A decentralized rural sewage treatment site intelligent optimization method and system

Abstract

The present application relates to the technical field of sewage treatment, and in particular to a method and system for intelligent optimization of decentralized rural sewage treatment sites. The present application first collects and cleans water quality parameters and operating parameter data, inputs the pre-processed characteristic data into the prediction model to obtain the prediction results, compares the prediction results with the real-time operating data to obtain the deviation data, and dynamically optimizes and distributes the model parameters accordingly, reducing the blindness of parameter setting in traditional methods, and being able to adjust the control strategy in real time according to the actual operating conditions, significantly reducing the operating costs while ensuring that the effluent meets the standards, and improving the processing efficiency.

Description

Technical Field

[0001] The present application relates to the technical field of sewage treatment, and in particular to a method and system for intelligent optimization of decentralized rural sewage treatment sites. Background Art

[0002] Rural domestic sewage is one of the main sources of rural environmental pollution. With the increase in rural environmental governance efforts, decentralized sewage treatment facilities have been widely used. However, due to the large fluctuations in water quality and quantity of rural domestic sewage and the scattered distribution of sites, how to achieve intelligent operation and management has become an urgent problem to be solved.

[0003] For the operation optimization of decentralized rural sewage treatment stations, the control method based on fixed operating parameters is currently mainly used. By setting the target value range of key operating parameters such as dissolved oxygen and reflow ratio, traditional control strategies such as PID control are used to adjust the operation.

[0004] The existing technology cannot adapt to the dynamic changes in the quality and quantity of influent water, resulting in unstable treatment effects. In addition, the operating parameter settings lack specificity, causing waste of energy and chemicals. This situation needs to be further improved. Summary of the invention

[0005] In order to solve the problem that existing rural sewage treatment sites lack pertinence and cause waste of energy and medicine, the present application provides a decentralized rural sewage treatment site intelligent optimization method and system, which adopts the following technical solutions:

[0006] In a first aspect, the present application provides a decentralized rural sewage treatment site intelligent optimization method, comprising the following steps:

[0007] Collecting real-time operation data and historical data including water quality parameters and operation parameters, and performing data cleaning on the real-time operation data and the historical data to obtain pre-processed characteristic data;

[0008] Inputting the preprocessed feature data into a preset prediction model to obtain a prediction result;

[0009] Comparing the prediction result with the real-time operation data to obtain deviation data;

[0010] The corresponding site parameters in the prediction model are optimized according to the deviation data, the optimized parameters are returned to the prediction model, and sent to the corresponding sewage treatment site.

[0011] By adopting the above technical scheme, existing rural sewage treatment sites often adopt fixed parameters to operate when facing fluctuations in influent water quality, resulting in low treatment efficiency and high energy and drug consumption. The present application first collects and cleans data such as water quality parameters and operating parameters, inputs the pre-processed characteristic data into the prediction model to obtain the prediction results, obtains deviation data by comparing with the real-time operating data, and dynamically optimizes and distributes the model parameters accordingly, reducing the blindness of parameter setting in traditional methods, and being able to adjust the control strategy in real time according to the actual operating conditions, significantly reducing operating costs while ensuring that the effluent meets the standards, and improving treatment efficiency.

[0012] Optionally, data cleaning is performed on the real-time operation data and the historical data to obtain pre-processed feature data, which specifically includes the following steps:

[0013] The real-time operation data is segmented according to the water inlet flow rate to obtain high load segment, medium load segment and low load segment data;

[0014] For the high load segment, medium load segment and low load segment data, respectively, an abnormal value detection threshold is set, and abnormal data is eliminated according to the corresponding abnormal value detection threshold;

[0015] The missing values ​​in the real-time operation data are supplemented according to the data of the same load segment in the historical data to obtain a complete data set;

[0016] The complete data set is standardized to obtain preprocessed feature data.

[0017] By adopting the above technical scheme, since the real-time operation data of rural sewage treatment sites often have quality problems such as outliers and missing values, and the fluctuation characteristics of the data under different load conditions are significantly different; this application first segments the data based on the inlet flow rate, and then sets the corresponding outlier detection threshold for each load segment to eliminate outliers, and then uses the historical data of the same load segment to supplement the missing values, and finally performs standardized processing to obtain characteristic data. Through the segmented adaptive processing strategy, the accuracy of data cleaning and outlier recognition rate is improved.

[0018] Optionally, the preprocessed feature data is input into a preset prediction model to obtain a prediction result, which specifically includes the following steps:

[0019] determining an operating state of the sewage treatment process according to the operating parameters;

[0020] Inputting the pre-processed characteristic data into a preset water quality prediction model, and calculating the pollutant degradation rate in combination with the operating status;

[0021] Predicting effluent water quality indicators based on the degradation rate of the pollutants;

[0022] Calculate the energy consumption data and drug consumption data required to achieve the effluent water quality indicators.

[0023] By adopting the above technical solution, the present application first determines the process operation status according to the operation parameters, then inputs the characteristic data into the water quality prediction model and calculates the pollutant degradation rate in combination with the operation status, and then predicts the effluent indicators, and finally calculates the energy consumption and drug consumption data required to meet the standards; by introducing operation status judgment and degradation kinetics analysis, the coordinated prediction of effluent water quality and resource consumption is achieved, the prediction accuracy is improved, and the operation cost is reduced.

[0024] Optionally, the operating state is determined according to the influent water quality parameters and water temperature changes, and the operating state includes a stable operating state, an impact load state, and a water temperature mutation state; determining the operating state of the sewage treatment process according to the operating parameters specifically includes the following steps:

[0025] Calculate the change rate of the inlet water quality parameter within the continuous monitoring time period, and determine that it is an impact load state when the change rate exceeds a preset fluctuation threshold and the duration exceeds a first set time period;

[0026] Monitor the water temperature change curve, and when the water temperature change rate exceeds the preset temperature threshold and the duration exceeds the second set time, it is determined to be a water temperature mutation state;

[0027] When the change rate of the inlet water quality parameter and the water temperature change rate do not exceed the corresponding preset thresholds, and the operating parameters remain within the set range, it is determined to be in a stable operating state.

[0028] By adopting the above technical scheme, since rural sewage treatment sites are often faced with interference factors such as sudden changes in influent water quality and seasonal temperature fluctuations, traditional operating status judgment methods often rely only on instantaneous data for judgment, and cannot accurately identify abnormal operating conditions such as continuous impact loads and temperature mutations; the present application first calculates the rate of change of the influent water quality parameters and combines the duration to determine whether it is in an impact load state, then monitors the water temperature change rate and its duration to determine whether a temperature mutation occurs, and finally determines whether it is in a stable operating state through comprehensive judgment; by introducing the change rate and duration, the accuracy and timeliness of the operating status judgment are improved.

[0029] Optionally, optimizing corresponding site parameters in the prediction model according to the deviation data and returning the optimized parameters to the prediction model specifically includes the following steps:

[0030] When the deviation data exceeds a preset deviation threshold, model parameter optimization is triggered;

[0031] Calculate the loss function value according to the deviation data;

[0032] Iteratively optimizing the corresponding site parameters in the prediction model until the loss function value is less than a preset function threshold;

[0033] The optimized parameters are returned to the prediction model.

[0034] By adopting the above technical solution, since the operating conditions of rural sewage treatment sites often change, the parameters of traditional prediction models often remain fixed and cannot adapt to the new operating environment in time, resulting in a gradual decrease in prediction accuracy; this application first triggers the optimization process by the deviation data exceeding the preset threshold, and then calculates the loss function value based on the deviation data, and then iteratively optimizes the parameters until the loss function value meets the requirements, and finally updates the optimized parameters to the prediction model, thereby improving the adaptability of the model and prediction accuracy.

[0035] Optionally, the water quality parameters include COD, ammonia nitrogen, total nitrogen and total phosphorus; the operating parameters include dissolved oxygen, water temperature, pH value, sludge concentration and reflow ratio.

[0036] Optionally, the energy consumption data includes an air supply control target value, which is calculated based on the influent COD, ammonia nitrogen and aerator oxygen utilization rate in the pretreated characteristic data;

[0037] The method further includes optimizing aeration control according to the air supply control target value, specifically including the following steps:

[0038] The actual air supply volume is detected by an air flow meter, and the fan operating frequency is automatically adjusted according to the air supply volume control target value;

[0039] The opening of the air valves in each section of the aerobic pool is adjusted according to the operating frequency of the fan and the real-time monitoring value of the dissolved oxygen concentration in each section of the aerobic pool.

[0040] By adopting the above technical scheme, the energy consumption of the aeration system of rural sewage treatment sites accounts for a large part of the total energy consumption. The traditional fixed-frequency operation or single dissolved oxygen control method cannot adapt to the dynamic changes of the influent load and the pollutant concentration, resulting in low aeration efficiency and serious energy waste. The present application first automatically adjusts the fan operating frequency based on the actual air supply detection value and the target value, and then dynamically adjusts the opening of the air valve of each section of the aerobic tank according to the fan frequency and the real-time monitoring value of the dissolved oxygen in each section. By establishing an air supply prediction model and a two-stage linkage control strategy, the refined regulation of the aeration system is achieved.

[0041] Optionally, the drug consumption data includes a phosphorus removal dosage control target value, and the phosphorus removal dosage control target value is calculated based on the influent total phosphorus and chemical phosphorus removal efficiency in the pretreated characteristic data;

[0042] The method further includes optimizing the dosing control according to the phosphorus removal dosing control target value, specifically including the following steps:

[0043] The actual dosage is detected by a dosage flow meter, and the operation frequency of the dosage pump is automatically adjusted according to the phosphorus removal dosage control target value;

[0044] The allocation ratio of the dosing positions is adjusted according to the operating frequency of the dosing pump and the total phosphorus concentration of the reaction tank effluent.

[0045] By adopting the above technical scheme, when the sewage treatment station adopts a fixed dosage and single-point addition method, excessive dosage will cause waste when the total phosphorus concentration of the inlet water is low, and insufficient dosage will cause the effluent to exceed the standard when the concentration is high; this application first automatically adjusts the operating frequency of the dosing pump according to the actual dosage detection value and the target value, and then dynamically adjusts the allocation ratio of different dosing positions according to the dosing pump frequency and the effluent total phosphorus monitoring value, and realizes precise regulation of the phosphorus removal system by establishing a dosage prediction model and a graded dosing control strategy.

[0046] In a second aspect, the present application provides a decentralized rural sewage treatment site intelligent optimization system, comprising:

[0047] Data acquisition module, used to collect real-time operation data and historical data including water quality parameters and operation parameters;

[0048] A data preprocessing module, used for performing data cleaning on the real-time operation data and the historical data to obtain preprocessed feature data;

[0049] A prediction module, used for inputting the preprocessed feature data into a preset prediction model to obtain a prediction result;

[0050] A comparison module, used to compare the prediction result with the real-time operation data to obtain deviation data;

[0051] The optimization module is used to optimize the corresponding site parameters in the prediction model according to the deviation data, return the optimized parameters to the prediction model, and send them to the corresponding sewage treatment site.

[0052] Optionally, the data preprocessing module includes:

[0053] A segmentation unit, used to segment the real-time operation data according to the size of the water inlet flow rate to obtain high load segment, medium load segment and low load segment data;

[0054] An abnormality detection unit, used to set abnormal value detection thresholds for the high load segment, medium load segment and low load segment data respectively, and remove abnormal data according to the corresponding abnormal value detection thresholds;

[0055] A data supplement unit, used to supplement the missing values ​​in the real-time operation data according to the data of the same load segment in the historical data to obtain a complete data set;

[0056] The standardization processing unit is used to perform standardization processing on the complete data set to obtain pre-processed feature data.

[0057] In summary, the present application includes at least one of the following beneficial technical effects:

[0058] This application first collects and cleans water quality parameters, operating parameters and other data, inputs the pre-processed characteristic data into the prediction model to obtain the prediction results, obtains the deviation data by comparing with the real-time operating data, and dynamically optimizes and issues the model parameters accordingly, reducing the blindness of parameter setting in traditional methods, and can adjust the control strategy in real time according to the actual operating conditions, significantly reducing the operating cost while ensuring that the effluent meets the standards, and improving the processing efficiency;

[0059] Since the real-time operation data of rural sewage treatment stations often have quality problems such as outliers and missing values, and the fluctuation characteristics of the data under different load conditions are significantly different; this application first segments the data based on the inlet flow rate, then sets the corresponding outlier detection threshold for each load segment to remove outliers, then uses the historical data of the same load segment to supplement the missing values, and finally performs standardization processing to obtain characteristic data. Through the segmented adaptive processing strategy, the accuracy of data cleaning and outlier recognition rate is improved;

[0060] This application first determines the process operation status based on the operating parameters, then inputs the characteristic data into the water quality prediction model and calculates the pollutant degradation rate in combination with the operation status, and then predicts the effluent indicators, and finally calculates the energy consumption and drug consumption data required to meet the standards; by introducing operation status judgment and degradation kinetics analysis, the coordinated prediction of effluent water quality and resource consumption is achieved, the prediction accuracy is improved, and the operating costs are reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is one of the flow diagrams of a decentralized rural sewage treatment site intelligent optimization method according to an embodiment of the present application;

[0062] Figure 2 It is a flow chart of step S110 in a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0063] Figure 3 It is a flow chart of step S120 in a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0064] Figure 4It is a flow chart of step S121 in a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0065] Figure 5 It is a flow chart of step S140 in a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0066] Figure 6 This is the second flow chart of a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0067] Figure 7 This is the third flow chart of a method for intelligent optimization of a decentralized rural sewage treatment site according to an embodiment of the present application;

[0068] Figure 8 This is a module schematic diagram of a decentralized rural sewage treatment site intelligent optimization system according to an embodiment of the present application;

[0069] Fig. 9 It is a diagram of the internal structure of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0070] The terms used in the following embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to be used as limitations to the present application. As used in the specification and appended claims of the present application, the singular expressions "one", "a kind of", "said", "above", "the" and "this" are intended to also include plural expressions, unless there is a clear indication to the contrary in the context. It should also be understood that the term "and / or" used in the present application refers to any or all possible combinations comprising one or more listed items.

[0071] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features, and in the description of the embodiments of the present application, unless otherwise specified, "plurality" means two or more.

[0072] The embodiments of the present application are further described in detail below in conjunction with the drawings in the specification.

[0073] In the first aspect, the present application provides a decentralized rural sewage treatment site intelligent optimization method, referring to Figure 1 , including the following steps:

[0074] S110, collecting real-time operation data and historical data including water quality parameters and operation parameters, performing data cleaning on the real-time operation data and historical data, and obtaining pre-processed characteristic data.

[0075] In this embodiment, real-time operation data refers to data collected in real time by various sensors at the sewage treatment site, including water quality parameters such as COD, ammonia nitrogen, etc., and operation parameters such as dissolved oxygen, pH value, etc. Historical data refers to past operation data stored for a long time.

[0076] Specifically, the data cleaning process includes steps such as outlier removal, missing value supplementation and data standardization. This application adopts a segmented outlier detection method based on load segments to improve the accuracy of outlier detection. Missing values ​​are supplemented based on data from similar load segments in the past.

[0077] S120: Input the preprocessed feature data into a preset prediction model to obtain a prediction result.

[0078] In this embodiment, the prediction model is a neural network model constructed based on a machine learning algorithm, which is used to predict effluent water quality indicators, energy consumption and drug consumption based on input data.

[0079] Specifically, the characteristic data is input into the prediction model in batches, and the model processes multiple batches of data in parallel to improve the prediction efficiency. The prediction results include index values ​​such as effluent COD and ammonia nitrogen, as well as control target values ​​such as gas supply and dosage.

[0080] S130: Compare the prediction result with the real-time operation data to obtain deviation data.

[0081] In this embodiment, the deviation data reflects the degree of difference between the predicted results and the actual operation data.

[0082] Specifically, the predicted water output index is compared with the measured value to obtain the water quality index deviation; the predicted gas supply is compared with the actual gas supply to obtain the gas supply deviation; the predicted dosage is compared with the actual dosage to obtain the dosage deviation.

[0083] S140. Optimize the corresponding site parameters in the prediction model according to the deviation data, return the optimized parameters to the prediction model, and send them to the corresponding sewage treatment site.

[0084] In this embodiment, site parameters refer to some characteristic parameters of each sewage treatment site, including reactor volume, designed sludge concentration and aeration intensity.

[0085] Specifically, by constructing a loss function, taking the deviation data as input, and iteratively optimizing the site parameters based on optimization algorithms such as gradient descent, the loss function value is minimized, thereby obtaining the optimized parameters. The optimized parameters will be fed back to the prediction model and sent to the on-site control system to guide the operation of sewage treatment. For example, according to the optimized reactor volume parameters, the residence time of sewage in each reactor is controlled; according to the optimized design sludge concentration parameters, the sludge concentration of the aeration tank is controlled; according to the optimized aeration intensity parameters, the ventilation volume of the aerator is adjusted.

[0086] In one embodiment, referring to Figure 2 In step S110, data cleaning is performed on the real-time operation data and the historical data to obtain pre-processed feature data, which specifically includes the following steps:

[0087] S111. Segment the real-time operation data according to the size of the water inlet flow rate to obtain high load segment, medium load segment and low load segment data.

[0088] In this embodiment, the water inlet flow rate is an important indicator for determining the load status of the sewage treatment system.

[0089] Specifically, according to the measured water inlet flow value, it is divided into three intervals, for example: the flow below a certain flow threshold Q1 is the low load segment, the flow above a certain flow threshold Q2 is the high load segment, and the flow between Q1 and Q2 is the medium load segment. The corresponding data are classified into three different load segment data sets: low, medium, and high.

[0090] S112, setting abnormal value detection thresholds for the high load segment, the medium load segment, and the low load segment data respectively, and eliminating abnormal data according to the corresponding abnormal value detection thresholds.

[0091] Specifically, the mean and standard deviation of the high, medium and low load segment data sets are calculated respectively, and then the corresponding upper and lower thresholds of outlier detection are determined according to the 3σ principle. For each load segment, data that exceeds its corresponding threshold range is judged as an outlier and removed.

[0092] S113. The missing values ​​in the real-time operation data are supplemented according to the data of the same load segment in the historical data to obtain a complete data set.

[0093] In this embodiment, the real-time operation data is traversed to identify records with missing values, and then the complete record closest to the record is found in the same load segment data of the historical data, and the values ​​of the corresponding parameters are used to fill the missing values.

[0094] S114, standardize the complete data set to obtain preprocessed feature data.

[0095] Specifically, the above-mentioned standardized calculation is performed on each parameter value in the data set to obtain normal distribution characteristic data with 0 mean and 1 standard deviation.

[0096] In one embodiment, referring to Figure 3 In step S120, the preprocessed feature data is input into a preset prediction model to obtain a prediction result, which specifically includes the following steps:

[0097] S121. Determine the operating status of the sewage treatment process according to the operating parameters.

[0098] In this embodiment, the operating state of the sewage treatment process directly affects the degradation efficiency of pollutants.

[0099] Specifically, according to the operating parameters such as inlet water quality, water temperature, etc., the operating state is divided into three situations: stable operation, impact load and water temperature mutation. For example, if the inlet water quality fluctuates violently for a long time, it is determined to be an impact load state.

[0100] S122. Input the pre-processed characteristic data into a preset water quality prediction model, and calculate the pollutant degradation rate in combination with the operating status.

[0101] In this embodiment, the model will learn the mapping relationship between characteristic data and pollutant degradation rate, and calculate the corresponding degradation rate values ​​of pollutants such as COD and ammonia nitrogen under given characteristic data and operating conditions.

[0102] S123. Predict effluent water quality indicators based on pollutant degradation rates.

[0103] In this embodiment, the degradation of pollutants during the treatment process directly determines the water quality index of the effluent.

[0104] Specifically, a calculation model for effluent water quality indicators is constructed, and the degradation rate of each pollutant and other factors are taken as input to calculate the predicted values ​​of a series of effluent indicators such as COD and ammonia nitrogen.

[0105] S124. Calculate the energy consumption data and drug consumption data required to achieve the effluent water quality index.

[0106] Specifically, based on the predicted effluent water quality indicators, combined with process parameters, equipment efficiency and other factors, the energy consumption and drug consumption data such as gas supply and drug addition required to achieve these indicators are calculated.

[0107] In one embodiment, referring to Figure 4 The operating state is determined according to the influent water quality parameters and water temperature changes, and the operating state includes a stable operating state, an impact load state, and a water temperature mutation state; in step S121, the operating state of the sewage treatment process is determined according to the operating parameters, which specifically includes the following steps:

[0108] S1211. Calculate the change rate of the inlet water quality parameter during the continuous monitoring period. When the change rate exceeds a preset fluctuation threshold and the duration exceeds a first set time, determine it as an impact load state.

[0109] In this embodiment, the inlet water quality parameters include conventional indicators such as COD, ammonia nitrogen, and total phosphorus. The system collects online monitoring data every 5 minutes and determines whether a shock load occurs by calculating the change rate within 30 minutes. When the COD change rate exceeds 20% / h or the ammonia nitrogen change rate exceeds 25% / h, and the state lasts for more than 1 hour, it is determined to enter the shock load state, and the system will automatically adjust the corresponding operating parameters to cope with the water quality shock.

[0110] Specifically, during the daily operation of a sewage treatment plant, online monitoring revealed that the influent COD rose rapidly from 200mg / L to 350mg / L within 2 hours, with a change rate of 37.5% / h, and this trend lasted for 1.5 hours. Based on this, the system determined that it had entered an impact load state and adjusted the aeration volume and reflow ratio in a timely manner, so that the effluent quality remained stable and met the discharge standards.

[0111] S1212, monitoring the water temperature change curve, when the water temperature change rate exceeds the preset temperature threshold and the duration exceeds the second set time, it is determined to be a water temperature sudden change state.

[0112] In this embodiment, water temperature is a key parameter affecting the biochemical system, and its changing trend needs to be monitored in real time. The system records water temperature data every 10 minutes and calculates the rate of change within 2 hours. When the water temperature change rate exceeds 2°C / h and lasts for more than 2 hours, it is determined to be a sudden change in water temperature. At this time, the operating parameters of the biochemical system need to be adjusted in time to maintain the treatment effect.

[0113] Specifically, during the winter operation of a sewage treatment plant, the inlet water temperature dropped from 18°C ​​to 12°C in one day, with a change rate of 2.5°C / h, and the cooling trend lasted for 3 hours. Based on this, the system determined that the water temperature had entered a sudden change state, and immediately launched an emergency control plan, ensuring the stability of the treatment effect by adjusting parameters such as sludge age and dissolved oxygen concentration.

[0114] S1213. When the change rate of the inlet water quality parameter and the water temperature change rate do not exceed the corresponding preset thresholds, and the operating parameters remain within the set range, it is determined to be a stable operating state.

[0115] In this embodiment, the determination of the stable operation state needs to meet the stability requirements of water quality and water temperature at the same time. The system monitors the change rate of the inlet water quality parameters and the change rate of the water temperature in real time. When these indicators are kept within the preset threshold range, and the process operation parameters such as dissolved oxygen, pH value, MLSS, etc. are within the normal range, it is determined to be a stable operation state, and the conventional operation parameter settings are maintained to ensure the continuous stability of the treatment effect.

[0116] In one embodiment, referring to Figure 5 In step S140, the corresponding site parameters in the prediction model are optimized according to the deviation data, and the optimized parameters are returned to the prediction model, which specifically includes the following steps:

[0117] S141. When the deviation data exceeds a preset deviation threshold, model parameter optimization is triggered.

[0118] In this embodiment, a preset deviation threshold is set to determine whether the model prediction effect meets the requirements. When the average deviation between the predicted value and the actual operating data within 24 consecutive hours exceeds 15%, or the single prediction deviation exceeds 25%, the system automatically triggers the parameter optimization process.

[0119] S142. Calculate the loss function value according to the deviation data.

[0120] In this embodiment, the loss function uses the root mean square error as an evaluation indicator, and the loss function value reflecting the prediction accuracy of the model is obtained by calculating the square root of the sum of the squares of the deviations between the predicted value and the actual value.

[0121] S143. Iteratively optimize the corresponding site parameters in the prediction model until the loss function value is less than a preset function threshold.

[0122] In this embodiment, the gradient descent algorithm is used to iteratively optimize the model parameters, and the loss function value is recalculated after each iteration until the loss function value is less than the preset 0.15 threshold or the maximum number of iterations is 100.

[0123] S144. Return the optimized parameters to the prediction model.

[0124] In one embodiment, the energy consumption data includes an air supply control target value, and the method further includes optimizing aeration control according to the air supply control target value, wherein the air supply control target value is calculated based on the influent COD, ammonia nitrogen and aerator oxygen utilization rate in the pre-treated characteristic data, referring to Figure 6 , specifically including the following steps:

[0125] S610, detecting the actual air supply volume through an air flow meter, and automatically adjusting the fan operating frequency according to the air supply volume control target value.

[0126] In this embodiment, the system collects actual air supply data once a minute through an air flow meter installed on the main air supply pipeline, and compares it with the air supply control target value calculated based on the water inlet characteristics. When the deviation between the actual air supply and the target value exceeds ±5%, the control system automatically adjusts the fan operating frequency through the frequency converter, and the adjustment step is set to 2Hz. After each adjustment, wait for 5 minutes to observe the effect to avoid frequent adjustments causing system instability.

[0127] S620, adjusting the opening of the air valves of each section of the aerobic pool according to the fan operating frequency and the real-time monitoring value of the dissolved oxygen concentration of each section of the aerobic pool.

[0128] In this embodiment, the aerobic pool is divided into three sections according to the requirements of the sewage treatment process, and each section is equipped with an independent dissolved oxygen online monitor and regulating air valve. The system automatically controls the opening of each air valve in a proportional adjustment manner according to the changes in the fan operating frequency and the real-time dissolved oxygen monitoring value of each section. Among them, the dissolved oxygen control target of the first section is 2.0mg / L, the second section is 1.5mg / L, and the third section is 1mg / L, and the adjustment interval is 5 minutes.

[0129] In one embodiment, the drug consumption data includes a phosphorus removal dosage control target value, and the method further includes optimizing the dosing control according to the phosphorus removal dosage control target value, wherein the phosphorus removal dosage control target value is calculated based on the influent total phosphorus and the chemical phosphorus removal efficiency in the pretreated characteristic data, and the reference Figure 7 , specifically including the following steps:

[0130] S710, detecting the actual dosage through the dosage flow meter, and automatically adjusting the operation frequency of the dosage pump according to the phosphorus removal dosage control target value.

[0131] In this embodiment, the system uses an electromagnetic flowmeter to monitor the actual addition flow of the phosphorus removal agent in real time, and the monitoring frequency is once every 5 minutes. According to the online monitoring value of the total phosphorus in the inlet water and the target total phosphorus requirement of the outlet water, the required dosage of the agent is calculated in combination with the chemical phosphorus removal efficiency. When the actual dosage deviates from the target value by more than ±5%, the frequency conversion controller automatically adjusts the operating frequency of the dosing pump, and the adjustment step is 1.5Hz. After each adjustment, observe for 15 minutes to ensure that the dosing system operates smoothly.

[0132] S720. Adjust the allocation ratio of the dosing positions according to the operating frequency of the dosing pump and the total phosphorus concentration of the reaction tank effluent.

[0133] In this embodiment, three dosing points of the phosphorus removal agent are set in the reaction tank, which are located at the entrance of the anaerobic tank, the entrance of the anoxic tank, and the entrance of the aerobic tank. The system dynamically adjusts the agent distribution ratio of the three dosing points according to the change of the operation frequency of the dosing pump and the online monitoring data of the total phosphorus in the effluent of the reaction tank. The initial distribution ratio is set to 4:3:3. When the total phosphorus in the effluent cannot stably meet the standard, the system will automatically optimize the distribution ratio, and the adjustment interval is 20 minutes.

[0134] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0135] On the second aspect, the present application provides a decentralized rural sewage treatment site intelligent optimization system. The decentralized rural sewage treatment site intelligent optimization system of the present application is described below in combination with the above-mentioned decentralized rural sewage treatment site intelligent optimization method.

[0136] Reference Figure 8 , a decentralized rural sewage treatment site intelligent optimization system, including:

[0137] The data acquisition module is used to collect real-time operation data and historical data including water quality parameters and operation parameters.

[0138] Among them, water quality parameters include COD, ammonia nitrogen, total nitrogen and total phosphorus; operating parameters include dissolved oxygen, water temperature, pH value, sludge concentration and reflow ratio.

[0139] The data preprocessing module is used to clean the real-time operation data and historical data to obtain preprocessed feature data.

[0140] The prediction module is used to input the preprocessed feature data into the preset prediction model to obtain the prediction result.

[0141] The comparison module is used to compare the prediction results with the real-time operation data to obtain deviation data.

[0142] The optimization module is used to optimize the corresponding site parameters in the prediction model according to the deviation data, return the optimized parameters to the prediction model, and send them to the corresponding sewage treatment site.

[0143] In one embodiment, the data preprocessing module includes:

[0144] The segmentation unit is used to segment the real-time operation data according to the size of the water inlet flow rate to obtain high load segment, medium load segment and low load segment data.

[0145] The anomaly detection unit is used to set anomaly detection thresholds for the high-load segment, medium-load segment and low-load segment data respectively, and eliminate abnormal data according to the corresponding anomaly detection thresholds.

[0146] The data supplement unit is used to supplement the missing values ​​in the real-time operation data according to the data of the same load segment in the historical data to obtain a complete data set.

[0147] The standardization processing unit is used to perform standardization processing on the complete data set to obtain pre-processed feature data.

[0148] In one embodiment, the prediction module includes:

[0149] The operating status judgment unit is used to determine the operating status of the sewage treatment process according to the operating parameters.

[0150] The water quality prediction unit is used to input the pre-processed characteristic data into a preset water quality prediction model and calculate the pollutant degradation rate in combination with the operating status.

[0151] The effluent water quality prediction unit is used to predict the effluent water quality indicators based on the degradation rate of pollutants.

[0152] The energy consumption and drug consumption prediction unit is used to calculate the energy consumption data and drug consumption data required to achieve the effluent water quality indicators.

[0153] In one embodiment, the operation status determination unit includes:

[0154] The fluctuation detection subunit is used to calculate the change rate of the inlet water quality parameters within the continuous monitoring time period. When the change rate exceeds a preset fluctuation threshold and the duration exceeds a first set time, it is determined to be an impact load state.

[0155] The temperature detection subunit is used to monitor the water temperature change curve. When the water temperature change rate exceeds a preset temperature threshold and the duration exceeds a second set time, it is determined to be a water temperature sudden change state.

[0156] The stable state judgment subunit determines that the stable operation state occurs when the change rate of the inlet water quality parameter and the water temperature change rate do not exceed the corresponding preset thresholds and the operation parameters remain within the set range.

[0157] In one embodiment, the optimization module includes:

[0158] The optimization trigger unit is used to trigger model parameter optimization when the deviation data exceeds a preset deviation threshold.

[0159] The loss function calculation unit is used to calculate the loss function value according to the deviation data.

[0160] The parameter optimization unit is used to iteratively optimize the corresponding site parameters in the prediction model until the loss function value is less than a preset function threshold.

[0161] The parameter updating unit is used to return the optimized parameters to the prediction model.

[0162] In one embodiment, the energy consumption and drug consumption prediction unit includes:

[0163] The air supply volume prediction subunit is used to calculate the air supply volume control target value as energy consumption data based on the influent COD, ammonia nitrogen and aerator oxygen utilization rate in the pre-treated characteristic data.

[0164] The aeration optimization subunit is used to optimize aeration control according to the air supply control target value, specifically including: detecting the actual air supply through the air flow meter, and automatically adjusting the fan operating frequency according to the air supply control target value; adjusting the opening of the air valve in each section of the aerobic pool according to the fan operating frequency and the real-time monitoring value of the dissolved oxygen concentration in each section of the aerobic pool.

[0165] In one embodiment, the energy consumption and drug consumption prediction unit further includes:

[0166] The phosphorus removal dosage prediction subunit is used to calculate the phosphorus removal dosage control target value as the drug consumption data based on the total phosphorus in the influent and the chemical phosphorus removal efficiency in the pretreated characteristic data.

[0167] The dosing optimization subunit is used to optimize the dosing control according to the phosphorus removal dosing dosage control target value, specifically including: detecting the actual dosing amount through the dosing flow meter, and automatically adjusting the dosing pump operation frequency according to the phosphorus removal dosing dosage control target value; adjusting the distribution ratio of the dosing position according to the dosing pump operation frequency and the total phosphorus concentration of the reaction tank effluent.

[0168] In one embodiment, the present application provides an electronic device, which may be a server, and its internal structure diagram may be as follows: Fig. 9 As shown. The electronic device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the electronic device is used to provide computing and control capabilities. The memory of the electronic device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the electronic device is used to store data. The network interface of the electronic device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for intelligent optimization of decentralized rural sewage treatment sites is implemented.

[0169] Those skilled in the art will understand that Fig. 9The structure shown in the figure is merely a block diagram of a partial structure related to the scheme of the present application, and does not constitute a limitation on the electronic device to which the scheme of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different arrangement of components.

[0170] In one embodiment, an electronic device is further provided, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiments when executing the computer program.

[0171] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the above-mentioned computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0172] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A decentralized rural sewage treatment site intelligent optimization method, characterized in that: The steps include: Collecting real-time operation data and historical data including water quality parameters and operation parameters, wherein the water quality parameters include COD, ammonia nitrogen, total nitrogen and total phosphorus; and the operation parameters include dissolved oxygen, water temperature, pH value, sludge concentration and reflow ratio; The real-time operation data is segmented according to the water inlet flow rate to obtain high load segment, medium load segment and low load segment data, wherein the data below the flow threshold Q1 is the low load segment, the data above the flow threshold Q2 is the high load segment, and the data between Q1 and Q2 is the medium load segment; For the high load segment, medium load segment and low load segment data, respectively, an abnormal value detection threshold is set, and abnormal data is eliminated according to the corresponding abnormal value detection threshold; The missing values ​​in the real-time operation data are supplemented according to the data of the same load segment in the historical data to obtain a complete data set; Performing standardization on the complete data set to obtain preprocessed feature data; The preprocessed feature data is input into a preset prediction model to obtain a prediction result, which specifically includes the following steps: S121, determining the operating state of the sewage treatment process according to the water quality parameters and the operating parameters; S122, inputting the pre-processed characteristic data into a preset water quality prediction model, and calculating the pollutant degradation rate in combination with the operating status; S123, predicting effluent water quality indicators based on the pollutant degradation rate; S124, calculating the energy consumption data and drug consumption data required to achieve the effluent water quality index; Comparing the prediction result with the real-time operation data to obtain deviation data; Optimizing the corresponding site parameters in the prediction model according to the deviation data, returning the optimized parameters to the prediction model, and sending them to the corresponding sewage treatment site; The operating state includes a stable operating state, an impact load state and a water temperature mutation state; determining the operating state of the sewage treatment process according to the water quality parameters and the operating parameters specifically includes the following steps: Calculate the change rate of the inlet water quality parameter within the continuous monitoring time period, and determine that it is an impact load state when the change rate exceeds a preset fluctuation threshold and the duration exceeds a first set time period; Monitor the water temperature change curve, and when the water temperature change rate exceeds the preset temperature threshold and the duration exceeds the second set time, it is determined to be a water temperature mutation state; When the change rate of the inlet water quality parameter and the water temperature change rate do not exceed the corresponding preset thresholds, and the operating parameters remain within the set range, it is determined to be in a stable operating state.

2. The method for intelligent optimization of decentralized rural sewage treatment sites according to claim 1, characterized in that: Optimizing the corresponding site parameters in the prediction model according to the deviation data, and returning the optimized parameters to the prediction model, specifically includes the following steps: When the deviation data exceeds a preset deviation threshold, model parameter optimization is triggered; Calculate the loss function value according to the deviation data; Iteratively optimizing the corresponding site parameters in the prediction model until the loss function value is less than a preset function threshold; The optimized parameters are returned to the prediction model.

3. The method for intelligent optimization of decentralized rural sewage treatment sites according to claim 1, characterized in that: The energy consumption data includes an air supply control target value, which is calculated based on the influent COD, ammonia nitrogen and aerator oxygen utilization rate in the pre-treated characteristic data; The method further includes optimizing aeration control according to the air supply control target value, specifically including the following steps: The actual air supply volume is detected by an air flow meter, and the fan operating frequency is automatically adjusted according to the air supply volume control target value; The opening of the air valves in each section of the aerobic pool is adjusted according to the operating frequency of the fan and the real-time monitoring value of the dissolved oxygen concentration in each section of the aerobic pool.

4. According to the intelligent optimization method for decentralized rural sewage treatment sites in claim 1, the drug consumption data includes a phosphorus removal dosage control target value, and the phosphorus removal dosage control target value is calculated based on the total phosphorus in the influent and the chemical phosphorus removal efficiency in the pre-treated characteristic data; The method further includes optimizing the dosing control according to the phosphorus removal dosing control target value, specifically including the following steps: The actual dosage is detected by a dosage flow meter, and the operation frequency of the dosage pump is automatically adjusted according to the phosphorus removal dosage control target value; The allocation ratio of the dosing positions is adjusted according to the operating frequency of the dosing pump and the total phosphorus concentration of the reaction tank effluent.

5. A decentralized rural sewage treatment site intelligent optimization system, characterized in that: include: A data acquisition module, for collecting real-time operation data and historical data including water quality parameters and operation parameters, wherein the water quality parameters include COD, ammonia nitrogen, total nitrogen and total phosphorus; and the operation parameters include dissolved oxygen, water temperature, pH value, sludge concentration and reflow ratio; A data preprocessing module, used for performing data cleaning on the real-time operation data and the historical data to obtain preprocessed feature data; A prediction module, used for inputting the preprocessed feature data into a preset prediction model to obtain a prediction result; A comparison module, used to compare the prediction result with the real-time operation data to obtain deviation data; An optimization module, used to optimize the corresponding site parameters in the prediction model according to the deviation data, return the optimized parameters to the prediction model, and send them to the corresponding sewage treatment site; The data preprocessing module comprises: A segmentation unit is used to segment the real-time operation data according to the size of the water inlet flow rate to obtain high load segment, medium load segment and low load segment data, wherein the data below the flow threshold Q1 is the low load segment, the data above the flow threshold Q2 is the high load segment, and the data between Q1 and Q2 is the medium load segment; An abnormality detection unit, used to set abnormal value detection thresholds for the high load segment, medium load segment and low load segment data respectively, and remove abnormal data according to the corresponding abnormal value detection thresholds; A data supplement unit, used to supplement the missing values ​​in the real-time operation data according to the data of the same load segment in the historical data to obtain a complete data set; A standardization processing unit, used for performing standardization processing on the complete data set to obtain pre-processed feature data; The prediction module comprises: An operating status determination unit, used to determine the operating status of the sewage treatment process according to the water quality parameters and the operating parameters; A water quality prediction unit, used to input the pre-processed characteristic data into a preset water quality prediction model, and calculate the pollutant degradation rate in combination with the operating status; An effluent water quality prediction unit, used to predict effluent water quality indicators based on the pollutant degradation rate; An energy consumption and drug consumption prediction unit, used to calculate the energy consumption data and drug consumption data required to achieve the effluent water quality index; The operating state includes a stable operating state, an impact load state, and a water temperature mutation state; the operating state judgment unit includes: A fluctuation detection subunit, used to calculate the change rate of the inlet water quality parameter within the continuous monitoring time period, and when the change rate exceeds a preset fluctuation threshold and the duration exceeds a first set time, it is determined to be an impact load state; The temperature detection subunit is used to monitor the water temperature change curve. When the water temperature change rate exceeds a preset temperature threshold and the duration exceeds a second set time, it is determined to be a water temperature mutation state; The stable state judgment subunit determines that the operating state is stable when the change rate of the inlet water quality parameter and the water temperature change rate do not exceed the corresponding preset thresholds and the operating parameters remain within the set range.

Images