A method, system, equipment and medium for optimizing and controlling water treatment processes

Abstract

This application relates to the field of automatic control technology, and discloses a method, system, equipment, and medium for optimizing control of a water treatment process. The method includes: determining the state of sensor-detected data using a multi-parameter state recognition algorithm to obtain a treatment stage switching command; modeling the change pattern of microbial activity based on the switching command to obtain a staged prediction model; optimizing the prediction model using a switching model predictive control algorithm to obtain a multi-objective control strategy; coordinating the calculation of setpoints based on the control strategy to obtain distributed control commands; and using the control commands to drive and control the equipment via a fieldbus to obtain a coordinated operating state. This application improves the control accuracy and multi-objective coordinated optimization capability of the water treatment process.

Description

Technical Field

[0001] This application relates to the field of automatic control technology, and in particular to a method, system, equipment and medium for optimizing control of water treatment processes. Background Technology

[0002] Existing water treatment process control methods mainly employ traditional PID control or simple automated control systems. These systems monitor single or a few water quality parameters, such as dissolved oxygen and pH, to achieve basic control of process parameters like aeration and recirculation. These control methods are typically based on empirical parameter settings and fixed control strategies, maintaining basic treatment efficiency under stable operating conditions, and have been widely used in wastewater treatment plants.

[0003] However, traditional control methods cannot accurately identify the transition timing of different reaction stages such as anaerobic, anoxic, and aerobic in the biochemical treatment process, resulting in a mismatch between the control strategy and the actual process stage. Secondly, a single control model is difficult to adapt to the differences in the changes in microbial activity at each treatment stage, resulting in insufficient control precision and response lag. Furthermore, existing methods lack a multi-objective coordinated optimization mechanism, and cannot simultaneously take into account multiple objectives such as water quality compliance, energy conservation and consumption reduction, and stable operation.

[0004] How to construct an intelligent control method that can accurately identify the switching of processing stages, adaptively adjust the control model, and achieve multi-objective coordinated optimization requires the control system to have the ability to identify the state by multi-parameter fusion, to establish corresponding predictive models based on the biochemical reaction characteristics of different stages, and to achieve dynamic switching and coordinated optimization of control strategies through advanced optimization algorithms, thereby overcoming the technical bottleneck of poor control performance of existing technologies under complex working conditions. Summary of the Invention

[0005] This application provides a method, system, equipment, and medium for optimizing the control of a water treatment process. It addresses the problem of existing technologies being unable to accurately identify treatment stage transitions and adaptively adjust control strategies by constructing a water treatment optimization method based on multi-parameter state recognition and switching model predictive control. This improves the control accuracy and multi-objective coordinated optimization capabilities of the water treatment process.

[0006] Firstly, this application provides a method for optimizing and controlling a water treatment process. This method includes: processing the detection data from dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors using a multi-parameter state recognition algorithm to obtain a treatment stage switching instruction; this includes: deploying dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors in anaerobic, anoxic, and aerobic tanks respectively; collecting water quality parameters of each treatment unit in real time to obtain a multi-parameter detection data sequence; performing ternary parameter fusion calculation on the multi-parameter detection data sequence; and comparing the dissolved oxygen concentration value with preset low-oxygen and high-oxygen thresholds. The dissolved oxygen state level is determined by comparison. Based on the dissolved oxygen state level, the oxidation-reduction potential (ORP) value is segmented and matched with negative potential threshold, zero potential threshold, and positive potential threshold to obtain the ORP state level. The pH change rate is calculated based on the ORP state level, and the difference between the pH change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator. The water quality gradient change indicator is verified temporally with the parameter change trend of multiple consecutive sampling periods to identify and confirm the transition timing of the anaerobic reaction stage, the anoxic reaction stage, and the aerobic reaction stage, and obtain the treatment stage switching instruction.

[0007] Optionally, the step of modeling the changes in microbial activity at each biochemical reaction stage according to the processing stage switching instruction to obtain a staged prediction model includes:

[0008] Based on the processing stage switching command, the phosphorus release kinetic parameters of the anaerobic reaction stage are identified and calculated, and the correlation analysis between microbial activity indicators and phosphorus release rate is performed to obtain the dynamic response parameters of the anaerobic stage.

[0009] The denitrification process parameters in the anoxic reaction stage are modeled and calculated according to the processing stage switching command. The denitrifying bacteria activity and total nitrogen removal rate are coupled and analyzed to obtain the dynamic response parameters in the anoxic stage.

[0010] The processing stage switching command is correlated with the nitrification and phosphorus uptake process in the aerobic reaction stage, and the activity changes of nitrifying bacteria and polyphosphate-accumulating bacteria are synchronously modeled to obtain the dynamic response parameters of the aerobic stage.

[0011] Based on the dynamic response parameters of the anaerobic stage, the anoxic stage, and the aerobic stage, the transfer characteristics are analyzed, and the input-output response relationship of each stage is converted into a predictive control structure to obtain stage transfer characteristic data.

[0012] The stage-transfer characteristic data are combined into models according to the time sequence relationship of anaerobic-anoxic-aerobic, and the switching conditions and boundary constraints of each stage model are set and processed to obtain a stage-based prediction model.

[0013] Optionally, the step of performing rolling optimization on the staged prediction model through a switching model predictive control algorithm to obtain a multi-objective control strategy includes:

[0014] The phased prediction model is input into the switching model prediction controller to set the prediction time domain, and the water quality change trend of multiple future control cycles is predicted and calculated to obtain the prediction time domain data.

[0015] Based on the predicted time-domain data, the target values ​​of chemical oxygen demand, total nitrogen concentration and total phosphorus concentration in the effluent are constrained, and the water quality compliance requirements are converted into optimization objective functions to obtain multi-objective constraint conditions.

[0016] Based on the multi-objective constraints, cost weights are allocated for aeration energy consumption, reagent consumption, and sludge production. Water quality control objectives and energy conservation and consumption reduction objectives are weighted and combined to obtain a comprehensive optimization objective function.

[0017] The comprehensive optimization objective function is solved using a rolling optimization algorithm, and the optimal control sequence in the control time domain is iteratively searched to obtain the rolling optimized control sequence.

[0018] Based on the rolling optimization control sequence, the control strategies for the anaerobic, hypoxic, and aerobic stages are extracted in segments. The optimal control parameters for each stage are combined according to the temporal relationship to obtain a multi-objective control strategy.

[0019] Optionally, the step of inputting the phased prediction model into the switching model prediction controller for prediction time domain setting, and performing prediction calculations on the water quality change trend for multiple future control cycles to obtain prediction time domain data includes:

[0020] The dynamic response parameters of the anaerobic stage, the hypoxic stage, and the aerobic stage in the phased prediction model are input into the switching model prediction controller to initialize the model parameters of each stage and obtain the initial state data of the controller.

[0021] Based on the controller's initial state data, time windows are set for the prediction time domain length and the control time domain length, dividing the time range of prediction control into multiple continuous control cycles to obtain time domain window configuration parameters.

[0022] The initial state values ​​of the water quality state variables at the current moment are set according to the time domain window configuration parameters. The current values ​​of dissolved oxygen concentration, oxidation-reduction potential and pH are used as the prediction starting point to obtain the prediction initial state vector.

[0023] The predicted initial state vector and the phased prediction model are recursively calculated to gradually predict the changes in water quality parameters in each future control cycle, thus obtaining the multi-cycle water quality change trajectory.

[0024] Based on the multi-cycle water quality change trajectory, the data on the changing trends of chemical oxygen demand removal rate, total nitrogen removal rate, and total phosphorus removal rate are integrated, and the predicted sequences of each water quality indicator are combined and arranged in chronological order to obtain the predicted time domain data.

[0025] Optionally, the step of coordinating and calculating the aeration rate setpoint, reflux ratio setpoint, and stirring intensity setpoint according to the multi-objective control strategy to obtain distributed control commands includes:

[0026] The control parameters for the anaerobic stage, the anoxic stage, and the aerobic stage in the multi-objective control strategy are separated into stage parameters, and the optimal control parameters for each stage are categorized and extracted to obtain a set of stage control parameters.

[0027] Based on the set of phased control parameters, the aeration set value of the aerobic tank is calculated and processed. The oxygen demand of nitrification reaction and the oxygen demand of phosphorus uptake by polyphosphate-accumulating bacteria are superimposed to obtain the optimized aeration set value.

[0028] Based on the set of phased control parameters, the set values ​​of the recirculation ratio in the anoxic tank and the recirculation ratio outside the aerobic tank are coupled and calculated. The denitrification efficiency requirement and the sludge concentration control requirement are balanced to obtain the optimized recirculation ratio set value.

[0029] The set of staged control parameters is correlated with the stirring intensity of the anaerobic and anoxic tanks, and the phosphorus release mixing intensity and denitrification mixing intensity are optimized respectively to obtain the optimized stirring intensity set value.

[0030] Based on the optimized aeration rate setting, optimized reflux ratio setting, and optimized stirring intensity setting, control instructions are encapsulated, and each control parameter is grouped and configured according to the execution device address to obtain distributed control instructions.

[0031] Optionally, the step of using the distributed control commands to drive and control the blower, return pump, and agitator via a fieldbus communication protocol to obtain the coordinated operating status of the water treatment equipment includes:

[0032] The distributed control commands are classified and parsed according to the blower control command, the reflux pump control command, and the agitator control command. The control parameters of each device are converted into data formats to obtain device-specific control data packages.

[0033] Based on the device-specific control data packet, the data frame structure of the fieldbus communication protocol is encapsulated, and the control command data and device address information are assembled according to the protocol format to obtain the bus communication data frame.

[0034] The frequency converter of the aerobic tank blower is driven by the bus communication data frame, and the optimized aeration volume setpoint is converted into a blower speed control signal to obtain the blower operation control status.

[0035] The bus communication data frame is connected to the frequency converter controller of the internal and external reflux pumps to perform pump speed adjustment control on the optimized reflux ratio set value, thereby obtaining the reflux pump operation control status.

[0036] Based on the bus communication data frames, instructions are transmitted to the motor controllers of the agitators in the anaerobic and anoxic tanks. The optimized stirring intensity setting value is converted into a motor power control signal, and synchronized with the operation control status of the blower and the operation control status of the return pump to obtain the coordinated operation status of the water treatment equipment.

[0037] Secondly, this application provides a water treatment process optimization control system, which includes:

[0038] The discrimination module is used to process the detection data from the dissolved oxygen sensor, oxidation-reduction potential sensor, and pH sensor using a multi-parameter state recognition algorithm to obtain processing stage switching instructions. This includes: deploying dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors in the anaerobic, anoxic, and aerobic tanks respectively; collecting water quality parameters from each treatment unit in real time to obtain a multi-parameter detection data sequence; performing ternary parameter fusion calculations on the multi-parameter detection data sequence; comparing the dissolved oxygen concentration value with preset low-oxygen and high-oxygen thresholds to determine the dissolved oxygen state level; and based on the... The dissolved oxygen status level is used to segment and identify the redox potential value. The potential value is matched with negative potential threshold, zero potential threshold, and positive potential threshold to obtain the redox potential status level. The acid-base change rate is calculated according to the redox potential status level. The difference between the acid-base change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator. The water quality gradient change indicator is verified with the parameter change trend of multiple consecutive sampling periods to identify and confirm the transition time of anaerobic reaction stage, anoxic reaction stage, and aerobic reaction stage, and obtain the treatment stage switching instruction.

[0039] The modeling module is used to model the changes in microbial activity at each biochemical reaction stage according to the processing stage switching instructions, and obtain a staged prediction model.

[0040] The optimization module is used to perform rolling optimization on the staged prediction model by switching model prediction control algorithms to obtain a multi-objective control strategy.

[0041] The control module is used to coordinate and calculate the aeration volume setpoint, reflux ratio setpoint, and stirring intensity setpoint according to the multi-objective control strategy to obtain distributed control commands.

[0042] The drive module is used to process the distributed control commands to drive and control the blower, return pump, and agitator through the fieldbus communication protocol, so as to obtain the coordinated operation status of the water treatment equipment.

[0043] Thirdly, a water treatment process optimization control device is provided, comprising: a memory and at least one processor, wherein the memory stores instructions; the at least one processor invokes the instructions in the memory to cause the water treatment process optimization control device to execute the above-described water treatment process optimization control method.

[0044] Fourthly, a computer-readable storage medium is provided, wherein instructions are stored therein, which, when executed on a computer, cause the computer to perform the aforementioned water treatment process optimization control method.

[0045] The technical solution provided in this application uses a multi-parameter state recognition algorithm to process the detection data from dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors, accurately identifying the transition times between different biochemical reaction stages such as anaerobic, anoxic, and aerobic processes during water treatment. This avoids the limitations of traditional methods that rely solely on a single parameter or time point for stage judgment. The method ensures the accuracy and timeliness of stage switching commands through ternary parameter fusion calculation and gradient change identification. Based on the treatment stage switching commands, the method models the changes in microbial activity at each biochemical reaction stage, establishing a stage-specific prediction model tailored to the characteristics of different reaction stages. This overcomes the problem that traditional unified models cannot adapt to the differentiated control needs of each stage. Through rolling optimization processing using a switching model predictive control algorithm, predictive control is combined with multi-objective optimization, achieving a coordinated balance between multiple objectives such as water quality compliance and energy conservation, solving the problem of poor overall performance caused by single-objective control in existing technologies. Coordinated calculations are performed on the aeration rate setpoint, reflux ratio setpoint, and stirring intensity setpoint. A distributed control architecture ensures the synchronous and coordinated operation of all actuators, avoiding the risk of single-point failures that may occur with centralized control.

[0046] In particular, the switching model predictive control algorithm fully considers the stage characteristics of the biochemical reaction process and the changing patterns of microbial activity. By establishing specialized predictive models for phosphorus release kinetics in the anaerobic stage, denitrification in the anoxic stage, and nitrification and phosphorus uptake in the aerobic stage, the control algorithm can accurately match the process requirements of different treatment stages. The multi-parameter state recognition algorithm is optimized for the correlation and temporal characteristics of key parameters such as dissolved oxygen, oxidation-reduction potential, and pH value in water treatment processes. Through ternary parameter fusion and gradient change detection, the accuracy and robustness of stage recognition are ensured. The distributed control command generation and execution mechanism fully considers the dynamic response characteristics and coordination requirements of equipment such as blowers, return pumps, and agitators. Real-time transmission of control commands and synchronous monitoring of equipment status are achieved through fieldbus communication protocols. Attached Figure Description

[0047] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0048] Figure 1 This is a schematic diagram of an embodiment of the water treatment process optimization control method of the present invention;

[0049] Figure 2 This is a schematic diagram of one embodiment of the water treatment process optimization control system of the present invention;

[0050] Figure 3 This is a schematic block diagram of the structure of the water treatment process optimization control equipment in an embodiment of the present invention. Detailed Implementation

[0051] This application provides a method, system, device, and medium for optimizing and controlling a water treatment process. The terms "first," "second," "third," "fourth," etc. (if present)," in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0052] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the water treatment process optimization control method in this application includes:

[0053] Step S101: The detection data of the dissolved oxygen sensor, redox potential sensor and pH sensor are processed by a multi-parameter state recognition algorithm to obtain the processing stage switching instruction.

[0054] Step S102: Model the microbial activity change pattern of each biochemical reaction stage according to the processing stage switching instruction to obtain a staged prediction model;

[0055] Step S103: The phased prediction model is subjected to rolling optimization processing through a switching model prediction control algorithm to obtain a multi-objective control strategy;

[0056] Step S104: Based on the multi-objective control strategy, coordinate the aeration rate setpoint, reflux ratio setpoint, and stirring intensity setpoint to obtain distributed control commands.

[0057] Step S105: The distributed control commands are used to drive and control the blower, reflux pump, and agitator through the fieldbus communication protocol to obtain the coordinated operation status of the water treatment equipment.

[0058] It is understood that the executing entity of this application can be a water treatment process optimization and control device, or it can be a terminal or a server; the specific implementation is not limited here. This application's embodiments use a server as an example for illustration.

[0059] Specifically, a multi-parameter state recognition algorithm is used to accurately identify different biochemical reaction stages in water treatment. This algorithm first deploys dissolved oxygen, oxidation-reduction potential (ORP), and pH sensors in anaerobic, anoxic, and aerobic tanks to collect water quality parameter data in real time, forming a multi-parameter detection data sequence. Its core technology, the ternary parameter fusion calculation, is as follows: First, the dissolved oxygen concentration value is compared with high and low oxygen thresholds, marking it as low, medium, or high oxygen states to obtain a dissolved oxygen state level. Based on this, the ORP value is segmented and matched with negative, zero, and positive potential threshold ranges to obtain an ORP state level. Then, based on this level, the pH change rate gradient is calculated. The difference between the current and previous pH values ​​is divided by the time interval to obtain the change rate, which is compared with a preset threshold. If the threshold is exceeded, a water quality gradient change indicator is generated. In the time-series verification stage, this identifier is compared and analyzed with the parameter change trends of multiple consecutive sampling periods. If the dissolved oxygen concentration continues to decrease and the redox potential tends to be negative within three consecutive sampling periods, it is determined to be the anaerobic reaction stage; if the dissolved oxygen concentration remains moderate and the redox potential fluctuates around zero, it is determined to be the hypoxic reaction stage; if the dissolved oxygen concentration increases significantly and the redox potential turns positive, it is determined to be the aerobic reaction stage, and finally the treatment stage switching instruction is generated.

[0060] Based on the treatment stage switching instructions, specialized models were developed to depict the changes in microbial activity at each biochemical reaction stage. In the anaerobic reaction stage, phosphorus release kinetic parameters were identified by analyzing the correlation between microbial activity indicators and phosphorus release rates. Data on microbial activity indicators under anaerobic conditions were collected, and changes in phosphorus release rates were monitored. A mathematical model was established through regression analysis to obtain dynamic response parameters for the anaerobic stage. In the anoxic reaction stage, the denitrification process parameter modeling focused on the coupling analysis of denitrifying bacteria activity and total nitrogen removal rate. Real-time monitoring of denitrifying bacteria activity changes was conducted, and combined with the total nitrogen concentration decrease rate, a multiple linear regression algorithm was used to identify quantitative relationships, forming dynamic response parameters for the anoxic stage. In the aerobic reaction stage, the nitrification and phosphorus uptake process was linked to the treatment stage switching instructions. The changes in nitrifying bacteria and polyphosphate-accumulating bacteria activity were modeled synchronously. Water quality data for the aerobic stage was collected, and their activity changes were monitored. The relationship with nitrogen and phosphorus removal efficiency was analyzed, and a mathematical model was established to obtain dynamic response parameters for the aerobic stage. The transfer characteristic analysis integrates the dynamic response parameters of the three stages of anaerobic, hypoxic, and aerobic, identifies the mutual influence and response characteristics of each stage, transforms the input-output response relationship into a predictive control structure, and sets switching conditions and boundary constraints according to the anaerobic-hypoxic-aerobic time sequence combination model to form a complete staged predictive model.

[0061] The phased prediction model is input into the switching model predictive controller for prediction time domain setting. The switching model predictive control algorithm can handle multimodal systems. Dynamic response parameters for anaerobic, anoxic, and aerobic stages are input into the controller for initialization. Time windows for the prediction and control time domains are set, dividing the prediction control time range into multiple continuous control cycles. Prediction time domain data is generated through recursive calculation, starting with the current dissolved oxygen concentration, redox potential, and pH. The phased prediction model is used to deduce the changes in water quality parameters for each future control cycle, resulting in multi-cycle water quality change trajectories. Multi-objective constraint settings convert the target values ​​of effluent chemical oxygen demand, total nitrogen concentration, and total phosphorus concentration into optimization objective functions, ensuring each indicator meets discharge requirements. Cost weight allocation assigns weights to aeration energy consumption, reagent consumption, and sludge production. These three are respectively correlated with blower power, flocculant and disinfectant dosage, and subsequent treatment costs. Water quality control and energy conservation objectives are weighted and combined into a comprehensive optimization objective function. The rolling optimization algorithm iteratively searches for the optimal control sequence in the control time domain. The optimal solution is recalculated in each control cycle. The control strategy is dynamically adjusted according to the real-time water quality. The optimal control parameters for each stage are extracted in segments and combined into a multi-objective control strategy according to time.

[0062] Based on a multi-objective control strategy, parameters are separated in stages, and control parameters for anaerobic, anoxic, and aerobic stages are categorized and extracted to form a set of staged control parameters. The aeration setpoint for the aerobic tank considers both the oxygen demand for nitrification and the oxygen demand for phosphorus uptake by polyphosphate-accumulating organisms (PAOs). The former is calculated based on ammonia nitrogen concentration and nitrification efficiency, while the latter is determined based on phosphorus removal and the PAO coefficient. The optimized aeration setpoint is obtained by superimposing these two parts. The coupled calculation of the return ratio setpoint involves the return ratio within the anoxic tank and the external return ratio within the aerobic tank. The internal return ratio affects denitrification efficiency, while the external return ratio affects sludge concentration control. The optimized return ratio setpoint is calculated by balancing these two requirements. The associated calculation of the stirring intensity setpoint is performed for both the anaerobic and anoxic tanks, optimizing the calculation based on phosphorus release mixing requirements and denitrification mixing requirements, respectively, to obtain the optimized stirring intensity setpoint. Finally, the optimized aeration, return ratio, and stirring intensity setpoints are grouped and configured according to the execution equipment address. The control commands for the blower, return pump, and stirrer each include the corresponding setpoints, forming a distributed control command.

[0063] Distributed control commands are categorized and parsed according to equipment type. Blower control commands, reflux pump control commands, and agitator control commands correspond to different equipment control parameters. After data format conversion, dedicated control data packets for each equipment are obtained. Fieldbus communication protocol, a commonly used communication standard in industrial automation, assembles the control command data and equipment address information according to the protocol format, forming a data frame structure conforming to the fieldbus communication standard. Blower drive control is implemented through a frequency converter, which converts the optimized aeration rate setpoint into a corresponding blower speed control signal. The frequency converter adjusts the blower's operating frequency based on the speed signal to achieve precise aeration rate control. The reflux pump's frequency converter controller receives the optimized reflux ratio setpoint and converts it into a corresponding pump speed adjustment control signal, controlling the reflux flow rate by adjusting the pump's operating speed. The agitator's motor controller receives the optimized agitation intensity setpoint and converts it into a motor power control signal, adjusting the agitator's speed and power output. Synchronization and coordination between devices are achieved through a fieldbus network. The operating status information of blowers, return pumps, and agitators is transmitted to the control center in real time. The control center coordinates and schedules according to the operating status of each device. When a device malfunctions, the operating parameters of other devices can be adjusted in a timely manner to ensure the stable operation of the entire water treatment process.

[0064] In one specific embodiment, the process of performing step S101 may specifically include the following steps:

[0065] Dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors were installed in the anaerobic, anoxic, and aerobic tanks, respectively, to collect water quality parameters of each treatment unit in real time and obtain multi-parameter detection data sequences.

[0066] The multi-parameter detection data sequence is subjected to ternary parameter fusion calculation, and the dissolved oxygen concentration value is compared with the preset low oxygen threshold and high oxygen threshold to obtain the dissolved oxygen state level.

[0067] Based on the dissolved oxygen state level, the redox potential value is segmented and identified, and the potential value is matched with the negative potential threshold, zero potential threshold and positive potential threshold interval to obtain the redox potential state level.

[0068] The acid-base change rate is calculated based on the oxidation-reduction potential state level, and the difference between the acid-base change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator.

[0069] The water quality gradient change markers are time-series verified with the parameter change trends of multiple consecutive sampling periods. The timing of the transition between the anaerobic reaction stage, the anoxic reaction stage, and the aerobic reaction stage is determined and confirmed to obtain the treatment stage switching instruction.

[0070] Specifically, dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors are installed in three different biochemical reaction zones: anaerobic, anoxic, and aerobic tanks. Each sensor detects its corresponding water quality parameter in real time using electrochemical principles. The dissolved oxygen sensor employs electrochemical membrane electrode technology, determining the dissolved oxygen concentration in the water by measuring the intensity of the reduction current at the electrode surface. The output analog signal is converted into a digital signal by an analog-to-digital converter. The oxidation-reduction potential sensor uses a platinum electrode as the indicator electrode and a reference electrode as the reference point, measuring the potential difference between the two electrodes to reflect the oxidation-reduction properties of the water. The potential difference value directly reflects the tendency of electron gain and loss in the water. The pH sensor is based on the ion selectivity principle of glass electrodes, determining the acidity or alkalinity of the water by measuring the potential change caused by the concentration of hydrogen ions. The voltage signal output by the sensor has a logarithmic relationship with the pH value. The data acquisition controller synchronously acquires the signals from each sensor according to a preset sampling frequency, arranging the data from the nine sensors in the three tanks in chronological order to form a multi-parameter detection data sequence containing information such as timestamp, sensor number, value, and unit.

[0071] The ternary parameter fusion calculation is the core of the multi-parameter state recognition algorithm. First, the dissolved oxygen concentration value is extracted from the multi-parameter detection data and compared with preset hypoxia (0.5 mg / L) and hyperxia (2 mg / L) thresholds. The hypoxia threshold represents the critical condition for an anaerobic environment, while the hyperxia threshold represents the initiation condition for an aerobic environment. The comparison process uses conditional branching logic: when the dissolved oxygen concentration is below the hypoxia threshold, it is marked as Level 1, indicating severe hypoxia; when it is between the two, it is marked as Level 2, indicating moderate oxygen; and when it is above the hyperxia threshold, it is marked as Level 3, indicating sufficient oxygen. The state levels are stored numerically for easy subsequent logical operations and conditional judgments. The three-level classification clearly distinguishes different oxygen environments, reflecting the survival status and metabolic intensity of microorganisms.

[0072] Segmented identification processing divides the redox potential into intervals based on the dissolved oxygen state level. The potential value range is typically between -400 mV and +400 mV, and the threshold is dynamically adjusted according to the dissolved oxygen state. In the first level state, the negative potential threshold is -200 mV, the zero potential is zero, and the positive potential is +100 mV, considering the negative potential characteristics of the anaerobic environment. In the second level state, the negative potential threshold is -100 mV, the zero potential remains zero, and the positive potential is +200 mV. Interval matching is achieved through numerical comparison, dividing the potential values ​​into four levels: reducing environment, weakly reducing environment, weakly oxidizing environment, and oxidizing environment.

[0073] The gradient calculation step quantifies the rate of change in pH based on the redox potential state. The rate of change is calculated by dividing the difference between the current pH value and the previous pH value by the time interval. A preset threshold for the rate of change is dynamically set based on the potential state: 0.02 pH units per minute in a reducing environment and 0.03 pH units per minute in an oxidizing environment. Differences are compared to determine if the rate of change exceeds the threshold. If it does, a water quality gradient change indicator is generated and stored using a Boolean type.

[0074] Timing verification ensures the accuracy of stage switching judgments by comparing and analyzing parameter change trends over five consecutive sampling periods using a sliding window mechanism. The window includes information on dissolved oxygen concentration, redox potential, and pH changes, with the trends determined through linear fitting. In the anaerobic stage, dissolved oxygen, redox potential, and pH all show a decreasing or stable trend; in the anoxic stage, dissolved oxygen is required to be at a low and stable level, redox potential fluctuates around zero, and pH is relatively stable; in the aerobic stage, dissolved oxygen is required to increase, redox potential is required to increase positively, and pH is required to increase. Multi-condition combination logic generates stage switching instructions, including stage identifier, switching timestamp, and confidence level information.

[0075] In one specific embodiment, the process of performing step S102 may specifically include the following steps:

[0076] Based on the processing stage switching command, the phosphorus release kinetic parameters of the anaerobic reaction stage are identified and calculated, and the correlation analysis between microbial activity indicators and phosphorus release rate is performed to obtain the dynamic response parameters of the anaerobic stage.

[0077] The denitrification process parameters in the anoxic reaction stage are modeled and calculated according to the processing stage switching command. The denitrifying bacteria activity and total nitrogen removal rate are coupled and analyzed to obtain the dynamic response parameters in the anoxic stage.

[0078] The processing stage switching command is correlated with the nitrification and phosphorus uptake process in the aerobic reaction stage, and the activity changes of nitrifying bacteria and polyphosphate-accumulating bacteria are synchronously modeled to obtain the dynamic response parameters of the aerobic stage.

[0079] Based on the dynamic response parameters of the anaerobic stage, the anoxic stage, and the aerobic stage, the transfer characteristics are analyzed, and the input-output response relationship of each stage is converted into a predictive control structure to obtain stage transfer characteristic data.

[0080] The stage-transfer characteristic data are combined into models according to the time sequence relationship of anaerobic-anoxic-aerobic, and the switching conditions and boundary constraints of each stage model are set and processed to obtain a stage-based prediction model.

[0081] Specifically, to model the changes in microbial activity at each stage of the biochemical reaction, it is first necessary to collect various data related to the switching of treatment stages. These data include switching commands for each reaction stage, dissolved oxygen concentration, pH value, oxidation-reduction potential, ammonia nitrogen concentration, total nitrogen concentration, and total phosphorus concentration. During the data collection phase, a sensor network is used to collect water quality parameters at each treatment stage in real time, and this data is transmitted to the central processing system via a fieldbus communication protocol. The collected data is stored in a database and undergoes preliminary cleaning and preprocessing to remove noise and outliers, ensuring the accuracy and reliability of the data. During data cleaning, reasonable thresholds are set to remove data points outside the normal range, thus ensuring the effectiveness of subsequent analysis. After data preparation, a multi-parameter state recognition algorithm is applied to conduct an in-depth analysis of the relationship between treatment stage switching commands and water quality parameters. By establishing a multiple regression model, with treatment stage switching commands as independent variables and water quality parameters as dependent variables, the changes in microbial activity at different reaction stages are identified. Specifically, based on historical data, linear regression, nonlinear regression, and other algorithms are used to fit the water quality parameters at each stage to obtain their respective mathematical models. These models can reflect the relationship between microbial activity and water quality parameters under different operating conditions.

[0082] Building upon this foundation, machine learning algorithms, such as support vector machines, decision trees, or neural networks, are further utilized to model microbial activity. Through training on historical data, the algorithm can identify potential patterns in microbial activity changes, predicting activity levels under specific conditions. By inputting treatment stage switching commands and corresponding water quality parameters into the training model, the algorithm outputs predicted microbial activity values ​​for each reaction stage. This process not only improves the model's accuracy but also enhances its adaptability, enabling it to remain effective under different treatment stages and operating conditions. During modeling, the temporal relationship between microbial activity and water quality parameters also needs to be considered. Time series analysis can capture the dynamic characteristics of microbial activity changes. Using time series analysis methods such as autoregressive moving average models to model microbial activity effectively analyzes its changing trends and predicts future activity levels. This step is crucial for achieving dynamic control because changes in water quality parameters often have a lag effect; timely prediction provides a basis for adjusting control strategies. After the model is established, it is compared with actual monitoring data, and the predictive performance of the model is evaluated through methods such as cross-validation to ensure the model's reliability and accuracy. By comparing the differences between the actual monitored microbial activity and the model predictions, the model parameters were further optimized to improve the model's fit.

[0083] In one specific embodiment, the process of executing step S103 may specifically include the following steps:

[0084] The phased prediction model is input into the switching model prediction controller to set the prediction time domain, and the water quality change trend of multiple future control cycles is predicted and calculated to obtain the prediction time domain data.

[0085] Based on the predicted time-domain data, the target values ​​of chemical oxygen demand, total nitrogen concentration and total phosphorus concentration in the effluent are constrained, and the water quality compliance requirements are converted into optimization objective functions to obtain multi-objective constraint conditions.

[0086] Based on the multi-objective constraints, cost weights are allocated for aeration energy consumption, reagent consumption, and sludge production. Water quality control objectives and energy conservation and consumption reduction objectives are weighted and combined to obtain a comprehensive optimization objective function.

[0087] The comprehensive optimization objective function is solved using a rolling optimization algorithm, and the optimal control sequence in the control time domain is iteratively searched to obtain the rolling optimized control sequence.

[0088] Based on the rolling optimization control sequence, the control strategies for the anaerobic, hypoxic, and aerobic stages are extracted in segments. The optimal control parameters for each stage are combined according to the temporal relationship to obtain a multi-objective control strategy.

[0089] Specifically, the process is initiated when the anaerobic reaction stage is determined based on the treatment stage switching command. Phosphorus release kinetics refers to the biochemical process by which polyphosphate-accumulating organisms (PAOs) release intracellularly stored phosphate under anaerobic conditions, and its rate is directly affected by the activity state of the microorganisms. The identification and calculation process first collects biomass density data of PAOs in the anaerobic tank, counts the number of PAO cells per unit volume of sludge using microscopic counting, and simultaneously measures the respiration rate of PAOs as an activity indicator. The respiration rate is measured under standard conditions using an oxygen consumption rate meter. The phosphorus release rate is calculated by continuously monitoring the changes in orthophosphate concentration in the anaerobic tank. The release rate is obtained by subtracting the phosphorus concentration of the previous moment from the current phosphorus concentration and dividing by the time interval. The correlation analysis uses linear regression, with the PAO activity indicator as the independent variable and the phosphorus release rate as the dependent variable. The regression coefficient is calculated using the least squares method to establish a quantitative relationship model between microbial activity and phosphorus release rate. During data processing, it is necessary to eliminate the interference of environmental factors such as temperature and pH. By incorporating environmental variables as covariates into the model through multiple regression analysis, the corrected dynamic response parameters of the anaerobic stage are obtained. These parameters include values ​​such as the basic phosphorus release rate constant, activity influence coefficient, and environmental correction factor.

[0090] The denitrification process parameter modeling and calculation are executed upon receiving the anoxic reaction stage switching command. Denitrification is a biological nitrogen removal process in which denitrifying bacteria reduce nitrates and nitrites to nitrogen gas in an anoxic environment. The population structure and quantity distribution of denitrifying bacteria in the anoxic tank are quantitatively detected using fluorescence in situ hybridization (FISH), obtaining biological indicator data on denitrifying bacteria activity. Denitrifying bacteria activity is characterized by measuring the enzyme activity levels of nitrate reductase and nitrite reductase using spectrophotometry. The total nitrogen removal rate is calculated by monitoring the difference in total nitrogen concentration between the influent and effluent of the anoxic tank. A coupling analysis is used to establish a mathematical model of the relationship between denitrifying bacteria activity and total nitrogen removal rate. Nonlinear regression is used to fit the data, and the model uses the Michaelis-Menten equation to describe the relationship between enzyme activity and removal rate. In data processing, influencing factors such as carbon source concentration, temperature, and dissolved oxygen concentration are corrected. Parameter sensitivity analysis is used to determine the weight coefficients of each factor, forming a dynamic response parameter matrix for the anoxic stage.

[0091] The nitrification and phosphorus uptake process is initiated upon receiving a switch command for the aerobic reaction stage. Nitrification is the gradual oxidation of ammonia nitrogen to nitrate by ammonia-oxidizing and nitrite-oxidizing bacteria, while phosphorus uptake is the process by which polyphosphate-accumulating bacteria absorb large amounts of phosphate and store it as polyphosphate under aerobic conditions. Simultaneous modeling requires monitoring the activity changes of nitrifying bacteria and polyphosphate-accumulating bacteria separately, with data acquisition using a real-time monitoring system. The modeling process considers the synergistic effect of the two biochemical processes, employing coupled differential equations to describe the growth kinetics of nitrifying bacteria and polyphosphate-accumulating bacteria. The equations are solved using numerical integration methods to obtain the dynamic response parameters of nitrifying bacteria and polyphosphate-accumulating bacteria activity over time during the aerobic stage.

[0092] Transfer characteristic analysis systematically integrates the dynamic response parameters of the anaerobic, anoxic, and aerobic stages. The analysis process first identifies the input and output variables for each stage. The response relationship transformation is achieved using the state-space method, converting the dynamic response parameters of each stage into matrix forms of state and output equations. The predictive control structure is represented by a multi-input multi-output transfer function matrix. The time-domain differential equation is transformed into a frequency-domain algebraic equation using Laplace transform, facilitating controller design and parameter tuning. Stage transfer characteristic data includes parameters such as the gain matrix, time constant matrix, and coupling coefficient matrix between each stage.

[0093] The model assemblage sequentially connects the stage-transfer characteristic data according to the anaerobic-anoxic-aerobic time sequence, establishing interface relationships between the models at each stage during the assemblage process. Switching conditions are set based on threshold judgments of water quality parameters, and boundary constraints ensure continuity and stability during stage switching. Continuity conditions for state variables and material balance constraints prevent numerical jumps or non-convergence during switching. Constraints include mass conservation constraints, energy balance constraints, and biological rationality constraints, which are incorporated into the optimization objective function using the Lagrange multiplier method. The staged prediction model employs a piecewise linearization method to handle the nonlinear characteristics of each stage, performing linear approximation near each operating point to form a set of piecewise continuous linear models.

[0094] In one specific embodiment, the process of inputting the phased prediction model into the switching model prediction controller for prediction time domain setting, and calculating the water quality change trend for multiple future control cycles, can specifically include the following steps:

[0095] The dynamic response parameters of the anaerobic stage, the hypoxic stage, and the aerobic stage in the phased prediction model are input into the switching model prediction controller to initialize the model parameters of each stage and obtain the initial state data of the controller.

[0096] Based on the controller's initial state data, time windows are set for the prediction time domain length and the control time domain length, dividing the time range of prediction control into multiple continuous control cycles to obtain time domain window configuration parameters.

[0097] The initial state values ​​of the water quality state variables at the current moment are set according to the time domain window configuration parameters. The current values ​​of dissolved oxygen concentration, oxidation-reduction potential and pH are used as the prediction starting point to obtain the prediction initial state vector.

[0098] The predicted initial state vector and the phased prediction model are recursively calculated to gradually predict the changes in water quality parameters in each future control cycle, thus obtaining the multi-cycle water quality change trajectory.

[0099] Based on the multi-cycle water quality change trajectory, the data on the changing trends of chemical oxygen demand removal rate, total nitrogen removal rate, and total phosphorus removal rate are integrated, and the predicted sequences of each water quality indicator are combined and arranged in chronological order to obtain the predicted time domain data.

[0100] Specifically, the switching model predictive controller is an advanced control algorithm specifically designed for multimodal dynamic systems. It automatically selects a suitable predictive model for control calculations based on the current operating state. During initialization, the dynamic response parameters for the anaerobic, anoxic, and aerobic stages of the phased predictive model are input into the corresponding modules of the controller. Anaerobic stage parameters include values ​​such as the phosphorus release rate constant, polyphosphate accumulation bacteria activity coefficient, and organic matter consumption rate; anoxic stage parameters include values ​​such as the denitrification rate constant, denitrifying bacteria activity coefficient, and carbon-to-nitrogen ratio influence factor; and aerobic stage parameters include values ​​such as the nitrification rate constant, phosphorus uptake rate constant, and oxygen transfer coefficient. Model parameter initialization is achieved through parameter matrix assignment. The dynamic response parameters for each stage are arranged in a matrix according to the state equation and output equation format, forming the controller's internal parameter database. The controller's initial state data includes key information such as the currently active model identifier, parameter matrix dimension information, model switching logic judgment conditions, and initial values ​​of state variables. This data lays the foundation for subsequent predictive calculations and control decisions.

[0101] The time window is calculated based on the time constant and response speed parameters in the controller's initial state data. The prediction time domain length is generally set to three to five times the maximum time constant to ensure prediction accuracy, while the control time domain length is relatively shorter to reduce computational burden. When setting the time window, the dominant time constant of the system is first calculated based on the dynamic response parameters of each stage: phosphorus release time constant dominates in the anaerobic stage, denitrification time constant dominates in the anoxic stage, and nitrification time constant dominates in the aerobic stage. The prediction time domain is divided using an equal-interval segmentation method, dividing the total prediction time domain into multiple control cycles with a fixed time step. The division of continuous control cycles needs to consider the limitations of sampling frequency and actuator response speed; typically, the control cycle length is an integer multiple of the sampling cycle and not less than the minimum action time of the actuator. Time domain window configuration parameters include the number of prediction steps, the number of control steps, sampling time, prediction start and end times, etc.

[0102] The initial state setting is the starting point for prediction calculations. Real-time sensor measurements capture current water quality state variables, such as dissolved oxygen concentration, oxidation-reduction potential, and pH. The measured data undergoes preprocessing, including filtering and denoising, outlier detection, and data calibration. Filtering and denoising uses a moving average method to eliminate noise; outlier detection removes data deviating from the normal range by setting thresholds; and data calibration corrects deviations based on sensor calibration curves. Dissolved oxygen concentration, oxidation-reduction potential, and pH together constitute the fundamental vector describing the water quality state. The initial state vector is stored as a column vector, with dimensions matching the number of state variables in the state equation.

[0103] Recursive calculation is the core component of predictive control algorithms. Based on the predicted initial state vector, it iterative calculations are performed using a phased prediction model. The recursive calculation employs numerical integration methods to solve the differential equation system; common methods include the Euler method and the Runge-Kutta method, the latter offering higher accuracy but requiring more computation. Within each control cycle, the state value for the next time step is calculated based on the current state and control input. The state transition equation describes the changes in state variables over time, and the output equation describes the relationship between the observed quantities and the state variables. The prediction process must consider model switching between different biochemical reaction stages. When a stage switching condition is encountered, the recursive calculation automatically switches to the next stage model, maintaining the continuity of state variables. The multi-cycle water quality change trajectory records the water quality parameter values ​​at each prediction time in time series form, including timestamps, dissolved oxygen concentration, redox potential, and predicted pH values.

[0104] Data integration transforms the raw predicted data of multi-period water quality change trajectories into evaluation indicators for treatment effectiveness. For example, the chemical oxygen demand (COD) removal rate is calculated by dividing the difference between the influent and effluent concentrations by the influent concentration; the total nitrogen (TNO) and total phosphorus (TP) removal rates are calculated using the same method. Trend analysis employs linear fitting, performing linear regression analysis on the changes in each removal rate over time. The slope of the regression line reflects the trend of the removal rate change; a positive slope indicates an upward trend, and a negative slope indicates a downward trend. The predicted sequences of each water quality indicator are arranged in chronological order to form an ordered array. The combination and arrangement process aligns the COD, TNO, and TPB removal rate sequences according to the same time index, forming a multi-dimensional predicted data matrix. The predicted time-domain data is stored in matrix form, with row indices corresponding to time nodes, column indices corresponding to different water quality indicators, and matrix elements representing the predicted values ​​of the indicators at the corresponding times.

[0105] In one specific embodiment, the process of executing step S104 may specifically include the following steps:

[0106] The control parameters for the anaerobic stage, the anoxic stage, and the aerobic stage in the multi-objective control strategy are separated into stage parameters, and the optimal control parameters for each stage are categorized and extracted to obtain a set of stage control parameters.

[0107] Based on the set of phased control parameters, the aeration set value of the aerobic tank is calculated and processed. The oxygen demand of nitrification reaction and the oxygen demand of phosphorus uptake by polyphosphate-accumulating bacteria are superimposed to obtain the optimized aeration set value.

[0108] Based on the set of phased control parameters, the set values ​​of the recirculation ratio in the anoxic tank and the recirculation ratio outside the aerobic tank are coupled and calculated. The denitrification efficiency requirement and the sludge concentration control requirement are balanced to obtain the optimized recirculation ratio set value.

[0109] The set of staged control parameters is correlated with the stirring intensity of the anaerobic and anoxic tanks, and the phosphorus release mixing intensity and denitrification mixing intensity are optimized respectively to obtain the optimized stirring intensity set value.

[0110] Based on the optimized aeration rate setting, optimized reflux ratio setting, and optimized stirring intensity setting, control instructions are encapsulated, and each control parameter is grouped and configured according to the execution device address to obtain distributed control instructions.

[0111] Specifically, phased parameter separation is a data processing procedure that categorizes and organizes the comprehensive control parameters output by the multi-objective control strategy according to the biochemical reaction stages. The control parameters output by the multi-objective control strategy include the setpoints of multiple control variables such as aeration rate, reflux ratio, and stirring intensity at different time periods. Parameter separation first divides the control parameter sequence into time periods based on the time nodes of the treatment stage switching commands: the anaerobic stage corresponds to the time period when the dissolved oxygen concentration is below 0.5 mg / L, the anoxic stage corresponds to the time period when the dissolved oxygen concentration is between 0.5 and 2 mg / L, and the aerobic stage corresponds to the time period when the dissolved oxygen concentration is above 2 mg / L. The classification and extraction process groups the control parameter values ​​within each time period according to parameter type. The control parameters for the anaerobic stage mainly include the stirring intensity setpoint and the influent flow rate control value; the control parameters for the anoxic stage include the internal reflux ratio setpoint and the stirring intensity setpoint; and the control parameters for the aerobic stage include the aeration rate setpoint and the external reflux ratio setpoint. The optimal control parameters are determined by comparing the relationship between the parameter values ​​and the objective function values ​​within each time period, selecting the parameter combination that optimizes the objective function as the optimal control parameters for that stage. The phased control parameter set is stored in a structured data format. Each phase corresponds to a parameter set, which contains information such as parameter name, parameter value, parameter unit, and applicable time range.

[0112] The aeration setpoint is calculated based on dissolved oxygen demand and microbial activity parameters during the aerobic stage, encompassing the processes of ammonia nitrogen oxidation and further nitrite oxidation. The oxygen demand for nitrification is determined by multiplying the influent ammonia nitrogen concentration by the theoretical oxygen consumption coefficient, which is 4.57 grams of oxygen per gram of ammonia nitrogen. Actual values ​​need to be adjusted based on nitrifying bacteria activity and environmental conditions. The oxygen demand for phosphorus uptake by polyphosphate-accumulating bacteria is based on their biochemical process of absorbing phosphate and synthesizing polyphosphate under aerobic conditions; approximately 1.5 grams of oxygen are required to remove each gram of phosphorus. The total oxygen demand is the sum of the oxygen demand for nitrification and polyphosphate uptake by polyphosphate-accumulating bacteria. The actual aeration requirement is then calculated based on oxygen transfer efficiency and a safety factor. The optimized aeration setpoint is obtained by dividing the total oxygen demand by the oxygen transfer coefficient, which is affected by factors such as aerator type, water depth, and water temperature.

[0113] The coupled calculation of the return ratio setpoint involves two parameters: the internal return ratio in the anoxic tank and the external return ratio in the aerobic tank. The internal return ratio affects denitrification efficiency and needs to be determined based on the target total nitrogen removal rate. The external return ratio is used to maintain a suitable sludge concentration in the biological treatment tank. A multi-objective optimization method is employed for the equilibrium treatment, adjusting the importance of denitrification efficiency and sludge concentration control through weighting coefficients to find the optimal combination of return ratios. Optimizing the return ratio setpoint is achieved through iterative calculation.

[0114] The mixing intensity correlation calculation optimizes the design based on the different mixing requirements of the anaerobic and anoxic tanks. The mixing intensity in the anaerobic tank must ensure sufficient contact between polyphosphate-accumulating bacteria and organic matter, avoiding over-mixing. The mixing intensity in the anoxic tank must ensure sufficient contact between denitrifying bacteria, nitrates, and organic matter, maintaining an anaerobic environment. The optimization process independently calculates the mixing intensity for the anaerobic and anoxic tanks, considering both minimizing energy consumption and maximizing treatment efficiency, and determines the most economical mixing intensity through cost-benefit analysis.

[0115] The control command encapsulation converts the optimized aeration rate setpoint, reflux ratio setpoint, and agitation intensity setpoint into a control command format that the actuator can recognize. The control command includes fields such as device address, function code, data value, and checksum, and uses standard industrial communication protocols, such as Modbus or Profibus, to transmit it to the actuator via a fieldbus network.

[0116] In one specific embodiment, the process of executing step S105 may specifically include the following steps:

[0117] The distributed control commands are classified and parsed according to the blower control command, the reflux pump control command, and the agitator control command. The control parameters of each device are converted into data formats to obtain device-specific control data packages.

[0118] Based on the device-specific control data packet, the data frame structure of the fieldbus communication protocol is encapsulated, and the control command data and device address information are assembled according to the protocol format to obtain the bus communication data frame.

[0119] The frequency converter of the aerobic tank blower is driven by the bus communication data frame, and the optimized aeration volume setpoint is converted into a blower speed control signal to obtain the blower operation control status.

[0120] The bus communication data frame is connected to the frequency converter controller of the internal and external reflux pumps to perform pump speed adjustment control on the optimized reflux ratio set value, thereby obtaining the reflux pump operation control status.

[0121] Based on the bus communication data frames, instructions are transmitted to the motor controllers of the agitators in the anaerobic and anoxic tanks. The optimized stirring intensity setting value is converted into a motor power control signal, and synchronized with the operation control status of the blower and the operation control status of the return pump to obtain the coordinated operation status of the water treatment equipment.

[0122] Specifically, equipment classification and parsing is a data parsing process that groups distributed control instructions according to the type of executing equipment. Distributed control instructions contain various control parameters and setpoints. Classification and parsing first categorizes control instructions by equipment identifier. For example, blower control instructions include parameters such as aeration volume and start / stop; return pump control instructions include parameters such as internal / external return ratio; and agitator control instructions include parameters such as agitation intensity. Data format conversion is the process of converting control parameters from standardized values ​​to equipment-specific formats. Different equipment's frequency converters and controllers have different data formats and ranges. Blower frequency converters use frequency setting control, requiring the aeration volume setpoint to be converted to a frequency value, considering the characteristics of the blower and piping; return pump frequency converters use speed percentage control, requiring the return ratio setpoint to be converted to a pump speed percentage value, based on the pump's performance and flow characteristics; agitator motor controllers use power settings, requiring the agitation intensity setpoint to be converted to a motor power value, considering the agitator's torque and load characteristics. Equipment-specific control data packets are stored in a structured format, containing fields such as equipment type identifier, parameter type, value, unit, and timestamp.

[0123] Data frame encapsulation is based on the standard format requirements of fieldbus communication protocols, which are widely used in industrial automation. Common protocols include Modbus, PROFIBUS, and CAN bus. The first step in encapsulation is to determine the communication protocol type and its data frame format. For example, a Modbus protocol data frame contains fields such as device address, function code, data content, and error check, each with a fixed byte length and format requirements. The device address is a unique identifier for a device in a fieldbus network, typically represented by an eight-bit or sixteen-bit value. Control command data must be encoded according to the protocol format; numerical data uses binary or hexadecimal encoding, while status data uses bitwise operations. Protocol format assembly arranges the device address, function code, control data, and checksum in a prescribed order to form a complete data frame. The data frame length is dynamically adjusted according to the amount of control data.

[0124] During the blower drive signal transmission process, control commands are transmitted to the blower frequency converter via a fieldbus network. The blower in the aerobic tank is a key device providing aeration, and its operating status directly affects the biochemical reaction effect. The first step in drive signal transmission is establishing a communication connection with the blower frequency converter, including a physical connection (achieved via a bus cable) and a logical connection (confirming identity and parameters through a handshake protocol). Optimizing the aeration setpoint requires conversion into a blower speed control signal; this process must consider the blower's performance characteristics and operating conditions. The speed control signal is achieved by adjusting the frequency converter's output frequency. Since there is a non-linear relationship between speed and aeration, the frequency converter automatically adjusts its output frequency and voltage upon receiving the signal, thereby driving the blower to reach the target speed.

[0125] The communication connection for the return pumps involves both the internal and external return pumps. The internal return pump is responsible for returning the mixed liquor from the aerobic tank to the anoxic tank, while the external return pump is responsible for returning the sludge from the secondary sedimentation tank to the biological treatment tank. The establishment of the communication connection is similar, using a fieldbus network to exchange data between the controller and the frequency converter. Optimizing the return ratio setpoint requires converting it to the corresponding pump speed setpoint, based on the pump's flow characteristic curve and the piping system characteristics. Internal return ratio control is achieved by adjusting the speed of the internal return pump, while the external return pump must overcome the gravity of the sludge and piping resistance.

[0126] The agitator command transmission enables independent control of the agitation equipment in both anaerobic and anoxic tanks. The motor controller achieves precise control of the agitation intensity by adjusting the motor's power supply voltage and frequency. The command transmission process sends the optimized agitation intensity setpoint to the motor controller via bus communication data frames, including command encoding, data verification, and communication confirmation. The optimized agitation intensity setpoint needs to be converted into a motor power control signal, taking into account the mechanical and load characteristics of the agitator.

[0127] Synchronization and coordination is a control mechanism that ensures the coordinated operation of multiple devices. The operating status of blowers, return pumps, and agitators must be consistent in terms of time and function. Coordinated control is achieved by monitoring the operational feedback information of each device. When one device malfunctions, the system automatically adjusts the operating parameters of other devices to maintain the overall treatment effect. This coordination mechanism ensures the efficient and stable operation of the water treatment system, improving overall treatment efficiency and water quality control.

[0128] The above describes the water treatment process optimization control method in the embodiments of this application. The following describes the water treatment process optimization control system in the embodiments of this application. Please refer to [link / reference]. Figure 2 One embodiment of the water treatment process optimization control system in this application includes:

[0129] The discrimination module 201 is used to process the data detected by the dissolved oxygen sensor, oxidation-reduction potential sensor, and pH sensor using a multi-parameter state recognition algorithm to obtain a treatment stage switching instruction. This includes: deploying dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors in the anaerobic, anoxic, and aerobic tanks respectively to collect water quality parameters of each treatment unit in real time, obtaining a multi-parameter detection data sequence; performing ternary parameter fusion calculation on the multi-parameter detection data sequence, comparing the dissolved oxygen concentration value with preset low-oxygen and high-oxygen thresholds to determine the dissolved oxygen state level; and based on the... The dissolved oxygen state level is used to segment and identify the redox potential value. The potential value is matched with negative potential threshold, zero potential threshold, and positive potential threshold to obtain the redox potential state level. The acid-base change rate is calculated according to the redox potential state level. The difference between the acid-base change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator. The water quality gradient change indicator is verified with the parameter change trend of multiple consecutive sampling periods to identify and confirm the switching time of the anaerobic reaction stage, the anoxic reaction stage, and the aerobic reaction stage, and obtain the treatment stage switching instruction.

[0130] The modeling module 202 is used to model the changes in microbial activity in each biochemical reaction stage according to the processing stage switching instruction, and obtain a staged prediction model.

[0131] Optimization module 203 is used to perform rolling optimization processing on the staged prediction model by switching model prediction control algorithm to obtain a multi-objective control strategy;

[0132] Control module 204 is used to perform coordinated calculations on the aeration volume setpoint, reflux ratio setpoint, and stirring intensity setpoint according to the multi-objective control strategy to obtain distributed control commands;

[0133] The drive module 205 is used to process the distributed control commands to drive the blower, return pump and agitator through the fieldbus communication protocol to obtain the coordinated operation status of the water treatment equipment.

[0134] above Figure 2 The water treatment process optimization control system in this embodiment of the invention is described in detail from the perspective of modular functional entities. The water treatment process optimization control equipment in this embodiment of the invention is described in detail from the perspective of hardware processing.

[0135] Reference Figure 3 This invention also provides a water treatment process optimization and control device, which can be a server, and its internal structure can be as follows: Figure 3As shown, the water treatment process optimization and control device includes a processor, memory, display screen, input device, network interface, and database connected via a system bus. The processor in this computer design provides computing and control capabilities. The memory of the water treatment process optimization and control device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system, computer programs, and 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 of the water treatment process optimization and control device is used to store the data corresponding to this embodiment. The network interface of the water treatment process optimization and control device is used for communication with external terminals via network connection. When the computer program is executed by the processor, it implements the above-described method.

[0136] 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 water treatment process optimization and control equipment to which the present invention is applied.

[0137] The present invention also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when the instructions are executed on a computer, cause the computer to perform the steps of the water treatment process optimization control method.

[0138] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0139] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a water treatment process optimization control device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0140] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for optimizing and controlling a water treatment process, characterized in that, The method includes: A multi-parameter state recognition algorithm is used to process the detection data from dissolved oxygen, oxidation-reduction potential, and pH sensors to determine the state and generate treatment stage switching instructions. This includes: deploying dissolved oxygen, oxidation-reduction potential, and pH sensors in the anaerobic, anoxic, and aerobic tanks respectively; collecting water quality parameters in each treatment unit in real time to obtain a multi-parameter detection data sequence; performing ternary parameter fusion calculations on the multi-parameter detection data sequence; comparing the dissolved oxygen concentration value with preset low-oxygen and high-oxygen thresholds to determine the dissolved oxygen state level; and based on the dissolved oxygen... The redox potential value is segmented and identified based on the state level. The potential value is matched with negative potential threshold, zero potential threshold, and positive potential threshold to obtain the redox potential state level. The acid-base change rate is calculated according to the redox potential state level. The difference between the acid-base change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator. The water quality gradient change indicator is verified with the parameter change trend of multiple consecutive sampling periods to identify and confirm the transition time of anaerobic reaction stage, anoxic reaction stage, and aerobic reaction stage, and obtain the treatment stage switching instruction. The microbial activity change patterns at each biochemical reaction stage are modeled based on the processing stage switching instructions to obtain a staged prediction model. The phased prediction model is subjected to rolling optimization processing through a switching model predictive control algorithm to obtain a multi-objective control strategy. Based on the multi-objective control strategy, the aeration rate setpoint, reflux ratio setpoint, and stirring intensity setpoint are coordinated and calculated to obtain distributed control commands. The distributed control commands are used to drive and control the blower, reflux pump, and agitator via a fieldbus communication protocol to obtain the coordinated operating status of the water treatment equipment.

2. The water treatment process optimization control method according to claim 1, characterized in that, The process of modeling the changes in microbial activity at each biochemical reaction stage based on the processing stage switching instruction yields a staged prediction model, including: Based on the processing stage switching command, the phosphorus release kinetic parameters of the anaerobic reaction stage are identified and calculated, and the correlation analysis between microbial activity indicators and phosphorus release rate is performed to obtain the dynamic response parameters of the anaerobic stage. The denitrification process parameters in the anoxic reaction stage are modeled and calculated according to the processing stage switching command. The denitrifying bacteria activity and total nitrogen removal rate are coupled and analyzed to obtain the dynamic response parameters in the anoxic stage. The processing stage switching command is correlated with the nitrification and phosphorus uptake process in the aerobic reaction stage, and the activity changes of nitrifying bacteria and polyphosphate-accumulating bacteria are synchronously modeled to obtain the dynamic response parameters of the aerobic stage. Based on the dynamic response parameters of the anaerobic stage, the anoxic stage, and the aerobic stage, the transfer characteristics are analyzed, and the input-output response relationship of each stage is converted into a predictive control structure to obtain stage transfer characteristic data. The stage-transfer characteristic data are combined into models according to the time sequence relationship of anaerobic-anoxic-aerobic, and the switching conditions and boundary constraints of each stage model are set and processed to obtain a stage-based prediction model.

3. The water treatment process optimization control method according to claim 1, characterized in that, The step of performing rolling optimization on the phased prediction model using a switching model predictive control algorithm to obtain a multi-objective control strategy includes: The phased prediction model is input into the switching model prediction controller to set the prediction time domain, and the water quality change trend of multiple future control cycles is predicted and calculated to obtain the prediction time domain data. Based on the predicted time-domain data, the target values ​​of chemical oxygen demand, total nitrogen concentration and total phosphorus concentration in the effluent are constrained, and the water quality compliance requirements are converted into optimization objective functions to obtain multi-objective constraint conditions. Based on the multi-objective constraints, cost weights are allocated for aeration energy consumption, reagent consumption, and sludge production. Water quality control objectives and energy conservation and consumption reduction objectives are weighted and combined to obtain a comprehensive optimization objective function. The comprehensive optimization objective function is solved using a rolling optimization algorithm, and the optimal control sequence in the control time domain is iteratively searched to obtain the rolling optimized control sequence. Based on the rolling optimization control sequence, the control strategies for the anaerobic, hypoxic, and aerobic stages are extracted in segments. The optimal control parameters for each stage are combined according to the temporal relationship to obtain a multi-objective control strategy.

4. The water treatment process optimization control method according to claim 3, characterized in that, The step involves inputting the phased prediction model into the switching model prediction controller for prediction time domain setting, and performing prediction calculations on the water quality change trend for multiple future control cycles to obtain prediction time domain data, including: The dynamic response parameters of the anaerobic stage, the hypoxic stage, and the aerobic stage in the phased prediction model are input into the switching model prediction controller to initialize the model parameters of each stage and obtain the initial state data of the controller. Based on the controller's initial state data, time windows are set for the prediction time domain length and the control time domain length, dividing the time range of prediction control into multiple continuous control cycles to obtain time domain window configuration parameters. The initial state values ​​of the water quality state variables at the current moment are set according to the time domain window configuration parameters. The current values ​​of dissolved oxygen concentration, oxidation-reduction potential and pH are used as the prediction starting point to obtain the prediction initial state vector. The predicted initial state vector and the phased prediction model are recursively calculated to gradually predict the changes in water quality parameters in each future control cycle, thus obtaining the multi-cycle water quality change trajectory. Based on the multi-cycle water quality change trajectory, the data on the changing trends of chemical oxygen demand removal rate, total nitrogen removal rate, and total phosphorus removal rate are integrated, and the predicted sequences of each water quality indicator are combined and arranged in chronological order to obtain the predicted time domain data.

5. The water treatment process optimization control method according to claim 1, characterized in that, The process of coordinating and calculating the aeration rate setpoint, reflux ratio setpoint, and stirring intensity setpoint according to the multi-objective control strategy to obtain distributed control commands includes: The control parameters for the anaerobic stage, the anoxic stage, and the aerobic stage in the multi-objective control strategy are separated into stage parameters, and the optimal control parameters for each stage are categorized and extracted to obtain a set of stage control parameters. Based on the set of phased control parameters, the aeration set value of the aerobic tank is calculated and processed. The oxygen demand of nitrification reaction and the oxygen demand of phosphorus uptake by polyphosphate-accumulating bacteria are superimposed to obtain the optimized aeration set value. Based on the set of phased control parameters, the set values ​​of the recirculation ratio in the anoxic tank and the recirculation ratio outside the aerobic tank are coupled and calculated. The denitrification efficiency requirement and the sludge concentration control requirement are balanced to obtain the optimized recirculation ratio set value. The set of staged control parameters is correlated with the stirring intensity of the anaerobic and anoxic tanks, and the phosphorus release mixing intensity and denitrification mixing intensity are optimized respectively to obtain the optimized stirring intensity set value. Based on the optimized aeration rate setting, optimized reflux ratio setting, and optimized stirring intensity setting, control instructions are encapsulated, and each control parameter is grouped and configured according to the execution device address to obtain distributed control instructions.

6. The water treatment process optimization control method according to claim 1, characterized in that, The process of using the distributed control commands to drive and control the blower, return pump, and agitator via a fieldbus communication protocol to obtain the coordinated operating status of the water treatment equipment includes: The distributed control commands are classified and parsed according to the blower control command, the reflux pump control command, and the agitator control command. The control parameters of each device are converted into data formats to obtain device-specific control data packages. Based on the device-specific control data packet, the data frame structure of the fieldbus communication protocol is encapsulated, and the control command data and device address information are assembled according to the protocol format to obtain the bus communication data frame. The frequency converter of the aerobic tank blower is driven by the bus communication data frame, and the optimized aeration volume setpoint is converted into a blower speed control signal to obtain the blower operation control status. The bus communication data frame is connected to the frequency converter controller of the internal and external reflux pumps to perform pump speed adjustment control on the optimized reflux ratio set value, thereby obtaining the reflux pump operation control status. Based on the bus communication data frames, instructions are transmitted to the motor controllers of the agitators in the anaerobic and anoxic tanks. The optimized stirring intensity setting value is converted into a motor power control signal, and synchronized with the operation control status of the blower and the operation control status of the return pump to obtain the coordinated operation status of the water treatment equipment.

7. A water treatment process optimization control system, characterized in that, For implementing the water treatment process optimization control method as described in any one of claims 1-6, the water treatment process optimization control system comprises: The discrimination module is used to process the detection data from the dissolved oxygen sensor, oxidation-reduction potential sensor, and pH sensor using a multi-parameter state recognition algorithm to obtain processing stage switching instructions. This includes: deploying dissolved oxygen sensors, oxidation-reduction potential sensors, and pH sensors in the anaerobic, anoxic, and aerobic tanks respectively; collecting water quality parameters from each treatment unit in real time to obtain a multi-parameter detection data sequence; performing ternary parameter fusion calculations on the multi-parameter detection data sequence; comparing the dissolved oxygen concentration value with preset low-oxygen and high-oxygen thresholds to determine the dissolved oxygen state level; and based on the... The dissolved oxygen status level is used to segment and identify the redox potential value. The potential value is matched with negative potential threshold, zero potential threshold, and positive potential threshold to obtain the redox potential status level. The acid-base change rate is calculated according to the redox potential status level. The difference between the acid-base change rate and the preset change rate threshold is compared to obtain the water quality gradient change indicator. The water quality gradient change indicator is verified with the parameter change trend of multiple consecutive sampling periods to identify and confirm the transition time of anaerobic reaction stage, anoxic reaction stage, and aerobic reaction stage, and obtain the treatment stage switching instruction. The modeling module is used to model the changes in microbial activity at each biochemical reaction stage according to the processing stage switching instructions, and obtain a staged prediction model. The optimization module is used to perform rolling optimization on the staged prediction model by switching model prediction control algorithms to obtain a multi-objective control strategy. The control module is used to coordinate and calculate the aeration volume setpoint, reflux ratio setpoint, and stirring intensity setpoint according to the multi-objective control strategy to obtain distributed control commands. The drive module is used to process the distributed control commands to drive and control the blower, return pump, and agitator through the fieldbus communication protocol, so as to obtain the coordinated operation status of the water treatment equipment.

8. A water treatment process optimization and control device, characterized in that, The method includes a memory and a processor, wherein the memory stores a computer program that can run on the processor, and the processor executes the computer program to implement the water treatment process optimization control method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is run by the processor, it causes the processor to execute the water treatment process optimization control method as described in any one of claims 1 to 6.

Images