammonia nitrogen concentration, aeration volume of the aerobic zone, ammonia oxidizing bacteria activity and other data uploaded by the A / O biological reactor processing unit in real time.

[0128] Based on the definition of each node in the material transformation association network, the real-time collected data is mapped to the corresponding node. For example, real-time ammonia nitrogen concentration data is mapped to the "A / O Bioreactor Treatment Unit - Ammonia Nitrogen" node, and real-time aeration rate data in the aerobic zone is mapped to the "A / O Bioreactor Treatment Unit - Aeration Rate in the Aerobic Zone" node, forming a real-time material transformation association network. The node values ​​in this real-time material transformation association network are all real-time data at the current moment.

[0129] Step S142: inputting the real-time material conversion association network into the dynamic equilibrium model to obtain the parameter matching relationship standard value of each processing unit.

[0130] The real-time material conversion association network is input into the dynamic equilibrium model, and the model outputs the standard value of the parameter matching relationship that each processing unit should have under the current operating state based on its internal parameter matching relationship mapping.

[0131] For example, for the A / O biological reactor treatment unit, the standard values ​​of the parameter matching relationship output by the dynamic equilibrium model may include the standard range of ammonia nitrogen concentration, the standard interval of aeration volume in the aerobic zone, the standard range of ammonia oxidizing bacteria activity, etc. These standard values ​​are determined based on the stable operation state of the system.

[0132] Step S143: extracting the real-time parameter value of each node in the real-time material conversion association network, comparing it with the corresponding parameter matching relationship standard value, and calculating the deviation rate.

[0133] The real-time parameter values ​​of each node are extracted from the real-time material conversion association network, such as the real-time concentration value of the "A / O biological reactor treatment unit-ammonia nitrogen" node and the real-time parameter value of the "A / O biological reactor treatment unit-aeration volume of the aerobic zone" node.

[0134] Compare each node's real-time parameter value with the corresponding parameter matching relationship standard value output by the dynamic balance model to calculate the deviation rate. The deviation rate is calculated as (real-time parameter value - standard value center value) / standard value range, where the standard value center value is the midpoint of the parameter matching relationship standard value interval, and the standard value range is the difference between the maximum and minimum values ​​in the interval.

[0135] Step S144: Nodes whose deviation rate exceeds a preset deviation rate threshold are screened out and marked as candidate abnormal associated nodes.

[0136] A preset deviation rate threshold is set to determine the deviation rate of each node. If a node's deviation rate exceeds this threshold, for example, if the real-time concentration value of the "A / O Bioreactor Treatment Unit - Ammonia Nitrogen" node is far above the standard value range and its deviation rate exceeds the preset deviation rate threshold, the node will be marked as a candidate abnormal association node.

[0137] After screening, a set of all candidate abnormal associated nodes is obtained.

[0138] Step S145: performing an impact range analysis on the candidate abnormal associated node, determining the impact degree of the abnormal state of the candidate abnormal associated node on the adjacent nodes, and generating an impact degree score.

[0139] Step S1451: Obtain a list of adjacent nodes of the candidate abnormal associated node in the material conversion associated network, where the adjacent nodes are nodes that have a direct connection relationship with the candidate abnormal associated node.

[0140] Taking the candidate abnormal association node "A / O biological reactor treatment unit-ammonia nitrogen" as an example, search for nodes that have a direct connection with it in the material conversion association network, such as the "A / O biological reactor treatment unit-ammonia oxidizing bacteria" node, the "A / O biological reactor treatment unit-aeration volume of the aerobic zone" node, the "secondary sedimentation tank treatment unit-suspended solids" node, etc., to form a list of adjacent nodes.

[0141] Step S1452: Calculate the correlation coefficient between the deviation rate of the candidate abnormally associated node and the deviation rate of the adjacent node, where the correlation coefficient is used to indicate the degree of synchronization of the deviation rate changes of the two.

[0142] The data sequence of the deviation rate of the candidate abnormal associated node "A / O biological reactor treatment unit-ammonia nitrogen" changing over time is compared with the data sequence of the deviation rate of each adjacent node changing over time.

[0143] The correlation coefficient is calculated by calculating the degree of correlation between two sets of data series. The correlation coefficient ranges from -1 to 1. A value close to 1 indicates a high degree of synchronization in the deviation rate changes between the two sets of data series, while a value close to -1 indicates a low degree of synchronization.

[0144] Step S1453: Calculate an influence diffusion index according to the correlation coefficient and the number of adjacent nodes. The influence diffusion index is positively correlated with the correlation coefficient and the number of adjacent nodes.

[0145] The impact diffusion index is calculated by comprehensively considering the correlation coefficient and the number of adjacent nodes. For each candidate abnormally associated node, the absolute value of the correlation coefficient with each adjacent node is taken, and then the sum is multiplied by the number of adjacent nodes to obtain the impact diffusion index.

[0146] For example, if a candidate abnormal associated node has three adjacent nodes, the corresponding absolute values ​​of the correlation coefficients are 0.8, 0.6, and 0.7, respectively. The sum is 2.1, which is multiplied by the number of adjacent nodes 3 to obtain an influence diffusion index of 6.3.

[0147] Step S1454: extracting the centrality value of the candidate abnormal association node in the material conversion association network, where the centrality value is used to represent the connection importance of the candidate abnormal association node in the material conversion association network.

[0148] The centrality value is determined by the number and strength of connections between the candidate anomaly-associated node and the material transformation network. The greater the number of connections and the stronger the strength of the connections, the higher the centrality value, indicating that the node is more important in the network and has a greater potential impact on other nodes due to its anomaly.

[0149] For example, the "A / O biological reactor treatment unit-ammonia nitrogen" node may be connected to multiple nodes, and its centrality value is relatively high.

[0150] Step S1455: performing weighted summation on the influence diffusion index and the centrality value to obtain an influence degree score, wherein the weight of the weighted summation is preset according to the node type.

[0151] Based on the type of candidate anomaly-related node, such as pollutant node, operating parameter node, or bacterial activity node, preset weights for the impact diffusion index and centrality value. For example, for a pollutant node, the weight of the impact diffusion index might be set to 0.6, and the weight of the centrality value might be set to 0.4.

[0152] Multiply the influence diffusion index by the corresponding weight and add the centrality value multiplied by the corresponding weight to obtain the influence degree score.

[0153] Step S1456: sorting the impact scores from high to low, and selecting a set number of candidate abnormal-related nodes with the highest ranking as key analysis objects.

[0154] Sort the impact scores of all candidate abnormal-related nodes and select the top-ranked nodes as the key analysis objects. The abnormal status of these nodes has a greater impact on the system and needs to be handled first.

[0155] Step S1457: Based on the impact score of the key analysis object, determine its impact level on adjacent nodes.

[0156] The impact of the key analysis object is divided into three levels: high, medium, and low. Scores above a certain range are classified as high impact, indicating that the abnormal state has a significant impact on adjacent nodes; scores within a certain range are classified as medium impact; and scores below a certain range are classified as low impact.

[0157] Step S146: determining abnormal related nodes that deviate from the equilibrium state from the candidate abnormal related nodes according to the impact degree score.

[0158] The candidate abnormal associated nodes whose impact score exceeds the preset impact threshold are determined as abnormal associated nodes. For example, if the score corresponding to the preset impact threshold is 5.0, the candidate abnormal associated node with an impact score of 6.3 is determined as an abnormal associated node.

[0159] These abnormal related nodes are nodes where parameter imbalance occurs in the processing unit, and their abnormal causes need to be further analyzed.

[0160] Step S147: extracting the real-time parameter values ​​and parameter matching relationship standard values ​​corresponding to the abnormal associated nodes, analyzing the parameter type causing the deviation, and determining it as an abnormal influencing factor, wherein the abnormal influencing factor is a pollutant or equipment operating parameter that causes parameter imbalance.

[0161] For the abnormal associated node "A / O biological reactor treatment unit-ammonia nitrogen", extract its real-time ammonia nitrogen concentration value and the ammonia nitrogen concentration standard range in the parameter matching relationship standard value, and analyze the cause of the deviation.

[0162] If it is found that the real-time value of the aeration volume in the aerobic zone is lower than the interval in the standard value of the parameter matching relationship, and the activity of ammonia-oxidizing bacteria is also lower than the standard range, combined with the connection relationship of the nodes in the material conversion association network, it is judged that the aeration volume in the aerobic zone may be the cause of the abnormal ammonia nitrogen concentration, and the aeration volume in the aerobic zone is determined as the abnormal influencing factor.

[0163] If the analysis finds that the abnormal parameters of the downstream treatment unit are caused by the excessively high concentration of a certain pollutant in the upstream water, the pollutant will be identified as the abnormal influencing factor.

[0164] Step S150: generating a diagnosis report including an abnormality processing path and a device control instruction according to the abnormality-related node and the abnormality impact factor, wherein the device control instruction is used to adjust the device operating parameters of the corresponding processing unit to restore system balance.

[0165] Step S151: Retrieve historical processing cases corresponding to the abnormality-related nodes and abnormality influencing factors from a preset fault diagnosis knowledge base, wherein the historical processing cases include abnormality cause analysis, processing path, and equipment adjustment records.

[0166] The fault diagnosis knowledge base stores case studies of various abnormalities that have occurred in municipal sewage treatment plants. Based on the identified abnormality-related nodes (such as "A / O biological reactor treatment unit - ammonia nitrogen") and the abnormality-influencing factors (such as aeration volume in the aerobic zone), the knowledge base is searched to retrieve all relevant historical treatment cases.

[0167] These historical processing cases record in detail the causes of the abnormalities at the time, such as insufficient aeration volume due to a failure of the aeration fan in the aerobic zone; the processing paths, such as repairing the aeration fan and adjusting the aeration volume; and specific records of equipment adjustments, such as adjusting the aeration volume from one value to another.

[0168] Step S152: Perform similarity matching on the historical processing cases, and select the case with the highest similarity to the current abnormal situation as a reference case.

[0169] Step S1521: extract abnormal related node features, abnormal impact factor features and processing unit status features in historical processing cases, and construct a case feature vector.

[0170] For each historical processing case, the type of abnormal related node (such as pollutant node), the processing unit where it is located, the deviation rate range, etc. are extracted as abnormal related node features; the type of abnormal influencing factor (such as operating parameters), the parameter value deviation range, etc. are extracted as abnormal influencing factor features; the operating status parameter range of other related processing units at that time is extracted as the processing unit status feature, such as the operating frequency range of the grid processing unit, the pH value range of the equalization tank processing unit, etc.

[0171] These features are arranged in a preset order to form a case feature vector for each historical case. Each element in the case feature vector corresponds to a specific feature value, which together describe the abnormal situation of the historical case.

[0172] Step S1522: extract the abnormal related node features, abnormal impact factor features and processing unit status features in the current abnormal situation, and construct a current feature vector.

[0173] For the current abnormal situation, the following features are also extracted: the type of abnormal related node, the processing unit it is in, the deviation rate range, etc.; the type of abnormal influencing factor, the parameter value deviation range, etc.; and the operating status parameter range of other current related processing units as processing unit status features.

[0174] Arrange these features in the same order as the case feature vector and construct the current feature vector so that it has the same structure as the case feature vector for similarity comparison.

[0175] Step S1523: Calculate the similarity between the case feature vector and the current feature vector to obtain a similarity score, and sort the historical processing cases in descending order based on the similarity score, and select the historical processing case that ranks first as the preliminary reference case.

[0176] When calculating the similarity between the case feature vector and the current feature vector, the similarity is calculated for each corresponding feature element. For example, for the feature of the processing unit where the abnormal associated node is located, if the historical case and the current situation are the same, the similarity of this feature is 1; if they are different, the similarity is 0.

[0177] For numerical features such as deviation rate ranges and parameter value deviation ranges, the degree of overlap between the two ranges is calculated. The higher the overlap, the higher the similarity. The similarities of all features are weighted and summed to obtain the similarity score between the historical case and the current anomaly.

[0178] The similarity scores of all historical processing cases are sorted from high to low, and the historical processing case ranked first is selected as the preliminary reference case.

[0179] Step S1524: Verify the feasibility of the processing path of the preliminary reference case in the current sewage treatment system, and check whether the equipment and processing units involved in the processing path are consistent with the current sewage treatment system. If feasible, the preliminary reference case is determined as the reference case; if not feasible, the historical processing case with the second highest ranking is selected for feasibility verification until a feasible reference case is found.

[0180] Review the equipment involved in the treatment path of the preliminary reference case, such as the model and quantity of aeration fans, the structure of the treatment unit, etc., and compare them with the equipment and treatment units of the current urban domestic sewage treatment plant.

[0181] If the equipment models, processing unit structures, etc. of the two are consistent, and the operation steps in the processing path can be implemented in the current system, such as the operation of adjusting the aeration volume can be realized in the current control system, then the preliminary reference case is considered feasible and will be determined as the reference case.

[0182] If the equipment involved in the preliminary reference case does not exist in the current system or the processing steps cannot be implemented, the case is not feasible. The historical processing case with the lower ranking is selected to perform the same feasibility verification until a feasible reference case is found.

[0183] Step S153: Based on the processing path of the reference case and combined with the material conversion association network of the current sewage treatment system, an abnormal processing path from the abnormal association node to the normal state is planned. The abnormal processing path includes the processing unit sequence that needs to be adjusted and the parameter adjustment direction.

[0184] The reference case's treatment path might be "check the aeration fan in the A / O bioreactor treatment unit - adjust the aeration rate in the aerobic zone - monitor changes in ammonia nitrogen concentration." By analyzing the connections and impacts between nodes in the current substance transformation network, the reference case's treatment path can be adjusted.

[0185] For example, the current substance transformation network shows that adjusting the aeration rate in the aerobic zone also affects the activity of ammonia-oxidizing bacteria, which in turn affects the degradation of ammonia nitrogen. Therefore, when planning abnormal treatment paths, in addition to adjusting the aeration rate, it is also necessary to add a step to monitor the activity of ammonia-oxidizing bacteria.

[0186] The final abnormality handling path is "check the aeration fan of the A / O biological reactor treatment unit - gradually increase the aeration volume of the aerobic zone to the standard value range of the parameter matching relationship - real-time monitoring of the activity of ammonia-oxidizing bacteria and ammonia nitrogen concentration - when the ammonia nitrogen concentration drops to the standard range, stabilize the aeration volume of the aerobic zone." It clearly states that the treatment unit that needs to be adjusted is the A / O biological reactor treatment unit, and the parameter adjustment direction is to increase the aeration volume of the aerobic zone.

[0187] Step S154: Based on the location of the abnormal associated node and the type of abnormal influencing factor, the core content of the diagnosis report is generated, and the core content includes a description of the abnormal phenomenon, an inference of the abnormal cause, and an explanation of the abnormal processing path.

[0188] The anomaly's associated node is "A / O Bioreactor Treatment Unit - Ammonia Nitrogen," and the influencing factor is the aeration rate in the aerobic zone. The core content of the diagnostic report describes the anomaly as "the real-time ammonia nitrogen concentration in the A / O Bioreactor Treatment Unit is higher than the parameter matching relationship standard value."

[0189] The cause of the anomaly is inferred from combining reference cases and current data, and it is inferred that "the aeration volume in the aerobic zone is lower than the standard range, resulting in insufficient activity of ammonia-oxidizing bacteria and decreased efficiency of ammonia nitrogen degradation."

[0190] The exception handling path description describes the planned exception handling path in detail, including the operation purpose and expected effect of each step.

[0191] Step S155: adding the identifier of the abnormal associated node, the specific parameters of the abnormal influencing factor and the step decomposition of the processing path to the diagnosis report to form a target diagnosis report.

[0192] The abnormal related node is clearly marked in the diagnostic report as "A / O biological reactor treatment unit-ammonia nitrogen", and the specific parameters of the abnormal influencing factor are "the real-time value of aeration volume in the aerobic zone is X, and the standard value range of the parameter matching relationship is YZ."

[0193] The abnormality handling path is broken down into specific steps, such as Step 1: Check the operating status of the aeration fan to confirm whether there is a fault; Step 2: If the fan is normal, gradually increase the aeration volume in the aerobic zone from X to Y through the control system; Step 3: Record the activity of ammonia-oxidizing bacteria and ammonia nitrogen concentration data at fixed intervals; Step 4: When the ammonia nitrogen concentration drops to the standard range, maintain the current aeration volume at a stable operation.

[0194] By adding these contents, a complete and well-organized target diagnosis report is formed.

[0195] Step S156: Determine the device identifier that needs to be regulated and the corresponding parameter adjustment value according to the exception handling path and the parameter adjustment direction in the target diagnosis report.

[0196] The exception handling path involves adjusting the aeration volume in the aerobic zone of the A / O biological reactor treatment unit. The corresponding device is the aeration fan of the treatment unit, and the device identifier is "Aeration Fan-F01".

[0197] According to the parameter matching relationship standard value interval YZ, combined with the current aeration volume X, the parameter adjustment value is determined to adjust the aeration volume from X to a value within the YZ interval. This value is determined based on historical data and the current system status, such as adjusting it to (Y+Z) / 2.

[0198] Step S157: combining the device identification, parameter adjustment value, and adjustment sequence to generate a device control instruction, wherein the device control instruction is used to adjust the device operating parameters of the corresponding processing unit.

[0199] The content of the equipment control instruction is "Equipment identification: Aeration fan-F01; Parameter adjustment value: adjust the aeration volume from X to (Y+Z) / 2; Adjustment sequence: First check the equipment status, and adjust it gradually after there is no fault. Each adjustment is done at a fixed interval to monitor the changes in relevant parameters."

[0200] The equipment control instruction is sent to the equipment control module of the A / O biological reactor processing unit through the central control system. The equipment control module adjusts the operating parameters of the aeration fan according to the equipment control instruction to restore the balance state of the system.

[0201] Figure 2A schematic diagram illustrates exemplary hardware and software components of a sewage treatment detection system 100 with diagnostic capabilities, which can implement the concepts of the present application, according to some embodiments of the present application. For example, a processor 120 can be used in the sewage treatment detection system 100 with diagnostic capabilities to perform the functions described in the present application.

[0202] The sewage treatment detection system 100 with diagnostic functions can be a general-purpose server or a special-purpose server, both of which can be used to implement the sewage treatment detection method with diagnostic functions of the present application. Although only one server is shown in this application, for convenience, the functions described in this application can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0203] For example, the sewage treatment detection system 100 with diagnostic function may include a network port 110 connected to the network, one or more processors 120 for executing program instructions, a communication bus 130, and different forms of storage media 140, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the sewage treatment detection system 100 with diagnostic function may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present application can be implemented according to these program instructions. The sewage treatment detection system 100 with diagnostic function also includes an I / O interface 150 between the computer and other input and output devices.

[0204] For ease of explanation, only one processor is described in the sewage treatment detection system 100 with diagnostic functions. However, it should be noted that the sewage treatment detection system 100 with diagnostic functions in the present application may also include multiple processors, so the steps performed by one processor described in the present application may also be performed jointly or individually by multiple processors. For example, if the processor of the sewage treatment detection system 100 with diagnostic functions executes step A and step B, it should be understood that step A and step B may also be executed jointly by two different processors or individually in one processor. For example, the first processor executes step A, the second processor executes step B, or the first processor and the second processor execute steps A and B together.

[0205] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer-executable instructions are preset. When a processor executes the computer-executable instructions, the above-mentioned sewage treatment detection method with diagnostic function is implemented.

[0206] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, multiple features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. A sewage treatment detection method with diagnostic function, characterized in that: The method comprises: Receive multi-source detection data uploaded by each treatment unit in the sewage treatment process, the multi-source detection data including the pollutant composition data of the sewage, the operating parameter data of the treatment equipment and the bacterial activity data in the reaction tank; Performing cross-unit correlation mapping processing on the multi-source detection data to generate a material transformation correlation network between the processing units, wherein the material transformation correlation network is used to reflect the mutual influence relationship between pollutants, equipment operating parameters and bacterial activity in different processing units; Constructing a dynamic equilibrium model of the sewage treatment system based on the material transformation association network, wherein the dynamic equilibrium model is used to reflect the parameter matching relationship of each treatment unit of the sewage treatment system under a stable operation state; Inputting multi-source detection data collected in real time into the dynamic balance model, parsing out abnormal associated nodes that deviate from the equilibrium state and corresponding abnormal influencing factors, wherein the abnormal associated nodes are nodes in the processing unit where parameter imbalance occurs, and the abnormal influencing factors are pollutants or equipment operating parameters that cause parameter imbalance; A diagnosis report including an abnormality processing path and an equipment control instruction are generated according to the abnormality associated node and the abnormality influencing factor, and the equipment control instruction is used to adjust the equipment operating parameters of the corresponding processing unit to restore the system balance.

2. The sewage treatment detection method with diagnostic function according to claim 1, characterized in that: The performing cross-unit association mapping processing on the multi-source detection data to generate a material conversion association network between each processing unit includes: Extracting characteristic pollutant indicators from the pollutant composition data of each processing unit from the multi-source detection data, wherein the characteristic pollutant indicators are the types and concentrations of pollutants that change significantly during the processing process; Extracting key operating parameters from the operating parameter data of each treatment unit, wherein the key operating parameters are equipment operating parameters that have a direct impact on the pollutant treatment effect; Extracting the dominant bacterial group index from the bacterial group activity data in the reaction tank, wherein the dominant bacterial group index is the type and activity of the bacterial group that plays a leading role in the pollutant degradation process; Establishing a processing unit association matrix, wherein the rows of the processing unit association matrix represent the preceding processing units, the columns represent the following processing units, and the matrix elements represent the degree of influence of the output material of the preceding processing unit on the input material of the following processing unit; The characteristic pollutant index, the key operating parameter, and the dominant bacterial group index are used as nodes, input into the processing unit association matrix, and the association strength value between the nodes in different processing units is calculated, where the association strength value is a quantitative value of the degree of mutual influence between two nodes; According to the association strength value, each node is connected according to the size of the association strength value to form a preliminary material transformation association network, and the preliminary material transformation association network is optimized, and nodes with association strength values ​​lower than a preset threshold are merged, and nodes and connection relationships with association strength values ​​higher than the preset threshold are retained; The processing unit identifier corresponding to each node is added to the optimized material transformation association network to generate the final material transformation association network between the processing units. The material transformation association network is used to reflect the mutual influence relationship between pollutants, equipment operating parameters and microbial activity in different processing units.

3. The sewage treatment detection method with diagnostic function according to claim 2, characterized in that: The step of extracting key operating parameters from the operating parameter data of each processing unit includes: Obtain all operating parameters and corresponding parameter values ​​in the operating parameter data of each processing unit; Collecting pollutant treatment effect data corresponding to different values ​​of each operating parameter, wherein the pollutant treatment effect data includes pollutant removal rate and pollutant concentration after treatment; Calculating the correlation coefficient between each operating parameter and the pollutant treatment effect data, wherein the correlation coefficient is used to indicate the degree of linear correlation between the two; Screening out operating parameters whose absolute values ​​of correlation coefficients exceed a preset correlation threshold as candidate key operating parameters; Performing a sensitivity analysis on the candidate key operating parameters, calculating the change in pollutant treatment effect when the parameter value changes by a unit, and obtaining a sensitivity coefficient; Sorting the sensitivity coefficients in descending order, and selecting the operating parameters with the highest ranking as key operating parameters, wherein the key operating parameters are equipment operating parameters that have a direct impact on the pollutant treatment effect; Also, the dominant bacterial index in the bacterial activity data in the extraction reaction pool, wherein the dominant bacterial index is the type and activity of the bacterial group that plays a leading role in the pollutant degradation process, including: Obtain all bacterial species and corresponding activity values ​​from the bacterial activity data in the reaction pool, wherein the activity values ​​include bacterial population number and metabolic rate; Analyze the degradation correlation between each bacterial species and characteristic pollutant indicators, where the degradation correlation is calculated by the ratio of the degradation rate of the characteristic pollutant in the presence of the bacterial species to the degradation rate in the absence of the bacterial species; Screen out bacterial species with degradation correlation exceeding the preset correlation threshold as candidate dominant bacterial species, and calculate the proportion of the activity value of each candidate dominant bacterial species in the total activity value of all bacterial species to obtain the activity proportion; Determining the candidate dominant bacterial group whose activity ratio exceeds a preset activity ratio threshold as a dominant bacterial group type, and extracting the activity value corresponding to the dominant bacterial group type as dominant bacterial group activity data; The dominant bacterial species and the corresponding dominant bacterial activity data are combined to obtain the dominant bacterial index in the bacterial activity data in the reaction tank. The dominant bacterial index is the bacterial species and activity that play a leading role in the pollutant degradation process.

4. The sewage treatment detection method with diagnostic function according to claim 1, characterized in that: The dynamic equilibrium model of the sewage treatment system is constructed based on the material transformation association network, including: Collecting multiple sets of material conversion association networks and corresponding parameter data sets of each treatment unit of the sewage treatment system during stable operation, wherein the parameter data sets include stable values ​​of pollutant composition data, operating parameter data, and bacterial activity data; Extracting node features from each group of the substance conversion association network to obtain a feature vector of each node, wherein the feature vector includes the pollutant concentration, operating parameter value, and bacterial activity value corresponding to the node; The node feature vectors within the same stable operation period are associated and aligned with the parameter data set to establish a mapping relationship between node features and stable parameters; A graph neural network algorithm is used to train the node feature vectors and mapping relationships after association alignment to construct an initial dynamic equilibrium model. The initial dynamic equilibrium model can output the corresponding parameter matching relationship according to the input material transformation association network; Selecting a material transformation association network and parameter data set during a portion of stable operation as a validation set, inputting the initial dynamic equilibrium model, and obtaining parameter matching relationship prediction results; Calculate the deviation between the parameter matching relationship prediction result and the actual parameter matching relationship in the validation set. If the deviation exceeds a preset deviation threshold, adjust the inter-layer connection weights of the graph neural network algorithm. Repeat the model training and parameter adjustment process until the deviation value is less than or equal to the preset deviation threshold, and obtain the final dynamic equilibrium model of the sewage treatment system. The dynamic equilibrium model is used to reflect the parameter matching relationship of each treatment unit of the sewage treatment system under stable operation.

5. The sewage treatment detection method with diagnostic function according to claim 4, characterized in that: The graph neural network algorithm is used to train the node feature vectors and mapping relationships after association alignment to construct an initial dynamic balance model, including: Inputting the node feature vector after association alignment into the input layer of the graph neural network, and performing dimension conversion on the node feature vector so that the dimension of the node feature vector matches the dimension of the hidden layer of the graph neural network; In the hidden layer of the graph neural network, the connection relationship between each node is described by an adjacency matrix, and the node feature vectors are aggregated based on the connection relationship to obtain an aggregated feature vector containing neighbor node information; Performing nonlinear activation processing on the aggregated feature vector to generate an activated feature vector, and inputting the activated feature vector into the output layer of the graph neural network. The activated feature vector is processed by a fully connected layer to output a parameter matching relationship prediction value of each processing unit, wherein the parameter matching relationship prediction value includes a pollutant concentration matching range, an operating parameter matching interval, and a bacterial community activity matching interval; Calculating the loss value between the predicted value of the parameter matching relationship and the actual parameter matching relationship, and using the back propagation algorithm to adjust the weight parameters of the graph neural network to minimize the loss value; Set a training iteration threshold. When the training iteration threshold is reached, stop training and determine the current graph neural network model as the initial dynamic equilibrium model.

6. The sewage treatment detection method with diagnostic function according to claim 1, characterized in that: The step of inputting the multi-source detection data collected in real time into the dynamic balance model and analyzing the abnormal associated nodes deviating from the equilibrium state and the corresponding abnormal impact factors includes: Perform node mapping processing on the multi-source detection data collected in real time, correspond the data to each node of the material transformation association network, and generate a real-time material transformation association network; Inputting the real-time material conversion association network into the dynamic equilibrium model to obtain the parameter matching relationship standard value of each processing unit; Extracting the real-time parameter value of each node in the real-time material conversion association network, comparing it with the corresponding parameter matching relationship standard value, and calculating the deviation rate; Nodes whose deviation rate exceeds the preset deviation rate threshold are screened out and marked as candidate abnormal association nodes; Performing an impact range analysis on the candidate abnormal associated node to determine the degree of impact of the abnormal state of the candidate abnormal associated node on adjacent nodes, and generating an impact degree score; Determining abnormal associated nodes that deviate from a balanced state from candidate abnormal associated nodes according to the impact degree score; The real-time parameter values ​​and parameter matching relationship standard values ​​corresponding to the abnormal associated nodes are extracted, the parameter types causing the deviations are analyzed, and the abnormal influencing factors are determined to be pollutants or equipment operating parameters that cause parameter imbalance.

7. The sewage treatment detection method with diagnostic function according to claim 6, characterized in that: The performing of an impact range analysis on the candidate abnormal associated node, determining the degree of impact of the abnormal state of the candidate abnormal associated node on adjacent nodes, and generating an impact degree score includes: Obtaining a list of adjacent nodes of the candidate abnormal association node in the material conversion association network, wherein the adjacent nodes are nodes that have a direct connection relationship with the candidate abnormal association node; Calculating a correlation coefficient between the deviation rate of the candidate abnormally associated node and the deviation rate of the adjacent node, wherein the correlation coefficient is used to indicate the degree of synchronization of the deviation rate changes of the two; Calculating an influence diffusion index based on the correlation coefficient and the number of adjacent nodes, wherein the influence diffusion index is positively correlated with the correlation coefficient and the number of adjacent nodes; Extracting a centrality value of a candidate abnormal association node in a material conversion association network, wherein the centrality value is used to represent the connection importance of the candidate abnormal association node in the material conversion association network; Performing a weighted summation of the influence diffusion index and the centrality value to obtain an influence degree score, wherein the weight of the weighted summation is preset according to the node type; Sort the impact scores from high to low, and select a set number of candidate abnormal-related nodes that are ranked high as key analysis objects; Based on the impact score of the key analysis object, the level of its impact on adjacent nodes is determined.

8. The sewage treatment detection method with diagnostic function according to claim 1, characterized in that: The generating of a diagnostic report including an abnormality handling path and a device control instruction according to the abnormality associated node and the abnormality impact factor includes: Retrieving historical processing cases corresponding to the abnormality-related nodes and abnormality influencing factors from a preset fault diagnosis knowledge base, wherein the historical processing cases include abnormality cause analysis, processing path, and equipment adjustment records; Perform similarity matching on the historical processing cases and select the case with the highest similarity to the current abnormal situation as a reference case; Based on the processing path of the reference case and the material transformation association network of the current sewage treatment system, an abnormal processing path from the abnormal association node to the normal state is planned. The abnormal processing path includes the order of processing units that need to be adjusted and the parameter adjustment direction; Based on the location of the abnormal associated node and the type of abnormal influencing factor, the core content of the diagnostic report is generated, which includes a description of the abnormal phenomenon, an inference of the abnormal cause, and an explanation of the abnormal processing path; Adding the identification of abnormal related nodes, specific parameters of abnormal influencing factors and step decomposition of processing paths to the diagnostic report to form a target diagnostic report; Determine the device identifier and corresponding parameter adjustment value that need to be regulated based on the exception handling path and the parameter adjustment direction in the target diagnostic report; The device identification, parameter adjustment value and adjustment sequence are combined to generate a device control instruction, which is used to adjust the device operating parameters of the corresponding processing unit.

9. The sewage treatment detection method with diagnostic function according to claim 8, characterized in that: The similarity matching of the historical processing cases is performed to select the case with the highest similarity to the current abnormal situation as a reference case, including: Extract abnormal related node features, abnormal impact factor features and processing unit status features from historical processing cases to construct case feature vectors; Extract the abnormal related node features, abnormal influencing factor features and processing unit status features in the current abnormal situation to construct the current feature vector; Calculate the similarity between the case feature vector and the current feature vector to obtain a similarity score, and sort the historical processing cases in descending order based on the similarity scores, and select the historical processing case that ranks first as the preliminary reference case; Verify the feasibility of the treatment path of the preliminary reference case in the current sewage treatment system, check whether the equipment and treatment units involved in the treatment path are consistent with the current sewage treatment system. If feasible, the preliminary reference case will be determined as the reference case; if not feasible, select the historical treatment case with the second highest ranking for feasibility verification until a feasible reference case is found.

10. A sewage treatment detection system with diagnostic function, characterized in that: It includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the sewage treatment detection method with diagnostic function as described in any one of claims 1 to 9.

Images