Intelligent and accurate aeration control method and system in sewage biochemical treatment process

Abstract

The invention belongs to the technical field of aeration control, and discloses an intelligent precise aeration control method and system in a sewage biochemical treatment process, and the method comprises the steps: carrying out the preprocessing of aeration monitoring data, and implementing the matching of working conditions; performing aeration cooperative control prediction on standard aeration monitoring data according to a working condition identification result, performing deviation calculation on an aeration adjustment amount, and optimizing the aeration adjustment amount to obtain an optimized aeration adjustment amount; and combining the optimized aeration amount adjustment amount with the multi-target constraint to generate an optimal aeration control strategy, thereby realizing intelligent and accurate aeration control in the sewage biochemical treatment process. According to the invention, dynamic requirements of reactions such as oxygenolysis and biological denitrification of organic matters in sewage on dissolved oxygen can be accurately grasped, aeration control which is more accurate and more efficient than traditional single dissolved oxygen control is realized, sewage treatment efficiency and stability are effectively improved, energy consumption and sludge yield are reduced, and effluent quality is improved.

Description

Technical Field

[0001] This invention relates to the field of aeration control technology, and in particular to a smart and precise aeration control method and system for wastewater biochemical treatment processes. Background Technology

[0002] Wastewater biological treatment is a core process that utilizes microbial metabolism to degrade pollutants such as organic matter, nitrogen, and phosphorus in wastewater. It mainly includes key biochemical reactions such as organic matter oxidation, nitrification, and denitrification, all of which are highly dependent on the precise supply of dissolved oxygen. Traditional aeration control often uses fixed thresholds or single parameter feedback, which can easily lead to insufficient aeration affecting treatment efficiency or excessive aeration causing energy waste. Therefore, designing a smart and precise aeration control method for wastewater biological treatment is crucial.

[0003] However, existing aeration control methods typically rely on a single parameter, such as aeration control based solely on dissolved oxygen. This lacks comprehensive perception and collaborative analysis of the multi-dimensional dynamic characteristics during wastewater biochemical treatment, making it difficult to accurately reflect the real-time demand for dissolved oxygen. At the same time, traditional control strategies, such as fixed-setpoint PID control, have weak adaptive capabilities and cannot effectively cope with fluctuations in water quality and quantity, seasonal changes, and complex operating conditions. This leads to excessive or insufficient aeration, resulting in problems such as high energy consumption, unstable denitrification efficiency, increased sludge production, and fluctuations in effluent quality. Furthermore, these methods are highly dependent on manual experience for adjustment, lacking sufficient intelligence and precision.

[0004] Therefore, how to provide a smart and precise aeration control method and system for the biochemical treatment of wastewater is an urgent problem to be solved. Summary of the Invention

[0005] This invention provides a smart and precise aeration control method and system for wastewater biochemical treatment processes to solve the aforementioned technical problems in the prior art.

[0006] According to a first aspect of the present invention, a method for intelligent and precise aeration control in a wastewater biochemical treatment process is provided.

[0007] In one embodiment, a smart and precise aeration control method for wastewater biochemical treatment processes includes:

[0008] Acquire aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform operating condition matching based on the standard aeration monitoring data, and obtain operating condition identification results.

[0009] Based on the operating condition identification results, the aeration coordinated control prediction is performed on the standard aeration monitoring data to obtain the aeration adjustment amount. The deviation of the aeration adjustment amount is calculated, and the aeration adjustment amount is optimized based on the deviation calculation results to obtain the optimized aeration adjustment amount.

[0010] The optimal aeration control strategy is generated by combining the optimized aeration volume adjustment with multi-objective constraints, and intelligent and precise aeration control of the wastewater biochemical treatment process is realized based on the optimal strategy.

[0011] According to a second aspect of the present invention, a smart and precise aeration control system for a wastewater biochemical treatment process is provided.

[0012] In one embodiment, the intelligent and precise aeration control system for wastewater biochemical treatment includes:

[0013] The operating condition matching module is used to acquire aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform operating condition matching based on the standard aeration monitoring data, and obtain operating condition identification results.

[0014] The aeration adjustment calculation module is used to predict the coordinated control of aeration based on the standard aeration monitoring data according to the working condition identification results, obtain the aeration adjustment amount, calculate the deviation of the aeration adjustment amount, and optimize the aeration adjustment amount based on the deviation calculation results to obtain the optimized aeration adjustment amount.

[0015] The aeration control strategy generation module is used to combine the optimized aeration volume adjustment with multi-objective constraints to generate the optimal aeration control strategy, and realize intelligent and precise aeration control in the wastewater biochemical treatment process based on the optimal strategy.

[0016] According to a third aspect of the present invention, a computer device is provided.

[0017] In some embodiments, the computer device includes a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps of the above-described intelligent and precise aeration control method for the wastewater biochemical treatment process.

[0018] According to a fourth aspect of the present invention, a computer-readable storage medium is provided.

[0019] In one embodiment, a computer program is stored on a computer-readable storage medium, and when the computer program is executed by a processor, it implements the steps of the above-described intelligent and precise aeration control method for the wastewater biochemical treatment process.

[0020] The technical solutions provided by the embodiments of the present invention may include the following beneficial effects:

[0021] 1. This invention employs a multi-parameter fusion precision aeration control strategy, deeply integrating multiple water quality parameters such as dissolved oxygen, oxidation-reduction potential, ammonia nitrogen, nitrite nitrogen, nitrate nitrogen, and pH (acidity / alkalinity), as well as hydraulic parameters such as liquid level and flow rate, to construct a comprehensive parameter system reflecting the wastewater biochemical treatment process. Through intelligent algorithms, this system comprehensively analyzes multiple parameters, accurately grasping the dynamic demand for dissolved oxygen in wastewater reactions such as organic matter oxidation and decomposition, and biological denitrification. This achieves more precise and efficient aeration control than traditional single dissolved oxygen control, effectively improving wastewater treatment efficiency and stability, reducing energy consumption and sludge production, and enhancing effluent quality.

[0022] 2. This invention utilizes LSTM (Long Short-Term Memory) prediction, fuzzy or PID (Proportional, Integral, and Derivative) control training and inference. It leverages big data analytics to mine and analyze massive amounts of historical operational data, and continuously optimizes the mathematical model and control strategy of the wastewater biochemical treatment process through deviation strategies. This enables the system to automatically adapt to changes in wastewater quality and quantity, as well as treatment needs under different seasons and operating conditions. It can predict dissolved oxygen trends in advance and make precise aeration adjustment decisions, significantly enhancing the system's adaptability and stability in complex operating conditions. This reduces manual intervention, achieving intelligent and refined management, and providing a brand-new intelligent control solution for the wastewater biochemical treatment industry.

[0023] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0024] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0025] Figure 1 This is a schematic flowchart illustrating a smart and precise aeration control method for a wastewater biochemical treatment process according to an exemplary embodiment.

[0026] Figure 2 This is a schematic diagram of the structure of a smart and precise aeration control system for a wastewater biochemical treatment process, according to an exemplary embodiment.

[0027] Figure 3 This is a schematic diagram of the structure of a computer device according to an exemplary embodiment.

[0028] Figure label:

[0029] 201. Operating condition matching module; 202. Aeration adjustment calculation module; 203. Aeration control strategy generation module. Detailed Implementation

[0030] The following description and accompanying drawings fully illustrate specific embodiments described herein to enable those skilled in the art to practice them. Some portions and features of certain embodiments may be included in or replace portions and features of other embodiments. The scope of the embodiments herein includes the entire scope of the claims and all available equivalents thereof. The various embodiments described herein are presented in a progressive manner, with each embodiment focusing on its differences from other embodiments; similar or identical parts between embodiments can be referred to interchangeably.

[0031] The modules in the apparatus or system of this application can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0032] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0033] Figure 1 An embodiment of the intelligent and precise aeration control method for a wastewater biochemical treatment process according to the present invention is shown.

[0034] In this optional embodiment, the intelligent and precise aeration control method for wastewater biochemical treatment includes:

[0035] Step S101: Obtain aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform operating condition matching based on the standard aeration monitoring data, and obtain operating condition identification results.

[0036] Step S102: Based on the working condition identification results, predict the aeration coordinated control of the standard aeration monitoring data to obtain the aeration adjustment amount, calculate the deviation of the aeration adjustment amount, and optimize the aeration adjustment amount based on the deviation calculation results to obtain the optimized aeration adjustment amount.

[0037] Step S103: The optimized aeration volume adjustment is combined with multi-objective constraints to generate the optimal aeration control strategy, and intelligent and precise aeration control of the wastewater biochemical treatment process is realized based on the optimal strategy.

[0038] In this optional embodiment, acquiring aeration monitoring data, preprocessing the aeration monitoring data to obtain standard aeration monitoring data, and performing operating condition matching based on the standard aeration monitoring data to obtain operating condition identification results include: acquiring aeration monitoring data during the wastewater biochemical treatment process, processing the aeration monitoring data for missing and outlier values ​​to obtain cleaned aeration monitoring data, and performing data format standardization conversion on the cleaned aeration monitoring data to obtain standard aeration monitoring data; performing feature fusion on the standard aeration monitoring data according to the process mechanism and generating associated features, and using fuzzy logic to perform operating condition matching and identification on the associated features to obtain operating condition identification results.

[0039] In this optional embodiment, the aeration monitoring data includes water quality data, hydraulic data, and aeration equipment operation data; the water quality data includes: dissolved oxygen data, oxidation-reduction potential data, ammonia nitrogen, nitrite nitrogen, and nitrate nitrogen data, and pH data; the hydraulic data includes: liquid level data and flow rate data; the aeration equipment operation data includes: variable frequency Roots blower operation data and microporous aerator group operation data.

[0040] In this optional embodiment, feature fusion includes: calculating the synergistic value of dissolved oxygen and redox potential, the ratio of ammonia nitrogen to nitrate nitrogen concentration, and the coupling value of flow rate and liquid level.

[0041] In this optional embodiment, the aeration adjustment amount is obtained by predicting the aeration coordinated control based on the standard aeration monitoring data according to the working condition identification result, and the deviation of the aeration adjustment amount is calculated. Based on the deviation calculation result, the aeration adjustment amount is optimized to obtain the optimized aeration adjustment amount. This includes: determining the initial aeration coordinated control weight based on the working condition identification result; predicting the aeration coordinated control based on the standard aeration monitoring data according to the initial aeration coordinated control weight and combined with the pre-acquired historical aeration monitoring data to obtain the aeration adjustment amount; calculating the deviation between the aeration adjustment amount and the preset target adjustment amount, and judging the level of the deviation calculation result according to the preset deviation grading rule to obtain the deviation level judgment result; matching the control path based on the deviation level judgment result, and compensating and correcting the aeration adjustment amount according to the level deviation based on the control path and the deviation calculation result to obtain the optimized aeration adjustment amount.

[0042] In this optional embodiment, the aeration adjustment amount is obtained by predicting the aeration adjustment based on the initial aeration collaborative control weights and combined with pre-acquired historical aeration monitoring data to perform aeration collaborative control. This includes: setting initial parameters and basic dissolved oxygen control targets for the standard aeration monitoring data using proportional, integral, and derivative control to obtain the basic aeration adjustment amount; fuzzifying the parameters of the basic aeration adjustment amount based on fuzzy inference to obtain the fuzzified aeration adjustment amount, and performing rule matching and defuzzification on the fuzzified aeration adjustment amount to obtain the preliminary aeration adjustment amount; constructing a neural network prediction model based on the pre-acquired historical aeration monitoring data and the initial aeration collaborative control weights, training and optimizing the neural network prediction model to obtain a multi-dimensional correlation model, and using the multi-dimensional correlation model to optimize the preliminary aeration adjustment amount to obtain the aeration adjustment amount.

[0043] In this optional embodiment, a neural network prediction model is constructed based on pre-acquired historical aeration monitoring data and initial aeration collaborative control weights. The neural network prediction model is then trained and optimized to obtain a multi-dimensional correlation model. The initial aeration adjustment amount is optimized using the multi-dimensional correlation model to obtain the aeration adjustment amount, which includes: determining the prediction target and boundary based on the initial aeration adjustment amount; constructing a time-series feature set based on pre-acquired historical aeration monitoring data; initializing the parameters of the neural network prediction model using the time-series feature set and initial aeration collaborative control weights to obtain a long short-term memory network model; dividing the time-series feature set into a training set and a validation set; using the preset minimum effluent water quality and minimum aeration energy consumption as the dual loss function; and iteratively training the long short-term memory network model using the gradient descent method to obtain the multi-dimensional correlation model; inputting the initial aeration adjustment amount into the multi-dimensional correlation model; sequentially passing through the long short-term memory network layer and the full-chain layer, and combining a gating unit to capture time-series dependencies to optimize and obtain the aeration adjustment amount.

[0044] In this optional embodiment, the matching rules for matching control paths based on deviation level judgment results include: if the deviation level judgment result is a minor deviation, then a proportional, integral, and derivative control path is matched; if the deviation level judgment result is a slight deviation, then a fuzzy rule-corrected control path is matched; if the deviation level judgment result is a severe deviation, then a multi-dimensional correlation model control path is matched.

[0045] In this optional embodiment, the optimal aeration control strategy is generated by combining the optimized aeration volume adjustment with multi-objective constraints. The intelligent and precise aeration control of the wastewater biochemical treatment process based on the optimal strategy includes: constructing multi-objective constraints with water quality indicators as the core constraint, and combining hydraulic spatial distribution requirements and equipment energy consumption requirements; optimizing the optimized aeration volume based on the multi-objective constraints and using a multi-objective optimization algorithm to obtain the optimal aeration control strategy; and converting the optimal aeration control strategy into aeration control commands and sending them to the programmable logic controller via an industrial communication protocol to drive the aeration equipment in the wastewater biochemical treatment process.

[0046] Figure 2 An embodiment of a smart and precise aeration control system for a wastewater biochemical treatment process according to the present invention is shown.

[0047] In this optional embodiment, the intelligent and precise aeration control system for the wastewater biochemical treatment process includes:

[0048] The working condition matching module 201 is used to acquire aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform working condition matching based on the standard aeration monitoring data, and obtain working condition identification results.

[0049] The aeration adjustment calculation module 202 is used to predict the aeration coordinated control of standard aeration monitoring data based on the working condition identification results, obtain the aeration adjustment amount, calculate the deviation of the aeration adjustment amount, optimize the aeration adjustment amount based on the deviation calculation results, and obtain the optimized aeration adjustment amount.

[0050] The aeration control strategy generation module 203 is used to combine the optimized aeration volume adjustment with multi-objective constraints to generate the optimal aeration control strategy, and realize intelligent and precise aeration control of the wastewater biochemical treatment process based on the optimal strategy.

[0051] To facilitate understanding of the above technical solutions of the present invention, the following further explains the above technical solutions of the present invention from the perspective of architecture and principle, as follows:

[0052] It should be further noted that the parameter monitoring module includes the following water quality parameter sensor group: dissolved oxygen sensor; oxidation-reduction potential (ORP) sensor; ammonia nitrogen, nitrite nitrogen, and nitrate nitrogen sensors; and pH sensor. The hydraulic parameter sensor group includes: level sensor and flow sensor. The dissolved oxygen sensor monitors the dissolved oxygen concentration at various points in the aeration tank in real time, offering high accuracy, rapid response, and adjustable data update frequency to ensure timely detection of even minor changes in dissolved oxygen. It has a wide measurement range (0-20 mg / L) and an accuracy of ±0.1 mg / L. Installed at different depths and locations in the aeration tank, it forms a three-dimensional monitoring network, comprehensively reflecting the dissolved oxygen distribution within the tank. The oxidation-reduction potential (ORP) sensor is used to determine the extent of oxidation-reduction reactions in wastewater, indirectly reflecting the oxidation and decomposition of organic matter. It works in conjunction with dissolved oxygen parameters to optimize aeration strategies. It has a measurement range of -500 mV to +1500 mV and an accuracy of ±10 mV, and is installed in key reaction areas of the aeration tank. Ammonia nitrogen, nitrite nitrogen, and nitrate nitrogen sensors: These sensors accurately monitor the dynamic changes in nitrogen forms in wastewater, providing a basis for aeration control in the biological denitrification process, achieving precise oxygen supply, and promoting the balanced nitrification and denitrification reactions. The measurement range and accuracy of each sensor are as follows: ammonia nitrogen (0-50 mg / L, ±0.5 mg / L), nitrite nitrogen (0-5 mg / L, ±0.1 mg / L), and nitrate nitrogen (0-100 mg / L, ±1 mg / L). They are placed in key locations such as the boundary between the aerobic and anoxic zones of the aeration tank. pH sensor: Wastewater biological treatment is sensitive to pH. This sensor monitors the pH value in the aeration tank in real time, ensuring that microorganisms are within the suitable pH range for growth, perfectly matching the aeration process with microbial metabolic activities. The measurement range is 0-14 pH, with an accuracy of ±0.1 pH. It is installed in a location with uniform water flow within the tank. The hydraulic parameter sensor group includes a level sensor: an ultrasonic level gauge for non-contact measurement of water level in the aeration tank, with a range of 0-10m and an accuracy of ±1cm. This provides a water level reference for adjusting the operation of the aeration equipment, preventing poor aeration or equipment damage due to abnormal water levels. It is installed at a suitable location on the aeration tank wall. Flow sensors are installed on the inlet pipe, return sludge pipe, and main aeration pipe to monitor inlet and outlet flow rates, return sludge flow rate, and aeration air flow rate. The inlet flow sensor has a range of 0-1000L / s and an accuracy of ±0.5%; the return sludge flow sensor has a range of 0-500L / s and an accuracy of ±1%; and the aeration air flow sensor has a range of 0-1000m³ / h and an accuracy of ±1.5%, ensuring accurate and controllable parameters throughout the wastewater flow and aeration cycle process, providing hydraulic data support for precise aeration control.

[0053] The aeration execution module includes a variable frequency Roots blower and a microporous aerator assembly. The variable frequency Roots blower features high efficiency and a wide air volume adjustment range. The blower speed is controlled by a frequency converter, achieving stepless and precise adjustment of the aeration volume. The blower's single-unit air volume is 150~1000 m³ / min (≈9000~60000 m³ / h), power is 200~1000 kW, and air pressure is 30~80 kPa. It can operate stably within a frequency range of 20Hz-50Hz, meeting the dynamic aeration requirements of the aeration tank under different working conditions. It also features low noise, long lifespan, and ensures a long-term stable air supply to the system. The equipment is equipped with an air intake filtration system that effectively filters impurities and dust from the air, preventing foreign objects from entering the blower and aeration pipes, causing blockages or damage. It also has an automatic cleaning function, periodically cleaning the filter to ensure unobstructed air intake and maintain efficient blower operation. For the microporous aerator group, a new type of silicone microporous aerator is used, which has excellent anti-clogging performance and high aeration uniformity. Installed at the bottom of the aeration tank in a matrix distribution, it produces small bubbles (e.g., 2-3 mm in diameter) with a large contact area with the wastewater. Each aerator is independently controllable, with its air intake adjusted by a solenoid valve. Based on the local dissolved oxygen demand feedback from the parameter monitoring module, the aeration intensity of each microporous aerator is precisely controlled, achieving a refined and uniform distribution of dissolved oxygen within the aeration tank. The aerator material is highly corrosion-resistant, enabling long-term stable operation in complex wastewater environments. Its smooth surface does not easily adhere to dirt, reducing maintenance workload. It also supports online monitoring and fault diagnosis; in the event of local blockage or damage, the system automatically alarms and switches to a backup aeration unit, ensuring uninterrupted aeration.

[0054] The intelligent control module includes a core controller and an intelligent decision-making system. The core controller utilizes a high-performance industrial-grade PLC (Programmable Logic Controller) as the system's core control unit, possessing powerful data processing capabilities, high reliability, and anti-interference performance, capable of adapting to the complex electromagnetic environment of wastewater treatment sites. The PLC connects to various parameter monitoring sensors and actuators via communication interfaces, acquiring monitoring data in real time, processing it rapidly, and outputting corresponding control signals according to preset control strategies to achieve precise control of the aeration equipment. It incorporates multiple intelligent control algorithms, such as fuzzy control, PID control, and neural network control, which can adaptively adjust control parameters and optimize aeration control effects based on the nonlinear and time-varying characteristics of the wastewater biochemical treatment process. Simultaneously, it has data storage capabilities, storing the current operating conditions, algorithms, and effect data in the core controller, ensuring storage for ≥3 years for future neural network learning. This includes monitoring parameters, equipment operating status, and control commands, providing data support for subsequent data analysis, fault diagnosis, and system optimization. PID control is the fundamental and precise control under stable operating conditions. Its core principle, based on the core logic of deviation and correction, uses proportional (P), integral (I), and derivative (D) actions to stabilize key monitored parameters such as dissolved oxygen (DO) and ammonia nitrogen within the preset target range. It is suitable for routine operating conditions with small fluctuations in water quality or quantity and stable microbial activity, serving as the basic stabilization layer for aeration control. Fuzzy control, on the other hand, provides rapid response under nonlinear and large disturbance conditions. Its core principle addresses the emphasized nonlinear characteristics of wastewater biochemical treatment, such as sudden increases in COD in industrial wastewater or abrupt changes in influent flow rate. Through three steps—fuzzification, rule-based reasoning, and defuzzification—it transforms the qualitative experience of multiple parameters, including DO, ORP, ammonia nitrogen, and flow rate (i.e., all parameters of the parameter monitoring module), into quantitative control commands without establishing a precise mathematical model. It serves as the disturbance response layer for aeration control. Neural network control is an adaptive optimization under long-term time-varying characteristics. Its core principle is based on the technical framework of big data analysis and machine learning to build a multi-dimensional correlation model of water quality, hydraulics and equipment. By learning historical operating data, that is, data stored in the core controller for ≥3 years, it optimizes PID parameters and fuzzy rules, predicts DO change trends, and adapts to time-varying characteristics, such as fluctuations in microbial activity caused by seasonal temperature changes and long-term drift of influent load. It is a long-term optimization layer for aeration control.A multi-dimensional correlation model of water quality, hydraulics, and equipment, built upon a framework of big data analysis and machine learning, is applied to the aeration process of wastewater biochemical treatment as follows: It integrates comprehensive data from the entire wastewater biochemical aeration process, including influent water quality (COD, NH3-N, TP, water temperature); hydraulic conditions (influent flow rate, tank level, sludge concentration MLSS, dissolved oxygen DO); and aeration equipment operation (air volume, frequency, air pressure, energy consumption, start / stop status). After preprocessing and feature correlation, machine learning algorithms are used to train and fit the coupling patterns of data from each dimension, establishing a nonlinear mapping correlation model of water quality load, hydraulic conditions, and aeration equipment operating parameters. The model takes real-time water quality and hydraulic data as input and accurately outputs optimal aeration air volume, frequency, and other control parameters to meet the current biochemical reaction requirements. It dynamically matches the oxygen demand of the aerobic tank's microbial nitrification degradation reaction, balancing water quality compliance with aeration energy saving, achieving intelligent and precise control of the aeration process, and adapting to complex conditions of fluctuating influent load.

[0055] Adaptive control is achieved through a four-step process: condition-triggered execution, hierarchical execution, collaborative optimization, and feedback closed-loop. The steps are as follows: real-time acquisition of multiple parameters and initial condition assessment (real-time monitoring); hierarchical algorithm execution (data analysis and decision-making); precise execution and feedback monitoring (precise aeration execution and feedback); and adaptive closed-loop adjustment (feedback and adaptive adjustment). The essence of multi-algorithm collaboration is that PID control ensures basic stability, fuzzy control handles sudden nonlinear disturbances, and neural network control adapts to long-term time-varying characteristics. Through the logical scheduling of the intelligent control module, an upgrade from passive correction to active prediction is achieved.

[0056] For intelligent decision-making systems, a mathematical model (i.e., a multi-dimensional correlation model) of the wastewater biochemical treatment process is constructed based on big data analytics and machine learning technologies. The model's input parameters include water quality, hydraulic data, and aeration equipment operating data, and it outputs the optimal aeration control strategy. The system receives monitoring data in real time, compares and analyzes it with the model, predicts the changing trend of dissolved oxygen in the aeration tank, and adjusts the aeration rate in advance to achieve precise aeration control. Clearly defining the model's objectives and boundaries is crucial; that is, the specific application scenarios and expected goals of the machine learning model must be clearly defined, along with the input and output boundary conditions, ensuring that the model development direction is highly consistent with business needs and avoiding deviations from the objectives during subsequent development. Full data collection and storage management involves systematically collecting complete datasets from relevant fields according to the model's objectives, establishing a standardized data storage mechanism, including raw data backup, version control, and access permission management, laying the data foundation for subsequent analysis and processing. Data preprocessing and cleaning involve quality checks on the collected raw data, handling missing values, outliers, and duplicate data, standardizing data formats, and performing data sampling or enhancement when necessary to ensure that the data quality meets the modeling requirements. Feature engineering and variable construction involve extracting and constructing meaningful feature variables based on business understanding and data analysis, performing feature selection, dimensionality reduction, and transformation, optimizing feature representation, and improving the model's ability to identify key features. Model structure design and selection involves choosing an appropriate algorithm framework, designing a reasonable network structure or model architecture, and determining the initial range of hyperparameters based on the problem type and data characteristics, thus preparing the model for training. Model training and iterative optimization involve training the model using preprocessed data, evaluating model performance through cross-validation, continuously optimizing model performance using hyperparameter tuning strategies, and recording changes in key metrics during training. Model validation and production deployment involves validating the model's generalization ability on independent test sets, evaluating actual performance through A / B testing, completing model packaging and interface development, establishing a monitoring mechanism, and then officially deploying the model online, continuously tracking its operational status. The optimal aeration control strategy derived from input parameters requires a closed-loop process encompassing parameter monitoring, data processing, algorithm computation, strategy generation, and feedback. The specific steps are as follows, strictly adhering to built-in modules and technical requirements throughout: Input parameter acquisition: The core of the parameter monitoring module is collecting three types of input parameters—water quality, hydraulics, and equipment—ensuring parameter coverage of all dimensions of wastewater biochemical reactions and providing raw data support for subsequent analysis. Input parameter preprocessing: The collected raw parameters undergo anomaly removal, deviation correction, and standardization to ensure the data input to the algorithm meets process logic and accuracy requirements. Input parameter feature fusion and operating condition identification: Based on a multi-parameter fusion-based precise aeration control strategy, the preprocessed input parameters are correlated with features and matched to operating conditions to clarify the core needs of the current wastewater biochemical reactions, such as organic matter degradation and denitrification, providing a basis for algorithm selection.Intelligent algorithms work collaboratively, using the results of operating condition identification to call upon built-in PID control, fuzzy control, and neural network control combined algorithms to calculate the aeration volume adjustment. The optimal aeration control strategy is generated by converting the airflow adjustment output by the algorithm into specific instructions recognizable by the aeration execution module, while incorporating equipment safety constraints to form the optimal strategy. A closed-loop strategy output and feedback mechanism sends the generated optimal strategy to the execution module, while continuous feedback through the parameter monitoring module ensures that the control effect meets actual requirements. The entire process strictly relies on three core modules: parameter monitoring, intelligent control, and aeration execution. Input parameters cover all sensor data, algorithm calculations follow built-in PID, fuzzy, and neural network logic, and strategy outputs adapt to equipment characteristics, achieving a closed loop from input parameters to the optimal control strategy. This perfectly aligns with the core objective of intelligent and precise aeration, ensuring a 20%-35% reduction in energy consumption and an effluent compliance rate of ≥98%.

[0057] This process, involving data alignment, multi-dimensional comparison, deviation assessment, and strategy adaptation, ensures that the comparison results provide accurate basis for adjusting aeration control strategies. The specific content of the real-time monitoring data and model comparison analysis focuses on the model output values ​​and real-time monitoring data, covering all dimensions of parameters related to water quality, hydraulics, and equipment, while simultaneously verifying the consistency of process logic. The core control parameter comparison focuses on key process targets and core control parameters, such as dissolved oxygen (DO) and ammonia nitrogen (NH4). + -N), nitrate nitrogen (NO3) --N) These parameters directly determine the aeration control effect; the focus of comparison is on numerical deviation and prediction accuracy. Auxiliary monitoring parameters are compared to verify process synergy. Auxiliary parameters include oxidation-reduction potential (ORP), pH value, and influent flow rate. Although these parameters do not directly control aeration, they reflect process synergy; the focus of comparison is on trend consistency. Equipment operating parameters are compared to verify the effectiveness of control commands. Equipment parameters include fan frequency, aerator differential pressure, and fan outlet pressure; the focus of comparison is on the matching degree between model commands and actual operating conditions. Process logic is compared to verify the fit between the model and actual operating conditions. Based on the multi-parameter fusion process mechanism, the consistency between the model's preset logic and actual monitoring data is compared. The implementation steps for real-time monitoring data and model comparison analysis are compared and analyzed. This is achieved in six steps: real-time reception, rapid preprocessing, model invocation, multi-dimensional comparison, deviation assessment, and result application. The system's real-time monitoring data and model comparison analysis is essentially a dynamic verification process based on parameter accuracy, process logic, and control effect. Multi-dimensional comparisons ensure consistency between model output and actual operating conditions, while deviation assessment provides clear direction for subsequent strategy adjustments. This perfectly aligns with the core design concepts of intelligent decision-making systems and adaptive adjustments, guaranteeing the accuracy and stability of aeration control. The prediction of dissolved oxygen (DO) trends in the aeration tank is driven by time-series data, multi-parameter fusion, and machine learning modeling. It is achieved through 10 steps, including data preparation, model building, real-time prediction, and feedback optimization, and is specifically adapted to the nonlinear and time-varying characteristics of wastewater biochemical treatment. The specific steps include: determining the prediction target and boundary; screening prediction input parameters; real-time data acquisition and historical data retrieval; data preprocessing; constructing a time-series feature set to capture the DO change pattern; and building an LSTM neural network prediction model. The core of the LSTM model includes an input layer, an LSTM layer (i.e., a three-gate, one-state core structure), an optional Dropout layer, a fully connected layer, and an output layer, with the layers progressing sequentially. Training requires data preprocessing, temporal dataset partitioning, model instantiation, and compilation (i.e., Adam optimizer and regression loss function, early stopping training, and prediction evaluation). Key parameters include the number of LSTM neurons, time step, and batch size. In the field of wastewater biochemical treatment, combined with the temporal characteristics of process indicators, LSTM is used to capture long-term dependencies to predict water quality and support intelligent control. The input is set to hourly-level time-series multiple features, i.e., influent water quality, process parameters, and environmental factors, and the output is core compliance indicators such as effluent COD and NH3-N, covering the complete causal chain. It can rely on existing monitoring equipment to collect data, construct samples to predict the indicators of the second hour based on the data of the previous hour, and implement the system by fine-tuning the core parameters; it can also be used for model training and optimization; real-time prediction execution; visualization of prediction trends and application for decision-making; feedback correction and closed-loop adaptive operation.Trend Analysis: Extract the predicted DO trend, i.e., rising, falling, or stable, and the predicted value for the next 5-10 minutes. Calculate the deviation from the target DO of 1.5-2.5 mg / L. For example, if the predicted value is 2.6 mg / L, the deviation is +0.1 mg / L. Threshold Judgment: If the deviation is ≤0.2 mg / L (minor deviation), maintain the current aeration rate; if the deviation is >0.2 mg / L, adjustment is required, proceed to the next step. Strategy Matching: For slight deviations (0.2-0.5 mg / L), use PID control for fine-tuning. For example, if DO is 0.3 mg / L low, increase the fan frequency by 1 Hz. For severe deviations (>0.5 mg / L), call fuzzy control, combining ammonia nitrogen and flow parameters to calculate the adjustment amount. For example, if DO drops sharply by 0.6 mg / L, increase the fan frequency to 3 Hz and increase the opening of local aerators by 20%. Command Execution: Convert the adjustment amount into fan frequency and aerator opening commands, and send them to the execution module via industrial Ethernet. The response time is less than 1 second. Feedback Correction: Real-time monitoring of adjusted DO; if the deviation returns to ≤0.2mg / L, maintain the instruction; otherwise, repeatedly determine the prediction target and boundary; filter prediction input parameters; real-time data acquisition and historical data retrieval; data preprocessing; until the target is met. Continuously learn and accumulate historical operating data and operational experience, optimize model parameters, improve prediction accuracy and control effect, support remote online upgrades and maintenance, and ensure the system is always in optimal operating condition. Data Preprocessing: Filter historical operating data of more than three months, i.e., DO, ammonia nitrogen, fan parameters, etc., remove outliers, classify by operating condition, i.e., stable or sudden change, and associate with corresponding operation records, such as manually adjusted aeration volume. Experience-based Regularization: Transform operational experience, such as DO needing to exceed 2.5mg / L when ammonia nitrogen suddenly increases, into quantitative rules, such as increasing the model's DO target value by 0.3mg / L when ammonia nitrogen is greater than 8mg / L, embedding these into the model constraints. For parameter iterative optimization, the machine learning model fine-tunes the LSTM weights using classified data, such as increasing the feature weights of ammonia nitrogen and aeration rate, and adjusting the learning rate from 0.001 to 0.0005. Control parameters: PID Kp / Ki / Kd are adjusted based on experience; for example, Kp is increased from 2.0 to 3.0 for sudden water quality changes. Validation and evaluation: The optimized model is tested using data from the past week. If the prediction bias decreases by ≥15% and the control response speed improves by ≥10%, it is considered effective. Periodic updates: The prediction target and boundaries are repeatedly determined weekly; prediction input parameters are filtered; real-time data acquisition and historical data retrieval are performed; data preprocessing is conducted, and continuous iteration is performed using new data and experience to ensure the model adapts to changes in operating conditions.

[0058] The communication module includes a wired communication submodule and a wireless communication submodule. The wired communication submodule uses industrial Ethernet, such as Profibus or Modbus TCP / IP, to tightly connect the parameter monitoring module, aeration execution module, and intelligent control module, ensuring high-speed, stable, and reliable data transmission with a bandwidth of 1000Mbps and a data transmission latency of less than 10ms. This meets the system's high real-time control requirements, features simple wiring, strong anti-interference capabilities, and is suitable for installation and deployment in the complex industrial environments of wastewater treatment plants. The wireless communication submodule is equipped with a 4G / 5G wireless communication module to upload key system operating data, such as dissolved oxygen concentration, aeration rate, and equipment status, to the cloud monitoring platform in real time. It also receives remote operation commands, enabling remote monitoring and management of the wastewater biochemical treatment aeration process. The wireless communication module features a low-power design, wide signal coverage, and strong penetration, ensuring stable signal transmission even in the complex architectural structures of large wastewater treatment plants, meeting the needs of applications such as mobile monitoring and multi-plant collaborative management.

[0059] The human-machine interface module includes a local monitoring terminal and a remote monitoring platform. The local monitoring terminal is a touchscreen industrial computer located in the wastewater treatment plant's central control room, equipped with an intuitive and simple graphical user interface. It displays various monitoring parameters of the aeration tank in real time, presented in dynamic charts and numerical forms; equipment operating status is indicated through visual methods such as icon flashing and color changes, showing normal, fault, and maintenance status; system alarm information is displayed via pop-up notifications and audible alarms; and historical data curves are provided, allowing on-site operators to promptly understand the system's operating status and perform on-site operations and parameter adjustments. It supports operation permission management, with different levels of operators having different access permissions to ensure the system's safe and stable operation. For example, ordinary operators can view real-time data and perform routine equipment start-up and shutdown operations; engineers can modify system parameters and optimize control strategies; and administrators can allocate and manage user permissions. It also has an operation log function, recording the time, operator, and operation content of each operation for easy traceability and querying. The remote monitoring platform is a cloud-based remote monitoring system accessible via a browser or mobile app. Authorized users can view detailed operational data of the wastewater biochemical treatment process anytime, anywhere, remotely operate aeration equipment, receive alarm notifications, and perform remote fault diagnosis and handling, achieving unattended operation and intelligent management. The platform supports centralized monitoring and management of multiple wastewater treatment plants, enabling unified configuration, data comparison and analysis, and performance evaluation of aeration systems in different plants, facilitating optimization of the overall wastewater treatment process and operating costs. Furthermore, it features data security protection mechanisms, employing encrypted transmission, firewalls, and user authentication technologies to ensure the security and confidentiality of system data.

[0060] The system's working principle includes: real-time monitoring, data analysis and decision-making, precise aeration execution, and feedback and adaptive adjustment. In the real-time monitoring phase, various sensors in the parameter monitoring module collect water quality and hydraulic parameters in the aeration tank in real time and transmit the data to the intelligent control module via the communication module according to the set sampling frequency. In the data analysis and decision-making phase, after receiving the monitoring data, the core controller uses a built-in intelligent control algorithm and intelligent decision-making system to analyze and process the data. Based on a pre-established wastewater biochemical treatment model, the real-time monitoring data is compared with the model's predicted values ​​to determine the degree of matching between the current dissolved oxygen level in the aeration tank and the wastewater treatment process. If the dissolved oxygen concentration is too low, it may inhibit microbial activity, affecting organic matter degradation and biological nitrogen removal efficiency; if the concentration is too high, it will cause energy waste and may lead to problems such as sludge loosening. Then, considering factors such as oxidation-reduction potential, nitrogen form changes, pH value, and hydraulic parameters, the optimal aeration control strategy is formulated through fuzzy reasoning and neural network optimization algorithms, including determining the operating frequency of each variable frequency Roots blower and adjusting the opening degree of the microporous aerator group, among other specific parameter settings. The pre-established wastewater biochemical treatment model revolves around the construction of biochemical reaction mechanisms and data-driven control. Specifically, the model's overall architecture adopts a two-layer fusion structure: the bottom mechanism layer is based on the biochemical reaction laws of the activated sludge process, such as nitrification or denitrification and organic matter degradation, constructing parameter correlation logic; the upper intelligent layer integrates machine learning and control algorithms to achieve prediction and decision-making, adapting to nonlinear and time-varying characteristics. The core sub-model and mathematical foundation include a water quality parameter prediction sub-model, namely an LSTM neural network, aiming to predict core parameters such as DO, ammonia nitrogen, and nitrate nitrogen for the next 5-10 minutes; the input is real-time and historical data, such as DO, ORP, influent flow rate, and blower frequency, including derived features such as DO change rate, ammonia nitrogen, and blower coupling value; the structure is a two-layer LSTM, with 32 and 16 neurons and one fully connected layer, outputting predicted values ​​and confidence intervals; the mathematical foundation is the capture of time-series dependencies through gating units, the loss function is mean squared error (MSE), and the optimizer is Adam with a learning rate of 0.001. The aeration control strategy sub-model, employing a multi-algorithm collaboration, utilizes PID control to adjust aeration volume under stable operating conditions. Fuzzy control addresses sudden changes in water quality by fuzzifying deviations such as dissolved oxygen (DO) and ammonia nitrogen (AM) and matching them to a rule base; for example, a sudden increase in AM due to low DO leads to a significant increase in aeration volume. Neural network correction optimizes PID parameters under denitrification-dominant conditions using LSTM; for instance, when AM exceeds 5 mg / L, Kp is increased to 3.0. The model's core modules and hierarchical connections include a three-layer structure: an input sensing layer, a core computation layer, and an output execution layer. The module interconnection involves the sensing layer collecting data, the computation layer performing coupled calculations, and the execution layer outputting control quantities.The core modules are a water quality parameter analysis module, a dissolved oxygen (DO) prediction module, an aeration volume control module, and an operating condition feedback correction module. These modules are bidirectionally interconnected, with the feedback module transmitting operating condition data back to the computational layer in real time for dynamic correction. The model training steps and key parameters include: sample set construction (collecting historical operating data such as COD, NH3-N, MLSS, water temperature, DO value, and aeration volume of the aerobic tank influent, preprocessing for noise reduction and normalization); model initialization (setting a learning rate of 0.01~0.05, 200~500 iterations, and a batch size of 32:3); training and fitting (iterative training based on gradient descent, using the dual loss functions of achieving effluent water quality standards and minimizing aeration energy consumption); validation and optimization (dividing the training and validation sets into a 7:3 ratio, correcting hyperparameters to a model fit ≥0.92); and final deployment (solidifying model parameters and integrating with the operating condition feedback closed loop). This model is adaptable to wastewater treatment applications and features input / output correlation. Inputs include real-time influent water quality (COD, NH3-N), MLSS, water temperature, real-time DO value, and liquid level in the aerobic tank. Outputs include precise aeration volume or frequency. Input data directly maps to the biochemical reaction load in the aerobic tank. The model couples data through biochemical reaction kinetics, and the output is matched to the microbial nitrification or degradation rate as needed, achieving precise DO control of 2.0~3.0 mg / L. It is suitable for aerobic tank operations in the biochemical treatment of municipal wastewater or industrial wastewater and can be applied based on the above parameters and connections. Input parameters include water quality parameters, hydraulic parameters, and equipment parameters; outputs include future DO trend predictions, blower frequency adjustments, and aerator opening commands. The process constraints are: DO target value of 1.5-2.5 mg / L, pH 6.5-8.5, and ammonia nitrogen ≤5 mg / L in effluent; equipment constraints are: blower frequency 20-50 Hz, outlet pressure ≤120% of design value, and aerator pressure difference ≤3 kPa to prevent clogging. The model iteration mechanism uses newly added historical data, including manual operational experience, for weekly fine-tuning, correcting LSTM weights (e.g., increasing the feature weights of ammonia nitrogen and aeration rate); and updating the fuzzy rule base (e.g., adjusting the correlation threshold between flow rate and DO during seasonal changes). This model, through mechanism and data fusion, achieves a closed loop from parameter prediction to aeration control, fully adapting to the dynamic characteristics of wastewater biological treatment.

[0061] The specific steps for comparing real-time monitoring data with model predictions are as follows: Data alignment involves matching real-time monitoring data with model predictions from the same time point, such as the DO prediction value at t+1 min, according to timestamps to ensure that the comparison objects are parameters at the same point in time. Deviation calculation involves the core parameters (DO, ammonia nitrogen): calculating the absolute deviation (measured value - predicted value) and the relative deviation (measured value - predicted value) / predicted value. 100%; Auxiliary parameters (ORP, flow rate) are used to calculate the deviation of the trend. For example, if the measured DO increases but the predicted DO decreases, a trend conflict is marked. Multi-dimensional verification includes numerical deviation verification: comparing whether the DO deviation is ≤0.2mg / L, i.e., the control accuracy threshold, and whether the relative deviation of ammonia nitrogen is ≤10%; trend consistency verification checks the synergistic trend of ORP and DO, such as whether ORP increases synchronously when DO increases, and the matching trend of flow rate and aeration volume, such as whether aeration volume increases synchronously when flow rate increases. The specific steps for judging the matching degree between dissolved oxygen level and wastewater treatment process include deviation level classification, i.e., Level 1 deviation is a minor deviation, with an absolute DO deviation ≤0.2mg / L and a relative ammonia nitrogen deviation ≤10%, and no trend conflict; Level 2 deviation is a slight deviation, i.e., 0.2mg / L < DO deviation ≤0.5mg / L, or 10% < ammonia nitrogen deviation ≤20%, with a basically consistent trend; Level 3 deviation is a serious deviation, i.e., DO deviation is greater than 0.5mg / L, or ammonia nitrogen deviation is greater than 20%, or there is a trend conflict. The process logic matching assessment includes: nitrification process is ammonia nitrogen degradation; if ammonia nitrogen is greater than 5 mg / L and DO ≥ 2.5 mg / L, it is considered a match; conversely, if ammonia nitrogen is high but DO is low, it is considered a mismatch. Denitrification process is nitrate nitrogen degradation; if nitrate nitrogen is greater than 20 mg / L and ORP is less than 100 mV and DO ≤ 0.5 mg / L, it is considered a match; conversely, if nitrate nitrogen is high but DO is high, it is considered a mismatch. Microbial activity is assessed based on pH being between 6.5 and 8.5 and DO being between 1.5 and 2.5 mg / L; if pH exceeds the limit, it is considered a mismatch. The comprehensive matching degree assessment includes: good matching (Level 1 deviation and all process logics matching); slight mismatch (Level 2 deviation and only one process logic mismatch); severe mismatch (Level 3 deviation and two or more process logic mismatches, requiring immediate adjustment of the aeration rate). Parameter acquisition and preprocessing include the simultaneous acquisition of data such as ORP, ammonia nitrogen / nitrate nitrogen / nitrite nitrogen, pH, and influent flow rate / level, removing outliers such as those exceeding the range, and standardizing to the [0,1] range. Parameter-process correlation mapping includes calculating the ORP-DO ratio (reflecting oxidation intensity) for ORP and DO synergy; calculating the ammonia nitrogen / nitrate nitrogen ratio (determining nitrification progress) and the nitrate nitrogen / nitrite nitrogen ratio (determining denitrification efficiency) for nitrogen speciation; verifying pH within the 6.5-8.5 range (microbial activity range); and calculating the flow rate-level coupling value (determining load fluctuations) for hydraulic parameters. Multi-parameter fusion and operating condition identification involves matching preset operating conditions based on correlation values, such as nitrification-dominant: high ammonia nitrogen and high ORP-DO ratio; denitrification-dominant: high nitrate nitrogen and low ORP. Control strategy output: Adjust the aeration rate according to the operating conditions. For example, increase the DO to 2.5 mg / L for nitrification-dominant systems and decrease the DO to 0.5 mg / L for denitrification-dominant systems. Simultaneously correct with pH and hydraulic parameters, for example, increase the aeration rate by 10% when the flow rate suddenly increases.

[0062] Data preparation involves collecting parameters such as DO, ORP, ammonia nitrogen, and flow rate. Preprocessing includes outlier removal, standardization, and extraction of features such as DO change rate and nitrogen ratio. Fuzzy inference generates a preliminary strategy by fuzzifying input parameters (e.g., describing DO deviation as low, medium, or high membership); matching a rule base (e.g., low DO and high ammonia nitrogen significantly increase aeration rate); and defuzzifying the output to determine the initial aeration rate adjustment (e.g., 30 m³ / h). Neural network optimization parameters involve inputting historical data and fuzzy output to an LSTM model; optimizing fuzzy rule weights (e.g., increasing ammonia nitrogen parameter weight to 40%); adjusting PID parameters (e.g., adjusting Kp from 2.0 to 2.5); and outputting the optimized adjustment. Strategy verification and execution compare the predicted DO of the optimized strategy with the target value. If the deviation is ≤0.2 mg / L, the strategy is sent to the aeration execution module; otherwise, the fuzzy inference process is repeated to generate the preliminary strategy and neural network optimization parameters. The precise aeration execution stage involves the core controller sending control commands to the aeration execution module based on the optimized control strategy. The variable frequency Roots blower adjusts its speed according to the received frequency signal, changing the aeration airflow. The solenoid valves of the microporous aerator group adjust the air intake according to control commands, achieving precise distribution of aeration volume to each area in the aeration tank. Simultaneously, the system monitors the operating status of the aeration equipment in real time, such as blower current, voltage, and aerator pressure, ensuring stable operation according to control commands and further guaranteeing the accuracy of the aeration effect. During the feedback and adaptive adjustment phase, the parameter monitoring module continuously collects changes in various parameters of the aeration tank and feeds them back to the intelligent control module. The system compares the real-time feedback data with the expected control target. If a deviation exceeds the allowable range, the intelligent decision-making system immediately reassesses the current operating conditions, analyzes the causes of the deviation (e.g., sudden changes in water quality, equipment failure), and automatically adjusts the control strategy, dynamically compensating and optimizing the aeration volume until parameters such as dissolved oxygen in the aeration tank stabilize within the optimal range, achieving closed-loop adaptive control of the system, as shown in Table 1 below.

[0063] Table 1 Allowable Deviation Range of Parameters

[0064]

[0065] The comparison steps between real-time feedback data and expected control targets include data alignment, matching feedback data (e.g., DO, ammonia nitrogen) with millisecond-level timestamps and monitoring point locations; and ensuring spatiotemporal consistency with expected targets (e.g., DO 1.5-2.5 mg / L). Deviation calculation involves the core parameters (DO, ammonia nitrogen), i.e., calculating the absolute deviation (feedback value - target value) or relative deviation ((feedback value - target value) / target value × 100%). Auxiliary parameters (ORP, flow rate) are used to verify whether the changing trend is consistent with the target trend; for example, if the target DO increases, does the feedback ORP increase synchronously? Deviation classification categorizes deviations into minor deviations, slight deviations, and severe deviations based on the calculation results, providing a basis for subsequent adjustments. The implementation steps of the automatic adjustment control strategy include deviation level identification, that is, determining the deviation level based on the comparison results between feedback data and the target, such as slight deviation, mild deviation, and severe deviation; algorithm matching, which uses PID to fine-tune parameters for slight deviation, superimposed fuzzy rule correction for mild deviation, and calls neural network optimization model for severe deviation; and instruction generation, which converts the optimized adjustment amount, such as fan frequency +2Hz and aerator opening degree +15%, into execution instructions and sends them to the aeration module. The dynamic compensation and optimization adjustment steps for aeration volume include: compensation calculation, i.e., calculating the basic compensation based on the magnitude of the deviation, for example, if DO is 0.3 mg / L low, the compensation air volume is increased by 10 m³ / h; combined with operating condition correction, for example, if the flow rate suddenly increases by 20%, the compensation amount is increased by 15%, and if ammonia nitrogen is high, an additional 5 m³ / h is added; graded adjustment includes macroscopic: adjusting the fan frequency to 20-50 Hz, stepless adjustment; microscopic: adjusting the aerator opening in zones, for example, opening to 80%-100% in areas with low DO; real-time verification: monitoring DO changes every 1 second, and if the deviation still exceeds the allowable range after compensation, repeating the compensation calculation and graded adjustment. The steps to determine the stability of parameters such as dissolved oxygen include numerical deviation verification, namely, absolute deviation of DO ≤ 0.2 mg / L, relative deviation of ammonia nitrogen ≤ 10%, pH between 6.5 and 8.5, and no deviation exceeding the tolerance for 3 minutes. The core algorithm and feature extraction of the intelligent aeration control system, namely the intelligent precision aeration control system for wastewater biochemical treatment, usually adopts a dual-drive architecture of mechanism model and AI algorithm, mainly including: prediction layer: LSTM neural network predicts influent load, and MPC model predicts DO demand; the core logic is to first use LSTM to predict future influent load as MPC perturbation input, and then use MPC combined with process constraints to optimize DO demand to achieve precise control. The core module and hierarchical connection of the mechanism model are designed with a three-layer core architecture, from top to bottom: water quality load analysis layer, biochemical reaction mechanism layer, and aeration demand calculation layer. The core module includes an influent water quality parameter module, a microbial nitrification / denitrification kinetics module, a dissolved oxygen (DO) mass transfer module, an aeration energy consumption coupling module, and a sludge concentration (MLSS) control module. The modules are connected in a unidirectional series and a bidirectional feedback, with the output of the preceding module serving as the input of the following module. The aeration feedback data corrects the kinetic module parameters in reverse, forming a closed loop.The key steps and parameters for constructing and training the mechanistic model are based on the core mechanism of activated sludge ASM2d, solidifying the core formulas for nitrifying / heterotrophic bacteria proliferation, substrate degradation, and oxygen mass transfer; calibrating the core process parameters as follows: sludge age 15-20 days, target DO value 2.0-3.0 mg / L, sludge concentration 3000-4000 mg / L, and oxygen utilization rate 0.25-0.35; inputting historical influent COD, NH3-N, water temperature, and flow rate data, substituting them into the mechanistic formula to iteratively solve for the theoretical value of aeration oxygen demand; combining actual aeration operation data to correct model deviations, calibrating kinetic coefficients until the error between theoretical and measured values ​​is ≤8%; solidifying the mechanistic model parameters, completing the model finalization, and serving as the constraint benchmark for the AI ​​algorithm. Key steps include data preprocessing, i.e., collecting influent / DO related time-series data, cleaning, normalizing, and time-series normalizing, and dividing the dataset, i.e., the LSTM needs to be reconstructed into a 3D input format. LSTM modeling and prediction involves building and training an LSTM network, tuning parameters, and outputting predicted future influent load values. MPC modeling and solving involves defining the prediction model, optimization objectives for achieving DO (discharge) targets and energy efficiency, and process constraints, transforming it into a planning problem to infer DO demand. Rolling execution and correction involve real-time data acquisition to update the model, continuously optimizing and executing only the current control variables, and using actual monitoring values ​​for feedback correction. System verification involves offline simulation to verify performance, with optional online deployment. The core connection is that the accuracy of LSTM prediction directly affects MPC performance; the combination of both ensures both accurate DO control and energy efficiency optimization. The control layer consists of PID control, fuzzy control, and reinforcement learning (DDPG). The core coordination logic includes hierarchical division of labor: DDPG performs high-level optimization, fuzzy control performs intermediate corrections, and PID performs low-level execution, adapting to the core requirements of precise aeration for energy saving and stable DO. The collaborative mechanism involves DDPG learning the optimal dynamic DO setpoint based on influent load and water quality targets; fuzzy control corrects DDPG output in real time to handle load changes, nonlinear disturbances, or PID parameters; and PID provides rapid response, precisely adjusting the opening of aeration fans / valvees to maintain stable DO. The implementation steps include: data preparation (collecting time-series data such as influent load, DO concentration, aeration energy consumption, and effluent water quality) and defining constraints such as equipment range and DO control interval. Modular modeling includes initial PID tuning of P / I / D parameters and setting basic DO control targets. Fuzzy control involves constructing a fuzzy rule base for DO deviation and load deviation to correct PID parameters, followed by offline calibration. DDPG (Defining reward functions for water quality compliance and minimum energy consumption) is used to train the agent to output dynamic DO setpoints. Module integration involves the DDPG output to fuzzy control for dynamic correction, and then to the PID receiving the target value to control the aeration actuator. Online deployment involves real-time data acquisition, rolling DDPG strategy updates, dynamic adaptation of fuzzy control to operating conditions, and millisecond-level PID response. Iterative optimization involves fine-tuning the fuzzy rules, DDPG reward function, and PID initial parameters based on actual energy consumption and DO stability data.The sensing layer collects data from multiple sensors, including dissolved oxygen (DO), ammonia nitrogen, flow rate, and sludge concentration. The feature extraction stage mainly extracts key features reflecting the state of the aeration process from these monitoring data, such as the rate of change of dissolved oxygen (DO); ammonia nitrogen removal rate; sludge settling characteristics (SVI); and aeration energy consumption indicators.

[0066] The core dimensions and objectives of multi-objective collaborative optimization are defined. Multi-objective collaborative optimization must simultaneously consider three core dimensions: water quality compliance, hydraulic stability, and equipment efficiency. The optimization objectives for each dimension are clear and complementary. The water quality dimension includes achieving standards for effluent ammonia nitrogen, total nitrogen, and COD as the core constraint, minimizing fluctuations. The hydraulic dimension includes uniform flow within the reactor, balanced spatial distribution of dissolved oxygen (DO), and stable sludge retention time (SRT). The equipment dimension includes minimizing aeration fan energy consumption, reducing equipment start-up and shutdown frequency, and extending the lifespan of vulnerable components such as fans and valves. The synergistic logic of neural networks and optimization algorithms forms a closed loop of prediction, optimization, and control. Their roles are clearly defined and they work synergistically: the neural network is responsible for accurate prediction, training on historical data to predict water quality changes, hydraulic states, and equipment energy consumption under different parameter combinations, providing a reliable predictive model for the optimization algorithm. Examples include LSTM predicting DO demand and CNN identifying sludge settling states. The optimization algorithm is responsible for global optimization, using the neural network's predictions as input and searching for the optimal combination of water quality, hydraulic, and equipment parameters under water quality compliance constraints, outputting optimal aeration rates, blower frequencies, and other control commands. Adaptable multi-objective optimization algorithms are needed, balancing global search capabilities with real-time performance, to adapt to the dynamic fluctuations in wastewater treatment scenarios. Mainstream multi-objective optimization algorithms include Non-Dominated Sorting Genetic Algorithm (NSGA-III) and Multi-Objective Particle Swarm Optimization (MOPSO), which excel at handling conflicting objectives, such as minimum energy consumption and optimal water quality. The collaborative optimization architecture comprises a three-layer structure employing a neural network prediction model, a multi-objective optimization algorithm, and PID / MPC control. The optimal parameters output by the optimization algorithm serve as the setpoints for the control layer, enabling dynamic adjustment. Lightweight adaptation involves engineering simplification of the optimization algorithm, such as reducing the number of iterations and dynamically allocating objective weights to ensure real-time response, for example, a control cycle of ≤5 minutes to meet the real-time requirements of the aeration system. The core advantage of multi-objective collaborative optimization is overcoming the limitations of single-objective optimization, avoiding the contradiction between excessive aeration for water quality and resulting in soaring energy consumption or sacrificing effluent compliance rate for energy saving. It enhances system robustness, adapting to dynamic fluctuations in influent load and ambient temperature; hydraulic parameter optimization reduces uneven DO distribution. It lowers overall costs while ensuring stable water quality compliance. It strengthens control precision, with the optimized control strategy improving DO control precision to ±0.1 mg / L.Key implementation recommendations include: clearly defining parameter constraint boundaries, i.e., first defining the feasible domain of each dimension of parameters, such as the upper limit of aeration volume and the DO control range, to avoid optimization results exceeding engineering realities; dynamically allocating target weights, i.e., dynamically adjusting the weights of water quality, hydraulic, and equipment targets based on influent load and water quality fluctuations, for example, prioritizing water quality weights when influent concentration rises sharply; incorporating mechanistic model constraints, i.e., integrating wastewater treatment mechanisms, such as activated sludge reaction kinetics, into the optimization process to avoid optimization results deviating from engineering realities; and using a particle swarm optimization algorithm: with aeration volume and dissolved oxygen setpoints as optimization variables, wastewater aeration process water quality compliance thresholds, dissolved oxygen 2.0~3.0 mg / L, and sludge concentration as constraint boundaries, and minimum aeration energy consumption as the fitness function, iteratively outputting optimal aeration parameters to adapt to the dynamic control of the aerobic tank biochemical reaction load. Genetic Algorithm: Aeration volume and frequency are encoded into gene sequences. With influent water quality load fluctuations as constraints and water quality compliance and energy conservation as adaptation goals, the optimal aeration scheme is selected through selection, crossover, and mutation iterations to match the nonlinear biological aeration requirements of activated sludge. Intelligent Optimization Algorithm: The core function of this algorithm is to solve complex industrial problems, adapting to nonlinear, multi-constraint, and multi-objective scenarios, such as process parameter optimization and energy consumption and water quality balance, overcoming the limitations of traditional methods. It replaces manual experience, automatically finding the optimal solution, reducing reliance on experts, and improving the scientific nature of decision-making. It achieves quantifiable gains, optimizing core indicators such as efficiency, cost, and quality, including reduced energy consumption, increased production capacity, and improved compliance rates. The algorithm implementation steps include modeling preparation, i.e., clarifying the optimization objective; constraints, i.e., equipment range and water quality standards; input and output data; algorithm construction, i.e., initializing parameters and building core logic, such as fuzzy rule base and sliding window strategy; training or optimization: training the model or iteratively optimizing using historical or real-time data, such as genetic algorithm selection, crossover, and mutation; offline simulation to verify the effect, and rolling iteration after online deployment; comparing the indicators before and after optimization to confirm the value of the algorithm and achieve effect closure.

[0067] This invention constructs a wastewater biochemical treatment parameter system through specific steps, including clarifying the system objectives. This involves covering all stages of the biochemical reaction, such as organic matter degradation, nitrification or denitrification, and supporting model prediction and control. The parameters must reflect four dimensions: water quality, hydraulics, equipment, and process status. Core parameter selection includes: water quality parameters such as DO, ORP, ammonia nitrogen / nitrate nitrogen / nitrite nitrogen, and pH (i.e., a water quality sensor set); hydraulic parameters such as influent flow rate, return sludge flow rate, and aeration tank level (i.e., a hydraulic sensor set); equipment parameters such as blower frequency, aerator differential pressure, and blower outlet pressure (i.e., aeration execution module); and process status parameters such as MLSS (sludge concentration) and sludge age (i.e., derived parameters calculated based on sludge discharge). Parameter attributes are defined, specifying the acquisition location for each parameter, such as DO in the three-dimensional region of the aeration tank, frequency (DO 1s / time), accuracy (DO ± 0.1 mg / L), and process correlations, such as ammonia nitrogen correlated with nitrification progress. Validation and simplification involve testing with historical data and eliminating redundant parameters, such as those with a correlation of less than 0.3 with DO, to ensure that parameters accurately reflect changes in biochemical state. For example, ORP can predict DO trends in advance. The connection between the parameter system and the wastewater biochemical treatment model includes the input basis of LSTM prediction and the training and inference of fuzzy / PID control, which relies entirely on the system's real-time / historical data. For example, DO and ammonia nitrogen data can be used to train a DO prediction sub-model. Accuracy support is provided by the completeness of the parameter system. For example, parameters across the entire nitrogen-containing chain determine the model's adaptability to processes such as denitrification. Parameter accuracy, for example, DO ±0.1 mg / L directly affects the model's prediction deviation ≤0.2 mg / L. Collaborative iteration involves the model using deviation data fed back from the parameter system. For example, if DO exceeds the tolerance, the model optimizes its own parameters, such as adjusting LSTM weights. Simultaneously, it guides the parameter system to supplement key parameters, such as adding sludge age to improve sludge load prediction accuracy. Parameter preprocessing involves collecting water quality (DO, ammonia nitrogen, etc.), hydraulic (flow rate, liquid level), and equipment (fan frequency, etc.) parameters, removing outliers beyond their range, and standardizing them to the [0,1] range to eliminate dimensional interference. Feature fusion is used to generate associated features based on process mechanisms, such as DO-ORP synergistic values, ammonia nitrogen / nitrate nitrogen ratios, and flow rate / liquid level coupling values, concentrating core information from multiple parameters. Operating condition matching and identification uses fuzzy logic to match 125 rules or K-means clustering, comparing the fused features with preset operating conditions, such as nitrification dominance or sudden water quality changes, to determine the current operating condition. Dynamic weight allocation uses neural networks, such as LSTM, to adjust parameter weights according to operating conditions; for example, during nitrification, the ammonia nitrogen weight is increased to 40% and the DO weight to 30%, highlighting key influencing factors. Output analysis results provide operating condition judgment conclusions, such as mild nitrification insufficiency, or generate control guidelines, such as requiring a 10% increase in aeration, to support subsequent strategy formulation. The intelligent collaborative control technology for aeration equipment refers to an aeration execution system that combines a variable frequency Roots blower with a microporous aerator group, and achieves intelligent collaborative control.Variable frequency Roots blowers provide a wide range of aeration airflow adjustment to meet the macroscopic changes in aeration volume required by the entire aeration tank. Microporous aerator groups, on the other hand, target specific areas within the aeration tank, independently and precisely controlling the air intake via solenoid valves based on real-time dissolved oxygen demand differences, achieving micro-zone aeration intensity adjustment. Working together, these two systems ensure uniform aeration while maximizing the flexibility and energy efficiency of the aeration system, reducing dissolved oxygen fluctuations from ±1.0 mg / L to ±0.2-0.3 mg / L compared to traditional aeration systems. A smart decision-making system based on big data and machine learning utilizes big data analytics to mine and analyze massive amounts of historical operational data, continuously optimizing the mathematical model and control strategies of the wastewater biochemical treatment process using machine learning algorithms. This smart decision-making system automatically adapts to changes in wastewater quality and quantity, as well as treatment needs under different seasons and operating conditions, predicting dissolved oxygen trends in advance and making precise aeration adjustment decisions. This significantly enhances the system's adaptability and stability in complex operating conditions, reduces manual intervention, and achieves intelligent and refined management, providing a new intelligent control solution for the wastewater biochemical treatment industry. The integrated remote monitoring and intelligent operation and maintenance function can integrate wired and wireless communication technologies to build a complete monitoring system that combines local and remote monitoring, enabling comprehensive, 24 / 7 real-time monitoring of the aeration system. Simultaneously, the intelligent operation and maintenance function based on the remote monitoring platform allows for remote guidance on equipment maintenance, timely detection and resolution of potential equipment problems, and improved system reliability and continuity.

[0068] The advantages of this invention lie in avoiding the over-aeration and under-aeration problems of traditional aeration methods, ensuring the aeration system always operates at optimal energy consumption, significantly reducing the operating costs of wastewater treatment plants and providing excellent benefits. It improves treatment efficiency and water quality by precisely adjusting dissolved oxygen levels, providing a suitable living environment for microorganisms, promoting efficient biochemical reactions such as organic matter degradation and biological nitrogen and phosphorus removal, thereby increasing wastewater treatment efficiency and making the effluent quality more stable and reliable, with major pollutant indicators meeting increasingly stringent wastewater treatment discharge standards. Stable and reliable operation is ensured through multiple safeguards, including multi-parameter monitoring, intelligent control algorithms, intelligent collaborative equipment regulation, and remote monitoring and maintenance, guaranteeing the system's strong anti-interference and self-adaptive capabilities. This allows it to reliably cope with fluctuations in wastewater quality and quantity, equipment failures, and other emergencies, reducing system downtime and improving the reliability of the aeration system and the operational stability of the wastewater treatment plant. Intelligent management is achieved through the collaborative application of a remote monitoring platform and local monitoring terminals, enabling intelligent management of the wastewater treatment aeration process. Operators can monitor the system's operation anytime, anywhere, and perform remote operation and management, reducing labor costs.

[0069] In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as follows: Figure 3As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores static and dynamic information data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements the steps in the above-described embodiment of the intelligent and precise aeration control method for the wastewater biochemical treatment process.

[0070] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present invention and does not constitute a limitation on the computer device to which the present invention is applied. A specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0071] In addition, the present invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described embodiment of the intelligent and precise aeration control method for the wastewater biochemical treatment process.

[0072] In addition, the present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the above-described embodiment of the intelligent and precise aeration control method for the wastewater biochemical treatment process.

[0073] Those skilled in the art will understand that implementing all or part of the processes in the intelligent and precise aeration control method for wastewater biochemical treatment processes described in the above embodiments can be accomplished by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the intelligent and precise aeration control method for wastewater biochemical treatment processes described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0074] This invention is not limited to the structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this invention is limited only by the appended claims.

Claims

1. A smart and precise aeration control method for wastewater biochemical treatment processes, characterized in that, include: Acquire aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform operating condition matching based on the standard aeration monitoring data, and obtain operating condition identification results. Based on the operating condition identification results, the aeration coordinated control prediction is performed on the standard aeration monitoring data to obtain the aeration adjustment amount. The deviation of the aeration adjustment amount is calculated, and the aeration adjustment amount is optimized based on the deviation calculation results to obtain the optimized aeration adjustment amount. The optimal aeration control strategy is generated by combining the optimized aeration volume adjustment with multi-objective constraints, and intelligent and precise aeration control of the wastewater biochemical treatment process is realized based on the optimal strategy.

2. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 1, characterized in that, The process of acquiring aeration monitoring data, preprocessing the aeration monitoring data to obtain standard aeration monitoring data, and performing operating condition matching based on the standard aeration monitoring data to obtain operating condition identification results includes: Acquire aeration monitoring data during the wastewater biochemical treatment process, process the missing and outlier values ​​of the aeration monitoring data to obtain the aeration monitoring data after cleaning, and perform data format standardization conversion on the cleaned aeration monitoring data to obtain standard aeration monitoring data. Based on the process mechanism, the standard aeration monitoring data are fused to generate associated features. Fuzzy logic is then used to match and identify the associated features under operating conditions to obtain the operating condition identification results.

3. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 2, characterized in that, The aeration monitoring data includes water quality data, hydraulic data, and aeration equipment operation data; The water quality data includes: dissolved oxygen data, oxidation-reduction potential data, ammonia nitrogen, nitrite nitrogen and nitrate nitrogen data, and pH data; The hydraulic data includes: liquid level data and flow rate data; The aeration equipment operation data includes: variable frequency Roots blower operation data and microporous aerator group operation data.

4. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 2, characterized in that, The feature fusion includes: calculating the synergistic value of dissolved oxygen and redox potential, the ratio of ammonia nitrogen to nitrate nitrogen concentration, and the coupling value of flow rate and liquid level.

5. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 1, characterized in that, The process involves predicting aeration coordination control based on standard aeration monitoring data according to the operating condition identification results, obtaining the aeration adjustment amount, calculating the deviation of the aeration adjustment amount, and optimizing the aeration adjustment amount based on the deviation calculation results. The optimized aeration adjustment amount includes: The initial aeration coordination control weights are determined based on the operating condition identification results. Based on the initial aeration coordination control weights and combined with the previously acquired historical aeration monitoring data, the aeration coordination control prediction is performed on the standard aeration monitoring data to obtain the aeration adjustment amount. The deviation between the aeration adjustment amount and the preset target adjustment amount is calculated, and the deviation calculation result is judged according to the preset deviation grading rules to obtain the deviation grading result. Based on the deviation level judgment result, the control path is matched, and the aeration adjustment amount is compensated and corrected according to the deviation level according to the control path and the deviation calculation result to obtain the optimized aeration adjustment amount.

6. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 5, characterized in that, The aeration adjustment amount is obtained by predicting the aeration coordination control based on the initial aeration coordination control weights and combining them with pre-acquired historical aeration monitoring data to obtain standard aeration monitoring data, including: The initial parameters and basic dissolved oxygen control target were set using proportional, integral and derivative control based on standard aeration monitoring data to obtain the basic aeration adjustment amount. Based on fuzzy reasoning, the basic aeration adjustment amount is parameterized to obtain the fuzzified aeration adjustment amount. Then, the fuzzified aeration adjustment amount is subjected to rule matching and defuzzification to obtain the preliminary aeration adjustment amount. A neural network prediction model is constructed based on the pre-acquired historical aeration monitoring data and the initial aeration collaborative control weights. The neural network prediction model is then trained and optimized to obtain a multi-dimensional correlation model. This multi-dimensional correlation model is then used to optimize the initial aeration volume adjustment to obtain the aeration adjustment amount.

7. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 6, characterized in that, The process involves constructing a neural network prediction model based on pre-acquired historical aeration monitoring data and initial aeration collaborative control weights, training and optimizing the neural network prediction model to obtain a multi-dimensional correlation model, and using this multi-dimensional correlation model to optimize the initial aeration volume adjustment, resulting in the following aeration adjustment amounts: The prediction target and boundary are determined based on the initial aeration volume adjustment, and a time series feature set is constructed based on the pre-acquired historical aeration monitoring data. The parameters of the neural network prediction model are initialized through the time series feature set and the initial aeration collaborative control weights to obtain the long short-term memory network model. The time-series feature set is divided into a training set and a validation set. The preset minimum effluent quality and aeration energy consumption are used as the dual loss functions. The long short-term memory network model is iteratively trained based on the gradient descent method to obtain a multi-dimensional correlation model. The initial aeration adjustment amount is input into a multi-dimensional correlation model, which is then passed through a long short-term memory network layer and a full-chain layer in sequence, combined with a gating unit to capture temporal dependencies, and optimized to obtain the aeration adjustment amount.

8. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 5, characterized in that, The matching rules for the control path based on the deviation level judgment result include: If the deviation level judgment result is a small deviation, then match the proportional, integral and derivative control paths; If the deviation level is determined to be a slight deviation, then the control path is corrected by matching fuzzy rules. If the deviation level assessment result is a severe deviation, then a multi-dimensional correlation model control path is matched.

9. The intelligent and precise aeration control method for wastewater biochemical treatment process according to claim 1, characterized in that, The process of combining optimized aeration rate adjustments with multi-objective constraints to generate an optimal aeration control strategy, and then implementing intelligent and precise aeration control for the wastewater biochemical treatment process based on this optimal strategy, includes: The core constraint is the water quality indicator compliance requirements, and multi-objective constraint conditions are constructed by combining the hydraulic spatial distribution requirements and the equipment energy consumption requirements. Based on multi-objective constraints, and using a multi-objective optimization algorithm to optimize the aeration adjustment amount, the optimal aeration control strategy is obtained. The optimal aeration control strategy is converted into aeration control commands and sent to the programmable logic controller via an industrial communication protocol to drive the aeration equipment in the wastewater biochemical treatment process.

10. A smart and precise aeration control system for a wastewater biochemical treatment process, characterized in that, The intelligent and precise aeration control system for the wastewater biochemical treatment process includes: The operating condition matching module is used to acquire aeration monitoring data, preprocess the aeration monitoring data to obtain standard aeration monitoring data, perform operating condition matching based on the standard aeration monitoring data, and obtain operating condition identification results. The aeration adjustment calculation module is used to predict the coordinated control of aeration based on the standard aeration monitoring data according to the working condition identification results, obtain the aeration adjustment amount, calculate the deviation of the aeration adjustment amount, and optimize the aeration adjustment amount based on the deviation calculation results to obtain the optimized aeration adjustment amount. The aeration control strategy generation module is used to combine the optimized aeration volume adjustment with multi-objective constraints to generate the optimal aeration control strategy, and realize intelligent and precise aeration control in the wastewater biochemical treatment process based on the optimal strategy.

Images