Impact mode diagnosis method, device and equipment of wastewater treatment reactor and storage medium

Abstract

The invention discloses an impact mode diagnosis method, device and equipment of a wastewater treatment reactor and a storage medium, and relates to the technical field of wastewater treatment.The impact mode diagnosis method comprises the steps that a target wastewater treatment reactor is used for treating wastewater, and a treated water sample is obtained; the method comprises the following steps: determining water sample component organic matter content and water sample component alkalinity based on each treated water sample, collecting online data when each wastewater is treated, and then establishing a characteristic change fingerprint database based on each online data, the water sample component organic matter content and the water sample component alkalinity; training the initial deep learning model by using the feature change fingerprint database so as to output offline data based on the online data by using the obtained target deep learning model; and clustering the offline data and the online data by using a preset clustering algorithm so as to diagnose the impact mode of the target wastewater treatment reactor by using a target expert diagnosis system constructed based on a clustering result. This makes it possible to improve the efficiency of diagnosing the impact pattern of the wastewater treatment reactor.

Description

Technical Field

[0001] The present invention relates to the technical field of wastewater treatment, and in particular to an impact mode diagnosis method, device, equipment and storage medium of a wastewater treatment reactor. Background Art

[0002] Driven by the "plastic ban", the market prospects of biodegradable plastics are relatively broad. With the increase in the output of biodegradable plastics, the corresponding treatment and discharge of wastewater generated in the production process has also become an urgent problem to be solved. Among them, production wastewater mainly includes three streams of stripping tower esterification wastewater, leaching tower wastewater and filter element cleaning wastewater. Its components are complex, with low pH (potential of hydrogen, i.e. hydrogen ion potential or hydrogen ion concentration index), high pollutant concentration and strong toxicity.

[0003] At present, UASB (Upflow Anaerobic Sludge Blanket, a high-efficiency anaerobic bioreactor) anaerobic biological treatment technology has been widely used in the treatment of high-concentration organic wastewater due to its advantages of high organic load, low cost, and low sludge production. However, due to the characteristics of low pH, high concentration and strong toxicity of acidic organic wastewater, it will have multiple impacts on the anaerobic digestion process, resulting in unstable reactor operation and even system collapse. Therefore, when treating acidic organic wastewater, timely diagnosis and identification of various impact modes on wastewater treatment reactors are of great significance to ensure the stability and economy of engineering applications.

[0004] As can be seen from the above, how to improve the efficiency of diagnosing the shock mode of a wastewater treatment reactor during the shock mode diagnosis process of a target wastewater treatment reactor is a problem that needs to be solved urgently. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a method, device, equipment and storage medium for diagnosing the impact mode of a wastewater treatment reactor, which can improve the efficiency of diagnosing the impact mode of a wastewater treatment reactor during the process of diagnosing the impact mode of a target wastewater treatment reactor. The specific scheme is as follows:

[0006] In a first aspect, the present application provides a method for diagnosing an impact mode of a wastewater treatment reactor, comprising:

[0007] A target wastewater treatment reactor is used to treat a number of wastewaters input through a number of preset pipelines to obtain corresponding treated water samples; the wastewater concentrations, wastewater components and wastewater acidity corresponding to each of the wastewaters are different;

[0008] Determine the organic matter content and alkalinity of the corresponding water sample components based on each of the treated water samples, and use the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor to collect the corresponding online data when each of the wastewaters is treated, and then establish a characteristic change fingerprint database based on each of the online data, the organic matter content of the water sample components and the alkalinity of the water sample components;

[0009] Using the information in the feature change fingerprint database to train an initial deep learning model to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data;

[0010] The offline data and the online data are clustered using a preset clustering algorithm, and a target expert diagnosis system is constructed based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

[0011] Optionally, the step of using the target wastewater treatment reactor to treat a plurality of wastewaters input through a plurality of preset pipelines to obtain corresponding treated water samples includes:

[0012] Prepare a plurality of wastewaters, and then input each of the wastewaters into a target wastewater treatment reactor through a plurality of preset pipelines at a preset flow rate for treatment;

[0013] A preset time relay is used to control a plurality of preset sampling pumps, so that each of the preset sampling pumps is used to sample each treatment result output by the target wastewater treatment reactor according to a preset sampling frequency to obtain a corresponding treated water sample.

[0014] Optionally, determining the organic matter content of the corresponding water sample components based on each of the treated water samples includes:

[0015] Using the preset organic matter content detection rules, the organic matter content of the water sample components corresponding to each of the treated water samples is detected respectively to obtain the organic matter content of the corresponding water sample components;

[0016] The water sample components corresponding to each of the treated water samples are titrated using a preset titration system to obtain corresponding titration results, and the alkalinity of the water samples corresponding to each of the water sample groups is determined based on each of the titration results.

[0017] Optionally, the collecting of online data corresponding to each wastewater being treated respectively by using a gas phase sensor and a liquid phase sensor in the target wastewater treatment reactor includes:

[0018] The gas phase sensor is used to detect the methane concentration, carbon dioxide concentration and hydrogen concentration in the gas path in the target wastewater treatment reactor to obtain the corresponding methane concentration, carbon dioxide concentration and hydrogen concentration, and then the gas flow meter is used to detect the instantaneous gas flow in the gas path to obtain the instantaneous gas flow result; the gas phase sensor is located in the gas path, and the gas flow meter is located at the end of the gas path;

[0019] Determine the methane production, the carbon dioxide production and the hydrogen production based on the methane concentration, the carbon dioxide concentration, the hydrogen concentration and the corresponding instantaneous gas flow results respectively;

[0020] The liquid phase sensor is used to collect the pH, conductivity and redox potential of the target wastewater treatment reactor to obtain the pH inside the reactor, the conductivity inside the reactor and the redox potential inside the reactor.

[0021] Optionally, the using the information in the feature change fingerprint database to train the initial deep learning model to obtain a target deep learning model, and using the target deep learning model to output corresponding offline data based on the online data, includes:

[0022] Using a first preset memory subnetwork to process each of the online data in the feature change fingerprint database to obtain corresponding first data to be pooled, and using a first preset pooling layer to process each of the first data to be pooled respectively to obtain corresponding first data to be processed;

[0023] Using a second preset memory subnetwork to process each of the first to-be-processed data to obtain corresponding second to-be-pooled data, and using a second preset pooling layer to process each of the second to-be-pooled data to obtain corresponding second to-be-processed data;

[0024] Using a preset memory unit, a preset input gate, a preset forget gate, and a preset output gate to process each of the second data to be processed, respectively, to obtain corresponding data to be regressed;

[0025] Each of the data to be regressed is sent to a preset fully connected layer, so as to use the preset fully connected layer to perform regression analysis on each of the data to be regressed to obtain offline data; the initial deep learning model includes the first preset memory subnetwork, the second preset memory subnetwork, the preset memory unit, the preset input gate, the preset forget gate and the preset output gate; the offline data includes the chemical oxygen demand of the inlet and outlet water, the components of each pollutant and the concentration of each pollutant.

[0026] Optionally, the diagnosing the impact mode of the target wastewater treatment reactor by using the target expert diagnosis system includes:

[0027] Based on the target expert diagnosis system and using the preset clustering algorithm, the offline data and the online data are clustered to obtain a clustering result; the clustering result includes a plurality of clusters, each of which corresponds to a different impact mode;

[0028] A number of rule codes are obtained from a preset diagnostic knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each of the rule codes and the clustering results, and the impact mode corresponding to the target wastewater treatment reactor is returned to the current user interface; the impact mode includes a no-impact mode, a load impact mode, a toxic pollutant impact mode and a temperature impact mode.

[0029] Optionally, before acquiring a plurality of rule codes from a preset diagnosis knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each of the rule codes and the clustering result, the method further includes:

[0030] Constructing diagnostic indicators corresponding to the load shock mode, the toxic pollutant shock mode and the temperature shock mode respectively based on preset experience, and constructing a preset diagnostic knowledge base based on each of the diagnostic indicators;

[0031] Wherein, when the impact mode corresponding to the target wastewater treatment reactor is the load impact mode, the pH value and carbon dioxide concentration inside the reactor are both less than the corresponding diagnostic indicators, the conductivity, methane concentration and effluent alkalinity inside the reactor are not greater than the corresponding diagnostic indicators, and the oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand inside the reactor are greater than the corresponding diagnostic indicators;

[0032] When the impact mode corresponding to the target wastewater treatment reactor is the toxic pollutant impact mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all less than the corresponding diagnostic indicators respectively, and the numerical decrease rate corresponding to the internal conductivity of the reactor and the effluent alkalinity is greater than the preset decrease rate threshold; the internal oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand of the reactor are respectively greater than the corresponding diagnostic indicators; the numerical increase rate corresponding to the effluent volatile fatty acids is greater than the preset increase rate threshold;

[0033] When the shock mode corresponding to the target wastewater treatment reactor is the temperature shock mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all lower than the corresponding diagnostic indicators respectively; the internal oxidation-reduction potential of the reactor, the effluent volatile fatty acids and the effluent chemical oxygen demand are respectively higher than the corresponding diagnostic indicators; the increase in the numerical value of the effluent chemical oxygen demand is lower than the preset increase threshold; the hydrogen concentration and the effluent volatile fatty acids are not lower than the corresponding diagnostic indicators.

[0034] In a second aspect, the present application provides an impact mode diagnostic device for a wastewater treatment reactor, comprising:

[0035] A treated water sample acquisition module is used to use a target wastewater treatment reactor to treat a number of wastewaters input through a number of preset pipelines to obtain corresponding treated water samples; the wastewater concentrations, wastewater components and wastewater acidity corresponding to each of the wastewaters are different;

[0036] An online data determination module is used to determine the organic matter content of the corresponding water sample components and the alkalinity of the water sample components based on each of the treated water samples, and to collect the corresponding online data when each of the wastewaters is treated respectively using the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor, and then to establish a characteristic change fingerprint database based on each of the online data, the organic matter content of the water sample components and the alkalinity of the water sample components;

[0037] An offline data determination module, used to train an initial deep learning model using the information in the feature change fingerprint database to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data;

[0038] The impact mode determination module is used to cluster the offline data and the online data using a preset clustering algorithm, and to construct a target expert diagnosis system based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

[0039] In a third aspect, the present application provides an electronic device, including:

[0040] Memory, used to store computer programs;

[0041] The processor is used to execute the computer program to implement the aforementioned impact mode diagnosis method for the wastewater treatment reactor.

[0042] In a fourth aspect, the present application provides a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the aforementioned impact mode diagnosis method for a wastewater treatment reactor.

[0043] As can be seen from the above, before the impact mode diagnosis of the target wastewater treatment reactor is performed, the present application needs to use the target wastewater treatment reactor to treat several wastewaters input through several preset pipes to obtain corresponding treated water samples; the wastewater concentration, wastewater components and wastewater acidity corresponding to each of the wastewaters are different; based on each of the treated water samples, the corresponding water sample component organic matter content and water sample component alkalinity are determined, and the gas phase sensor and liquid phase sensor in the target wastewater treatment reactor are used to collect the corresponding online data when each of the wastewaters is treated, and then a feature change fingerprint database is established based on each of the online data, the organic matter content of the water sample components and the alkalinity of the water sample components; the initial deep learning model is trained using the information in the feature change fingerprint database to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data; the offline data and the online data are clustered using a preset clustering algorithm, and a target expert diagnosis system is constructed based on the obtained clustering results, so as to use the target expert diagnosis system to diagnose the impact mode of the target wastewater treatment reactor.

[0044] It can be seen that the present application firstly processes several wastewaters input through several preset pipelines by using the target wastewater treatment reactor to obtain the corresponding treated water samples; then, the corresponding water sample component organic matter content and water sample alkalinity are determined based on each treated water sample, and the gas phase sensor and liquid phase sensor in the target wastewater treatment reactor are used to collect the corresponding online data when each wastewater is treated separately, and then a feature change fingerprint database is established based on each online data, the organic matter content of the water sample component and the alkalinity of the water sample; then, the initial deep learning model is trained using the information in the feature change fingerprint database to obtain the target deep learning model, so as to use the target deep learning model and output the corresponding offline data based on the online data; finally, the initial expert diagnosis system is trained using the offline data and the online data to obtain the target expert diagnosis system, so as to use the target expert diagnosis system to diagnose the impact mode of the target wastewater treatment reactor. In this way, the efficiency of diagnosing the impact mode of the wastewater treatment reactor is improved, thereby improving the stability and safety of the production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0046] Figure 1A flow chart of a method for diagnosing an impact mode of a wastewater treatment reactor disclosed in the present application;

[0047] Figure 2 A schematic diagram of the model structure of a specific deep learning model disclosed in this application;

[0048] Figure 3 A specific schematic diagram of a rule encoding corresponding to a specific impact mode disclosed in this application;

[0049] Figure 4 A schematic diagram of a specific clustering algorithm for diagnosing impact patterns disclosed in this application;

[0050] Figure 5 A schematic diagram of the structure of an impact mode diagnostic device for a wastewater treatment reactor disclosed in the present application;

[0051] Figure 6 This is a structural diagram of an electronic device disclosed in this application. DETAILED DESCRIPTION

[0052] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] At present, due to the advantages of UASB anaerobic biological treatment technology such as high organic load, low cost and low sludge production, it has been widely used in the treatment of high-concentration organic wastewater. However, due to the characteristics of low pH, high concentration and strong toxicity of acidic organic wastewater, it will have multiple impacts on the anaerobic digestion process, resulting in unstable operation of the reactor and even causing system collapse. Therefore, when treating acidic organic wastewater, timely diagnosis and identification of various impact modes on wastewater treatment reactors are of great significance to ensure the stability and economy of engineering applications. To this end, the present application provides a method for diagnosing the impact mode of a wastewater treatment reactor, which can improve the efficiency of diagnosing the impact mode of a wastewater treatment reactor during the impact mode diagnosis process of a target wastewater treatment reactor.

[0054] See also Figure 1 As shown, an embodiment of the present invention discloses a method for diagnosing an impact mode of a wastewater treatment reactor, comprising:

[0055] Step S11, using a target wastewater treatment reactor to treat a number of wastewaters input through a number of preset pipelines to obtain corresponding treated water samples; the wastewater concentrations, wastewater components and wastewater acidity corresponding to each of the wastewaters are different.

[0056] In actual processes, wastewater treatment faces various types of shocks, such as an increase in single or multiple pollutant loads, a sharp drop in pH, temperature fluctuations, etc. In order to simulate the complex situations in actual processes, the embodiment of the present application is designed to add a single or combined pollutant load to the influent entering the target wastewater treatment reactor, reduce the pH value of the influent, and shock the target wastewater treatment reactor through temperature adjustment, so as to more realistically reflect the actual working conditions.

[0057] In this embodiment, it is necessary to prepare acidic organic wastewater of different concentrations and components, so as to use the prepared wastewater to impact the UASB reactor in different modes, that is, to flow a number of prepared wastewaters into the UASB reactor at a preset flow rate through a feed pipe, so as to impose a specific, controllable external interference or change on the target wastewater treatment reactor, in order to simulate various abnormal conditions or fluctuations that may be encountered in the actual wastewater treatment process, and to detect the online data corresponding to the process of wastewater treatment in the target wastewater treatment reactor. Specifically, using the target wastewater treatment reactor to treat a number of wastewaters input through a number of preset pipes to obtain corresponding treated water samples can include: preparing a number of wastewaters, and then inputting each wastewater into the target wastewater treatment reactor for treatment at a preset flow rate and through a number of preset pipes.

[0058] Furthermore, after obtaining the treatment results of the target wastewater treatment reactor for the wastewater, the embodiment of the present application needs to use a sampling pump controlled by a time relay to perform regular sampling of the effluent output by the target wastewater treatment reactor after treating the specified acidic organic wastewater, and obtain a number of treated water samples. That is, multiple sampling pumps are controlled by a time relay so that the sampling pumps sample equal amounts of effluent water at specified time points. Among them, the sampling frequency is set according to the duration of each impact to ensure that representative effluent samples are obtained at key time nodes, so that the sampling interval and frequency can be flexibly adjusted to meet the experimental requirements under different impact modes and improve the accuracy of sampling and the integrity of data.

[0059] Step S12: determine the organic matter content and alkalinity of the corresponding water sample components based on each of the treated water samples, and use the gas phase sensor and liquid phase sensor in the target wastewater treatment reactor to collect the corresponding online data when each of the wastewaters is treated, and then establish a characteristic change fingerprint database based on the online data, the organic matter content of the water sample components and the alkalinity of the water sample components.

[0060] In this embodiment, after obtaining a number of treated water samples, the embodiment of the present application needs to perform characteristic parameter detection and organic content detection on the water sample components corresponding to each treated water sample, so as to obtain the characteristic values ​​and various organic content corresponding to each water sample component. Among them, the embodiment of the present application adopts a preset titration system to perform alkalinity titration and VFA (Volatile Fatty Acids) titration on the water sample components of each treated water sample, that is, slowly drip a solution of known concentration into the solution of the substance to be tested, and observe the changes in the reaction at the same time. When the reaction reaches the predetermined endpoint, record the volume of the standard solution used. Based on this volume and the known concentration of the standard solution, the content of the substance to be tested can be calculated. Specifically, determining the organic matter content of the corresponding water sample components based on each treated water sample may include: using a preset organic matter content detection rule to respectively detect the organic matter content of the water sample components corresponding to each treated water sample to obtain the corresponding organic matter content of the water sample components; using a preset titration system to titrate the water sample components corresponding to each treated water sample to obtain corresponding titration results, and determining the water sample alkalinity corresponding to each water sample group based on each titration result.

[0061] In addition, the embodiment of the present application needs to install a preset gas phase sensor and a preset liquid phase sensor on the target wastewater treatment reactor, so as to use the preset gas phase sensor and the preset liquid phase sensor to collect online data during the operation of the reactor. Specifically, using the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor to collect the corresponding online data when each wastewater is treated separately can include: using the gas phase sensor to detect the methane concentration, carbon dioxide concentration and hydrogen concentration in the gas path of the target wastewater treatment reactor to obtain the corresponding methane concentration, carbon dioxide concentration and hydrogen concentration, and then using the gas flow meter to detect the instantaneous gas flow in the gas path to obtain the instantaneous gas flow result; the gas phase sensor is located in the gas path, and the gas flow meter is located at the end of the gas path; based on the methane concentration, carbon dioxide concentration, hydrogen concentration and the corresponding instantaneous gas flow results, the methane production, carbon dioxide production and hydrogen production are determined; using the liquid phase sensor to collect the pH, conductivity and redox potential of the target wastewater treatment reactor to obtain the internal pH of the reactor, the internal conductivity of the reactor and the internal redox potential of the reactor. In a specific embodiment, the wastewater treatment reactor may be a UASB reactor, which is not specifically limited herein.

[0062] In this embodiment, the gas generated when the wastewater is anaerobic digested by the target wastewater treatment reactor will enter the gas path after the solid sludge and the liquid effluent are separated by a three-phase separator inside the UASB reactor.

[0063] It is worth mentioning that the preset gas phase sensor is installed in the gas circuit, and the end of the gas circuit is equipped with a preset gas flow meter for collecting instantaneous gas flow. In a specific embodiment, the preset gas phase sensor and the preset gas flow meter both use the 485 communication protocol. Among them, the preset gas phase sensor can detect the concentrations of methane, carbon dioxide and hydrogen in the gas circuit, respectively. Subsequently, the methane concentration, carbon dioxide concentration and hydrogen concentration detected by the preset gas phase sensor are combined with the instantaneous gas flow measured by the gas flow meter to determine the methane production, carbon dioxide production and hydrogen production.

[0064] Furthermore, within one impact cycle, the above gas production is calculated by integrating the instantaneous gas production, and the calculation formula is as follows:

[0065] Q 累计 =q×Δt;

[0066] Among them, Q 累计 is the gas production, q is the instantaneous flow collected by the preset gas flow meter, and Δt is the sampling time interval. It is worth mentioning that when facing the need to process large amounts of data, the embodiment of the present application can realize the automatic calculation of the cumulative gas production through programming.

[0067] Furthermore, the preset liquid phase sensor is installed in the upper area near the effluent of the target wastewater treatment reactor to monitor the liquid phase parameters such as pH value, ORP (Oxidation-Reduction Potential) value and conductivity in the reactor in real time. Among them, the data transmission method adopted by the preset liquid phase sensor is a standard current signal to ensure the stability and accuracy of the measurement results.

[0068] In this embodiment, after obtaining the online data, various organic matter contents and alkalinity corresponding to different impact modes, the embodiment of the present application needs to establish a diagnostic fingerprint database corresponding to different impact modes based on the online data, the organic matter content of each water sample component and the alkalinity of each water sample component.

[0069] Step S13: Use the information in the feature change fingerprint database to train the initial deep learning model to obtain a target deep learning model, and use the target deep learning model to output corresponding offline data based on the online data.

[0070] Subsequently, the diagnostic fingerprint database needs to be preprocessed. Since there are many influencing factors in the data collection process, which may cause partial missing of the data set, the presence of abnormal values, or different dimensions and ranges of different fingerprint feature values, the embodiment of the present application needs to perform data preprocessing on the collected online data, and then select a suitable preprocessing method based on the diagnostic fingerprint database and the corresponding functional requirements of the diagnostic fingerprint database.

[0071] First, the online data is normalized, that is, the data of different unit dimensions are uniformly mapped to the interval (0, 1). The mathematical expression is as follows:

[0072] ;

[0073] in, is the standardized data, that is, the transformed eigenvalue; is the original data, that is, the original value of a feature; is the standard deviation corresponding to the above characteristics, that is, the degree of dispersion of the data, which measures the variability of the data; is the mean of the above characteristics, that is, the average value of all water samples on the above characteristics.

[0074] Further, in a specific embodiment, the pH value of the reactor is in the range of 0-14; the unit of the conductivity of the reactor is us / cm, and the value range is 1-5000; the unit of the ORP of the reactor is mV, and the value range is -2000-2000; the methane concentration, carbon dioxide concentration, and hydrogen concentration are all percentages, and the range is 0-100%; the unit of the effluent alkalinity and the effluent volatile fatty acid is mg / L (calculated as CaCO3). The unit of COD (Chemical Oxygen Demand) is mg / L, and the range is 200-13000mg / L.

[0075] Furthermore, the online data is input into the pre-deep learning model to train the pre-deep learning model, and the obtained standard deep learning model is used as the input feature of the online data to obtain the corresponding offline data. It is worth mentioning that due to the different sampling frequencies of each sensor, when the time steps are inconsistent, the embodiment of the present application needs to process the online data to align the time series of the online data. In addition, in order to process data with different time steps, you can choose to create an independent sub-network for processing the online data collected by each sensor, and merge the processing results in subsequent steps. That is, the online data collected by each sensor is processed through an LSTM (Long Short-Term Memory, i.e., long short-term memory network) sub-network, and the connection layer is used to merge the outputs of all sub-networks.

[0076] Specifically, using the information in the feature change fingerprint database to train the initial deep learning model to obtain the target deep learning model, and using the target deep learning model to output the corresponding offline data based on the online data, can include: using the first preset memory sub-network to process each online data in the feature change fingerprint database to obtain the corresponding first to-be-pooled data, and using the first preset pooling layer to process each first to-be-pooled data respectively to obtain the corresponding first to-be-processed data; using the second preset memory sub-network to process each first to-be-processed data to obtain the corresponding second to-be-pooled data, and using the second preset pooling layer to process each first to-be-processed data respectively to obtain the corresponding second to-be-pooled data. Each second data to be pooled is processed to obtain corresponding second data to be processed; each second data to be processed is processed by using a preset memory unit, a preset input gate, a preset forget gate and a preset output gate to obtain corresponding data to be regressed; each data to be regressed is sent to a preset fully connected layer, so as to perform regression analysis on each data to be regressed by using the preset fully connected layer to obtain offline data; the initial deep learning model includes a first preset memory subnetwork, a second preset memory subnetwork, a preset memory unit, a preset input gate, a preset forget gate and a preset output gate; the offline data includes chemical oxygen demand of inlet and outlet water, components of each pollutant and concentration of each pollutant.

[0077] In a specific implementation, the model structure diagram of the target deep learning model is as follows: Figure 2 As shown: Among them, the convolution kernel size of the first convolution layer is 3×3, the stride is 1, the number of input channels is 6, and the number of output channels is 32. After the first convolution, the output tensor is (batch_size, time_steps -2, 32). After the convolution, the above output results are passed through a maximum pooling layer with a pooling layer kernel size of 2 and a stride of 2, and then the output tensor is (batch_size, (time_steps - 2) / 2 + 1, 32). In addition, the second convolution kernel size is 3×3, the stride is 1, the number of input channels is 32, the number of output channels is 64, and the output tensor is (batch_size, (time_steps - 2) / 2 -1, 64). It is worth mentioning that after the second convolutional layer, a maximum pooling layer is also required, and the kernel size of the second pooling layer is 2, the step size is 2, and the output tensor is (batch_size,{(time_steps - 2) / 2 -3)} / 2+1, 64), where batch_size is the number of samples input to the network during each training, and batch_size is the number of samples set according to the specific feature time step.

[0078] Subsequently, the output result obtained after the second layer of pooling is used as the input of the LSTM model, so that the LSTM introduces memory units (i.e., cell states) and three gating mechanisms (i.e., input gates, forget gates, and output gates) to the output result obtained after the second layer of pooling to avoid the problem of gradient disappearance, thereby learning long-term dependencies. In a specific implementation, the input of the LSTM layer is (batch_size,{(time_steps-2) / 2 -3)} / 2+1, 64), and return_sequences=False is set, then the output shape is (batch_size, 64). Finally, the output of the LSTM model is used as the input of the fully connected layer, and the fully connected layer is subjected to regression analysis to finally obtain offline data.

[0079] Step S14: clustering the offline data and the online data using a preset clustering algorithm, and constructing a target expert diagnosis system based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

[0080] In this embodiment, after obtaining the offline data output by the deep learning model, the embodiment of the present application needs to cluster the offline data and the online data using a preset clustering algorithm to obtain the corresponding clustering results, and use the clustering results to establish a target expert diagnosis system, so as to use the target expert diagnosis system to perform intelligent diagnosis on the impact mode of the target wastewater treatment reactor. Specifically, using the target expert diagnosis system to diagnose the impact mode of the target wastewater treatment reactor may include: clustering the offline data and the online data based on the target expert diagnosis system and using a preset clustering algorithm to obtain clustering results; the clustering results include several clusters, each cluster corresponding to a different impact mode; obtaining several rule codes from a preset diagnosis knowledge base, so as to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each rule code and the clustering result, and returning the impact mode corresponding to the target wastewater treatment reactor to the current user interface; the impact mode includes a no-impact mode, a load impact mode, a toxic pollutant impact mode, and a temperature impact mode.

[0081] Furthermore, in the process of diagnosing the impact mode corresponding to the target wastewater treatment reactor using the target expert diagnosis system, the embodiment of the present application needs to obtain a number of rule codes from the preset diagnosis knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each rule code and the clustering result. Among them, the specific schematic diagram of the rule code corresponding to the impact mode is as follows Figure 3Specifically, before obtaining a plurality of rule codes from a preset diagnostic knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each rule code and the clustering result, it can also include: constructing diagnostic indicators corresponding to the load impact mode, the toxic pollutant impact mode and the temperature impact mode respectively based on preset experience, and constructing a preset diagnostic knowledge base based on each diagnostic indicator.

[0082] In the first specific embodiment, when the impact mode corresponding to the target wastewater treatment reactor is the load impact mode, the pH and carbon dioxide concentration inside the reactor are both lower than the corresponding diagnostic indicators, the internal conductivity, methane concentration and effluent alkalinity of the reactor are not greater than the corresponding diagnostic indicators, and the internal oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand of the reactor are respectively greater than the corresponding diagnostic indicators.

[0083] In a second specific embodiment, when the impact mode corresponding to the target wastewater treatment reactor is the toxic pollutant impact mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all less than the corresponding diagnostic indicators, and the numerical decrease rate corresponding to the internal conductivity of the reactor and the effluent alkalinity is greater than the preset decrease rate threshold; the internal oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand of the reactor are respectively greater than the corresponding diagnostic indicators; the numerical increase rate corresponding to the effluent volatile fatty acids is greater than the preset increase rate threshold;

[0084] In a third specific embodiment, when the shock mode corresponding to the target wastewater treatment reactor is the temperature shock mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all lower than the corresponding diagnostic indicators; the internal oxidation-reduction potential of the reactor, the effluent volatile fatty acids and the effluent chemical oxygen demand are respectively higher than the corresponding diagnostic indicators; the numerical increase in the effluent chemical oxygen demand is lower than the preset increase threshold; the hydrogen concentration and the effluent volatile fatty acids are not lower than the corresponding diagnostic indicators.

[0085] It is worth mentioning that after obtaining the target expert diagnosis system, the embodiment of the present application needs to use other impact data to verify the target expert diagnosis system, so as to adjust the parameters of the target expert diagnosis system using the verification results.

[0086] Furthermore, in a specific implementation, since the target expert diagnosis system is an expert diagnosis system based on the K-means clustering algorithm, when the target expert diagnosis system processes the input data, the input data will be divided into the following four clusters: Cluster 1: normal operating state, Cluster 2: load shock mode, Cluster 3: toxic pollutant shock mode and Cluster 4: temperature shock mode, and the schematic diagram is as shown in FIG. Figure 4 As shown. The K-means clustering algorithm can automatically divide the data into different clusters, each cluster represents a different impact mode, but these clusters cannot directly tell the specific meanings they represent, such as normal operation or impact mode. Therefore, the embodiment of the present application can analyze the clustering results based on domain knowledge and experience to give each cluster an actual meaning or label, that is, to build an expert diagnosis knowledge base.

[0087] As can be seen from the above, the embodiment of the present application first uses the target wastewater treatment reactor to treat several wastewaters input through several preset pipelines to obtain corresponding treated water samples; then, based on each treated water sample, the corresponding water sample component organic matter content and water sample alkalinity are determined, and the gas phase sensor and liquid phase sensor in the target wastewater treatment reactor are used to collect the corresponding online data when each wastewater is treated separately, and then a feature change fingerprint database is established based on each online data, water sample component organic matter content and water sample alkalinity; then, the initial deep learning model is trained using the information in the feature change fingerprint database to obtain the target deep learning model, so as to use the target deep learning model and output the corresponding offline data based on the online data; finally, the initial expert diagnosis system is trained using offline data and online data to obtain the target expert diagnosis system, so as to use the target expert diagnosis system to diagnose the impact mode of the target wastewater treatment reactor. In this way, the efficiency of diagnosing the impact mode of the wastewater treatment reactor is improved, thereby improving the stability and safety of the production process.

[0088] Accordingly, see Figure 5 As shown, the present application also provides an impact mode diagnostic device for a wastewater treatment reactor, comprising:

[0089] The treated water sample acquisition module 11 is used to treat a number of wastewaters input through a number of preset pipelines using a target wastewater treatment reactor to obtain corresponding treated water samples; the wastewater concentrations, wastewater components and wastewater acidity corresponding to each of the wastewaters are different;

[0090] The online data determination module 12 is used to determine the organic matter content and alkalinity of the corresponding water sample components based on each of the treated water samples, and use the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor to collect the corresponding online data when each of the wastewaters is treated, and then establish a characteristic change fingerprint database based on each of the online data, the organic matter content of the water sample components and the alkalinity of the water sample components;

[0091] An offline data determination module 13 is used to train an initial deep learning model using the information in the feature change fingerprint database to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data;

[0092] The impact mode determination module 14 is used to cluster the offline data and the online data using a preset clustering algorithm, and to construct a target expert diagnosis system based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

[0093] As can be seen from the above, before the embodiment of the present application performs the shock mode diagnosis of the wastewater treatment reactor, it is first necessary to use the target wastewater treatment reactor to treat several wastewaters input through several preset pipes to obtain the corresponding treated water samples; then, based on each treated water sample, the corresponding water sample component organic matter content and water sample alkalinity are determined, and the gas phase sensor and liquid phase sensor in the target wastewater treatment reactor are used to collect the corresponding online data when each wastewater is treated separately, and then a feature change fingerprint database is established based on each online data, water sample component organic matter content and water sample alkalinity; then, the initial deep learning model is trained using the information in the feature change fingerprint database to obtain the target deep learning model, so as to use the target deep learning model and output the corresponding offline data based on the online data; finally, the initial expert diagnosis system is trained using the offline data and the online data to obtain the target expert diagnosis system, so as to use the target expert diagnosis system to diagnose the shock mode of the target wastewater treatment reactor. In this way, the efficiency of diagnosing the shock mode of the wastewater treatment reactor is improved, thereby improving the stability and safety of the production process.

[0094] In some specific implementations, the processed water sample acquisition module 11 may specifically include:

[0095] A wastewater preparation unit, used for preparing a plurality of wastewaters, and then inputting each of the wastewaters into a target wastewater treatment reactor through a plurality of preset pipelines at a preset flow rate for treatment;

[0096] The water sampling unit is used to control a plurality of preset sampling pumps using a preset time relay, so as to use each of the preset sampling pumps to sample each treatment result output by the target wastewater treatment reactor according to a preset sampling frequency to obtain a corresponding treated water sample.

[0097] In some specific implementations, the online data determination module 12 may specifically include:

[0098] An organic matter content detection unit, used to detect the organic matter content of the water sample components corresponding to each of the treated water samples respectively using a preset organic matter content detection rule to obtain the organic matter content of the corresponding water sample components;

[0099] The water sample titration unit is used to titrate the water sample components corresponding to each of the treated water samples using a preset titration system to obtain corresponding titration results, and determine the water sample alkalinity corresponding to each of the water sample groups based on each of the titration results.

[0100] In some specific implementations, the online data determination module 12 may specifically include:

[0101] A gas information acquisition unit, used to detect the methane concentration, carbon dioxide concentration and hydrogen concentration in the gas path in the target wastewater treatment reactor using the gas phase sensor to obtain the corresponding methane concentration, carbon dioxide concentration and hydrogen concentration, and then detect the instantaneous gas flow in the gas path using a gas flow meter to obtain the instantaneous gas flow result; the gas phase sensor is located in the gas path, and the gas flow meter is located at the end of the gas path;

[0102] A gas production determination unit, configured to determine the methane production, the carbon dioxide production and the hydrogen production based on the methane concentration, the carbon dioxide concentration, the hydrogen concentration and the corresponding instantaneous gas flow results;

[0103] The reactor internal information determination unit is used to collect the pH, conductivity and redox potential of the target wastewater treatment reactor using the liquid phase sensor to obtain the reactor internal pH, reactor internal conductivity and reactor internal redox potential.

[0104] In some specific implementations, the offline data determination module 13 may specifically include:

[0105] A first data processing unit is used to process each of the online data in the feature change fingerprint database using a first preset memory subnetwork to obtain corresponding first data to be pooled, and to process each of the first data to be pooled respectively using a first preset pooling layer to obtain corresponding first data to be processed;

[0106] A second data processing unit is used to process each of the first to-be-processed data using a second preset memory subnetwork to obtain corresponding second to-be-pooled data, and to process each of the second to-be-pooled data using a second preset pooling layer to obtain corresponding second to-be-processed data;

[0107] A third data processing unit is used to process each of the second to-be-processed data using a preset memory unit, a preset input gate, a preset forget gate, and a preset output gate to obtain corresponding to-be-regressed data;

[0108] The offline data determination subunit is used to send each of the to-be-regressed data to a preset fully-connected layer, so as to use the preset fully-connected layer to perform regression analysis on each of the to-be-regressed data to obtain offline data; the initial deep learning model includes the first preset memory subnetwork, the second preset memory subnetwork, the preset memory unit, the preset input gate, the preset forget gate and the preset output gate; the offline data includes the chemical oxygen demand of the inlet and outlet water, the components of each pollutant and the concentration of each pollutant.

[0109] In some specific implementations, the impact mode determination module 14 may specifically include:

[0110] A data clustering unit, configured to cluster the offline data and the online data based on the target expert diagnosis system and using the preset clustering algorithm to obtain a clustering result; the clustering result includes a plurality of clusters, each of which corresponds to a different impact mode;

[0111] A rule code determination unit is used to obtain a number of rule codes from a preset diagnostic knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each of the rule codes and the clustering results, and return the impact mode corresponding to the target wastewater treatment reactor to the current user interface; the impact mode includes a no-impact mode, a load impact mode, a toxic pollutant impact mode and a temperature shock mode.

[0112] In some specific embodiments, the impact mode diagnostic device for the wastewater treatment reactor may further include:

[0113] A diagnostic knowledge base construction unit is used to construct diagnostic indicators corresponding to the load shock mode, the toxic pollutant shock mode and the temperature shock mode based on preset experience, and to construct a preset diagnostic knowledge base based on each of the diagnostic indicators; wherein, when the shock mode corresponding to the target wastewater treatment reactor is the load shock mode, the internal pH of the reactor and the carbon dioxide concentration are both less than the corresponding diagnostic indicators, the internal conductivity of the reactor, the methane concentration and the effluent alkalinity are not greater than the corresponding diagnostic indicators, and the internal oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand of the reactor are respectively greater than the corresponding diagnostic indicators; when the shock mode corresponding to the target wastewater treatment reactor is the toxic pollutant shock mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all less than the corresponding diagnostic indicators respectively, the numerical decrease speed of the reactor internal conductivity and the effluent alkalinity is greater than the preset decrease speed threshold; the reactor internal redox potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand are respectively greater than the corresponding diagnostic indicators; the numerical increase speed of the effluent volatile fatty acids is greater than the preset increase speed threshold; when the impact mode corresponding to the target wastewater treatment reactor is the temperature impact mode, the reactor internal pH, reactor internal conductivity, effluent water alkalinity, methane concentration, carbon dioxide concentration and methane production are all less than the corresponding diagnostic indicators respectively; the reactor internal redox potential, effluent volatile fatty acids and effluent chemical oxygen demand are respectively greater than the corresponding diagnostic indicators; the numerical increase amplitude of the effluent chemical oxygen demand is less than the preset amplitude increase threshold; the hydrogen concentration and effluent water volatile fatty acids are not less than the corresponding diagnostic indicators.

[0114] Furthermore, the present application also discloses an electronic device. Figure 6 It is a structural diagram of an electronic device 20 shown according to an exemplary embodiment, and the content in the figure cannot be regarded as any limitation on the scope of use of this application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input and output interface 25 and a communication bus 26. Among them, the memory 22 is used to store a computer program, and the computer program is loaded and executed by the processor 21 to implement the relevant steps in the impact mode diagnosis method of the wastewater treatment reactor disclosed in any of the aforementioned embodiments. In addition, the electronic device 20 in this embodiment can specifically be an electronic computer.

[0115] In this embodiment, the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and the external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input and output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs and is not specifically limited here.

[0116] In addition, the memory 22, as a carrier for storing resources, can be a read-only memory, a random access memory, a disk or an optical disk, etc. The resources stored thereon can include an operating system 221, a computer program 222, etc., and the storage method can be temporary storage or permanent storage.

[0117] The operating system 221 is used to manage and control the hardware devices and computer program 222 on the electronic device 20, and can be Windows Server, Netware, Unix, Linux, etc. In addition to including a computer program that can be used to complete the impact mode diagnosis method of the wastewater treatment reactor executed by the electronic device 20 disclosed in any of the aforementioned embodiments, the computer program 222 can further include a computer program that can be used to complete other specific tasks.

[0118] Furthermore, the present application also discloses a computer-readable storage medium for storing a computer program; wherein, when the computer program is executed by a processor, the aforementioned disclosed method for diagnosing the impact mode of a wastewater treatment reactor is implemented. The specific steps of the method can be referred to the corresponding contents disclosed in the aforementioned embodiments, and will not be repeated here.

[0119] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0120] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0121] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0122] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0123] The technical solution provided by the present application is introduced in detail above. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technicians in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for diagnosing impact modes of a wastewater treatment reactor, characterized in that: include: Using a target wastewater treatment reactor to treat a number of wastewaters input through a number of preset pipelines to obtain corresponding treated water samples; The wastewater concentration, wastewater components and wastewater acidity corresponding to each of the wastewaters are different; Determine the organic matter content and alkalinity of the corresponding water sample components based on each of the treated water samples, and use the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor to collect the corresponding online data when each of the wastewaters is treated, and then establish a characteristic change fingerprint database based on each of the online data, the organic matter content of the water sample components and the alkalinity of the water sample components; Using the information in the feature change fingerprint database to train an initial deep learning model to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data; The offline data and the online data are clustered using a preset clustering algorithm, and a target expert diagnosis system is constructed based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

2. The method for diagnosing the impact mode of a wastewater treatment reactor according to claim 1, characterized in that: The method of using the target wastewater treatment reactor to treat a plurality of wastewaters input through a plurality of preset pipelines to obtain corresponding treated water samples includes: Prepare a plurality of wastewaters, and then input each of the wastewaters into a target wastewater treatment reactor through a plurality of preset pipelines at a preset flow rate for treatment; A preset time relay is used to control a plurality of preset sampling pumps, so that each of the preset sampling pumps is used to sample each treatment result output by the target wastewater treatment reactor according to a preset sampling frequency to obtain a corresponding treated water sample.

3. The method for diagnosing the impact mode of a wastewater treatment reactor according to claim 1, characterized in that: The determining of the organic matter content and alkalinity of the corresponding water sample components based on each of the treated water samples comprises: Using the preset organic matter content detection rules, the organic matter content of the water sample components corresponding to each of the treated water samples is detected respectively to obtain the organic matter content of the corresponding water sample components; The water sample components corresponding to each of the treated water samples are titrated using a preset titration system to obtain corresponding titration results, and the alkalinity of the water samples corresponding to each of the water sample groups is determined based on each of the titration results.

4. The method for diagnosing the impact mode of a wastewater treatment reactor according to claim 1, characterized in that: The method of collecting the online data corresponding to each wastewater being treated respectively by using the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor includes: The gas phase sensor is used to detect the methane concentration, carbon dioxide concentration and hydrogen concentration in the gas path in the target wastewater treatment reactor to obtain the corresponding methane concentration, carbon dioxide concentration and hydrogen concentration, and then the gas flow meter is used to detect the instantaneous gas flow in the gas path to obtain the instantaneous gas flow result; the gas phase sensor is located in the gas path, and the gas flow meter is located at the end of the gas path; Determine the methane production, the carbon dioxide production and the hydrogen production based on the methane concentration, the carbon dioxide concentration, the hydrogen concentration and the corresponding instantaneous gas flow results respectively; The liquid phase sensor is used to collect the pH, conductivity and redox potential of the target wastewater treatment reactor to obtain the pH inside the reactor, the conductivity inside the reactor and the redox potential inside the reactor.

5. The method for diagnosing the impact mode of a wastewater treatment reactor according to claim 1, characterized in that: The method of training the initial deep learning model using the information in the feature change fingerprint database to obtain a target deep learning model, and outputting corresponding offline data based on the online data using the target deep learning model, includes: Using a first preset memory subnetwork to process each of the online data in the feature change fingerprint database to obtain corresponding first data to be pooled, and using a first preset pooling layer to process each of the first data to be pooled respectively to obtain corresponding first data to be processed; Using a second preset memory subnetwork to process each of the first to-be-processed data to obtain corresponding second to-be-pooled data, and using a second preset pooling layer to process each of the second to-be-pooled data to obtain corresponding second to-be-processed data; Using a preset memory unit, a preset input gate, a preset forget gate, and a preset output gate to process each of the second data to be processed, respectively, to obtain corresponding data to be regressed; Each of the data to be regressed is sent to a preset fully connected layer, so as to use the preset fully connected layer to perform regression analysis on each of the data to be regressed to obtain offline data; the initial deep learning model includes the first preset memory subnetwork, the second preset memory subnetwork, the preset memory unit, the preset input gate, the preset forget gate and the preset output gate; the offline data includes the chemical oxygen demand of the inlet and outlet water, each pollutant component and the corresponding pollutant concentration of each of the pollutant components, and the organic matter content of the water sample components.

6. The method for diagnosing the impact mode of a wastewater treatment reactor according to any one of claims 1 to 5, characterized in that: The method of diagnosing the impact mode of the target wastewater treatment reactor by using the target expert diagnosis system includes: Based on the target expert diagnosis system and using the preset clustering algorithm, the offline data and the online data are clustered to obtain a clustering result; the clustering result includes a plurality of clusters, each of which corresponds to a different impact mode; A number of rule codes are obtained from a preset diagnostic knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each of the rule codes and the clustering results, and the impact mode corresponding to the target wastewater treatment reactor is returned to the current user interface; the impact mode includes a no-impact mode, a load impact mode, a toxic pollutant impact mode and a temperature impact mode.

7. The method for diagnosing the impact mode of a wastewater treatment reactor according to claim 6, characterized in that: Before acquiring a plurality of rule codes from a preset diagnosis knowledge base to diagnose the impact mode corresponding to the target wastewater treatment reactor based on each of the rule codes and the clustering result, the method further includes: Constructing diagnostic indicators corresponding to the load shock mode, the toxic pollutant shock mode and the temperature shock mode respectively based on preset experience, and constructing a preset diagnostic knowledge base based on each of the diagnostic indicators; Wherein, when the impact mode corresponding to the target wastewater treatment reactor is the load impact mode, the pH value and carbon dioxide concentration inside the reactor are both less than the corresponding diagnostic indicators, the conductivity, methane concentration and effluent alkalinity inside the reactor are not greater than the corresponding diagnostic indicators, and the oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand inside the reactor are greater than the corresponding diagnostic indicators; When the impact mode corresponding to the target wastewater treatment reactor is the toxic pollutant impact mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all less than the corresponding diagnostic indicators respectively, and the numerical decrease rate corresponding to the internal conductivity of the reactor and the effluent alkalinity is greater than the preset decrease rate threshold; the internal oxidation-reduction potential, hydrogen concentration, effluent volatile fatty acids, effluent water volatile fatty acids and effluent chemical oxygen demand of the reactor are respectively greater than the corresponding diagnostic indicators; the numerical increase rate corresponding to the effluent volatile fatty acids is greater than the preset increase rate threshold; When the shock mode corresponding to the target wastewater treatment reactor is the temperature shock mode, the internal pH of the reactor, the internal conductivity of the reactor, the effluent alkalinity, the methane concentration, the carbon dioxide concentration and the methane production are all lower than the corresponding diagnostic indicators respectively; the internal oxidation-reduction potential of the reactor, the effluent volatile fatty acids and the effluent chemical oxygen demand are respectively higher than the corresponding diagnostic indicators; the increase in the numerical value of the effluent chemical oxygen demand is lower than the preset increase threshold; the hydrogen concentration and the effluent volatile fatty acids are not lower than the corresponding diagnostic indicators.

8. An impact mode diagnostic device for a wastewater treatment reactor, characterized in that: include: A treated water sample acquisition module is used to use a target wastewater treatment reactor to treat a number of wastewaters input through a number of preset pipelines to obtain corresponding treated water samples; The wastewater concentration, wastewater components and wastewater acidity corresponding to each of the wastewaters are different; An online data determination module is used to determine the organic content of the corresponding water sample components based on each of the treated water samples, and to collect the corresponding online data when each of the wastewaters is treated respectively using the gas phase sensor and the liquid phase sensor in the target wastewater treatment reactor, and then to establish a characteristic change fingerprint database based on each of the online data, the organic content of the water sample components and the alkalinity of the water sample components; An offline data determination module, used to train an initial deep learning model using the information in the feature change fingerprint database to obtain a target deep learning model, so as to use the target deep learning model and output corresponding offline data based on the online data; The impact mode determination module is used to cluster the offline data and the online data using a preset clustering algorithm, and to construct a target expert diagnosis system based on the obtained clustering results, so as to diagnose the impact mode of the target wastewater treatment reactor using the target expert diagnosis system.

9. An electronic device, characterized in that: include: Memory, used to store computer programs; A processor is used to execute the computer program to implement the steps of the impact mode diagnosis method for a wastewater treatment reactor as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein when the computer program is executed by a processor, the steps of the impact mode diagnosis method for a wastewater treatment reactor as described in any one of claims 1 to 7 are implemented.

Images