Air floatation separation system based on big data model

Abstract

The application discloses a gas floatation separation system based on a big data model and belongs to the technical field of water treatment, which comprises a data acquisition module, a big data model processing module and a real-time control module; wherein the data acquisition module is used for acquiring water quality parameters and equipment operation parameters of the gas floatation separation system and generating real-time monitoring data; the big data model processing module is used for acquiring historical operation data, performing prediction processing on the real-time monitoring data and the historical operation data and generating an optimized parameter set; and the real-time control module is used for adjusting control parameters of the gas floatation separation system according to the optimized parameter set and performing an optimized separation operation. Through intelligent data acquisition, big data processing and control functions, the application realizes intelligent optimized control of the gas floatation separation process and improves the separation efficiency and stability.

Description

Technical Field

[0001] This invention relates to the field of water treatment technology, and in particular to an air flotation separation system based on a big data model. Background Technology

[0002] Air flotation separation, as a highly efficient solid-liquid or liquid-liquid separation technology, plays a crucial role in water treatment, particularly in the pretreatment of industrial wastewater, municipal sewage, and drinking water. This technology generates a large number of tiny bubbles in water, which attach to suspended particles or flocs, forming a float with a density less than water. These float to the surface, forming a scum layer, which is then removed by a skimmer, thus achieving the separation of pollutants and the purification of water.

[0003] In existing technologies, the operation and control of air flotation separation systems largely rely on the experience of operators or simple automated control logic. Operators typically adjust the frequency of the reagent dosing pump and the operating parameters of the bubble generator manually based on offline monitoring results of the influent water quality or visual observation on-site. Although some automated systems use online instruments for single-parameter feedback control, such as adjusting the reagent dosage solely based on the effluent turbidity, this control method is relatively simplistic, and the setpoints of its control parameters are often fixed or require periodic manual modification.

[0004] However, the aforementioned existing technical solutions have significant limitations. Because influent water quality parameters such as turbidity, pH, and temperature often exhibit complex and irregular dynamic changes, relying solely on manual experience or single-parameter feedback control has a significant lag, making it difficult to quickly and accurately match optimal operating parameters. This lag and mismatch often lead to unstable treatment results, sometimes resulting in substandard effluent quality, and sometimes requiring excessive addition of chemicals or excessive gas supply to ensure treatment effectiveness, causing serious waste of chemicals and energy and increasing system operating costs. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides an air flotation separation system based on a big data model. Through intelligent data acquisition, big data processing, and control functions, it achieves intelligent optimization control of the air flotation separation process, thereby improving the efficiency and stability of the separation.

[0006] The above objectives can be achieved through the following approach: The air flotation separation system based on a big data model includes a data acquisition module, a big data model processing module, and a real-time control module. The data acquisition module obtains the influent water quality parameters and equipment operating parameters of the air flotation separation system, generating real-time monitoring data. The big data model processing module acquires historical operating data, performs predictive processing on the real-time monitoring data and historical operating data, and generates an optimized parameter set. The real-time control module adjusts the control parameters of the air flotation separation system according to the optimized parameter set, executing optimized separation operations.

[0007] Furthermore, the data acquisition module includes: a water quality parameter acquisition unit, an equipment parameter acquisition unit, and a parameter processing unit; wherein, the water quality parameter acquisition unit is used to monitor the influent turbidity, pH value, and temperature, and generate a water quality parameter set; the equipment parameter acquisition unit is used to monitor the pressure and flow parameters of the bubble generator, and generate an equipment parameter set; the parameter processing unit is connected to the water quality parameter acquisition unit and the equipment parameter acquisition unit respectively, and is used to combine the water quality parameter set and the equipment parameter set to generate real-time monitoring data.

[0008] Furthermore, the step of combining the water quality parameter set and the equipment parameter set to generate real-time monitoring data includes: integrating the water quality parameter set and the equipment parameter set to obtain a comprehensive parameter set; and standardizing the comprehensive parameter set to generate real-time monitoring data.

[0009] Furthermore, the big data model processing module includes: a data acquisition unit, a predictive analysis unit, a parameter optimization unit, and an update processing unit; wherein, the data acquisition unit is used to acquire historical operation data; the predictive analysis unit is connected to the data acquisition unit and the parameter processing unit respectively, and is used to perform multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results; the parameter optimization unit is connected to the predictive analysis unit, and is used to calculate the optimal bubble size, reagent dosage, and airflow rate based on the intermediate prediction results to generate preliminary optimization parameters; the update processing unit is connected to the parameter optimization unit, and is used to update the preliminary optimization parameters to generate an optimized parameter set.

[0010] Furthermore, the step of performing multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results includes: performing multivariate analysis on the real-time monitoring data and the historical operation data to obtain the flocculation index, the gas-solid bonding probability factor, and the scum layer stability coefficient; and using the flocculation index, the gas-solid bonding probability factor, and the scum layer stability coefficient to generate intermediate prediction results.

[0011] Furthermore, the real-time control module includes: a parameter setting unit, an execution control unit, and a feedback adjustment unit; wherein, the parameter setting unit, connected to the update processing unit, is used to set the target pressure value of the bubble generator and the target rate value of the reagent dosing pump according to the optimized parameter set, and generate a control command set; the execution control unit, connected to the parameter setting unit, is used to execute the control command set and adjust the control parameters of the air flotation separation system; the feedback adjustment unit, connected to the execution control unit, is used to collect effluent quality parameters in real time and, in conjunction with the adjusted control parameters, perform optimized separation operations.

[0012] Furthermore, the real-time acquisition of effluent quality parameters, combined with the adjusted control parameters, to perform optimized separation operations includes: monitoring effluent turbidity values ​​and generating effluent monitoring data; acquiring quality standards for measuring effluent quality, comparing the effluent monitoring data with the quality standards, and generating deviation indicators; performing feedback adjustment processing on the deviation indicators to generate parameter correction instructions; and adjusting the target pressure value of the bubble generator and the target rate value of the reagent dosing pump based on the parameter correction instructions to perform optimized separation operations.

[0013] Furthermore, the process of monitoring effluent turbidity values ​​and generating effluent monitoring data includes: continuously and periodically collecting effluent turbidity values ​​to obtain a sequence of raw turbidity sampling values; and performing sliding window mean filtering and data standardization on the raw turbidity sampling value sequence to generate effluent monitoring data.

[0014] Furthermore, updating the preliminary optimization parameters and generating an optimization parameter set includes: updating the preliminary optimization parameters to obtain corrected optimization parameters; and using the corrected optimization parameters to perform operating condition adaptation verification to generate an optimization parameter set.

[0015] Compared with the prior art, the present invention has the following advantages: This invention realizes predictive intelligent control of the air flotation separation process. By performing in-depth analysis of real-time monitoring data and historical operation data through a big data model processing module, it can predict the system operation trend in advance and generate a forward-looking set of optimized parameters. This changes the lag of traditional control methods, enabling the system to actively respond to fluctuations in influent water quality rather than responding passively.

[0016] This invention improves the overall operating efficiency and economy of the air flotation separation system. Based on the optimized parameter set output by the model, the system can accurately adjust the dosage of reagents and the working state of the bubble generator. While ensuring that the effluent water quality meets the standards, it avoids excessive addition of chemical reagents and ineffective energy consumption of equipment, thereby reducing resource management and operating costs.

[0017] This invention enhances the stability and robustness of the control system. The system not only relies on model prediction optimization, but also uses the feedback adjustment unit in the real-time control module to monitor and correct the effluent quality in real time. It combines the global optimization capability of the model with the precise correction capability of feedback control, ensuring that even when there are deviations in the model prediction or sudden changes in the operating conditions, the system can still converge to a stable and efficient operating state.

[0018] Other features and advantages of the invention will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the invention. The objects and other advantages of the invention may be realized and obtained by means of the structures pointed out in the description, claims and drawings. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art 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.

[0020] Figure 1 This is a framework diagram of an air flotation separation system based on a big data model, according to an embodiment of the present invention.

[0021] Figure 2 This is a schematic diagram of the air flotation separation system based on a big data model according to an embodiment of the present invention.

[0022] Figure 3 This is a flowchart of the real-time control module in an embodiment of the present invention. Detailed Implementation

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Reference Figure 1 One embodiment of the present invention proposes an air flotation separation system based on a big data model. Through intelligent data acquisition, big data processing and control functions, it realizes intelligent optimization control of the air flotation separation process, thereby improving the separation efficiency and stability.

[0025] like Figure 2As shown, the system in this embodiment specifically includes: a data acquisition module, a big data model processing module, and a real-time control module; wherein, The data acquisition module is used to acquire the influent water quality parameters and equipment operating parameters of the air flotation separation system and generate real-time monitoring data. The big data model processing module is used to acquire historical operation data, perform predictive processing on the real-time monitoring data and the historical operation data, and generate an optimized parameter set. The real-time control module is used to adjust the control parameters of the air flotation separation system according to the optimized parameter set and to perform optimized separation operations.

[0026] Specifically, the data acquisition module captures the real-time and comprehensive operating status of the air flotation separation system, including influent water quality parameters and equipment operating parameters, forming real-time monitoring data reflecting the current operating conditions. Subsequently, the big data model processing module combines this real-time data with accumulated historical operating data, utilizing big data analysis and prediction models to understand the complex nonlinear relationships between various system variables and proactively predict the separation effect under the current conditions. Based on this prediction, the module further calculates the optimal combination of control parameters to achieve the best separation efficiency through optimization algorithms, generating an optimized parameter set. Finally, the real-time control module receives this optimization command and precisely controls the relevant actuators of the air flotation separation system, translating the optimization decision into actual physical operations. This completes the entire control closed loop from data perception to intelligent decision-making to precise execution, achieving proactive and predictive control of the air flotation separation process.

[0027] Furthermore, the data acquisition module includes: a water quality parameter acquisition unit, an equipment parameter acquisition unit, and a parameter processing unit; wherein, The water quality parameter acquisition unit is used to monitor the turbidity, pH value and temperature of the influent and generate a water quality parameter set; The equipment parameter acquisition unit is used to monitor the pressure and flow parameters of the bubble generator and generate a set of equipment parameters; The parameter processing unit is connected to both the water quality parameter acquisition unit and the equipment parameter acquisition unit, and is used to combine the water quality parameter set and the equipment parameter set to generate real-time monitoring data.

[0028] Specifically, the water quality parameter acquisition unit first performs its function by deploying high-precision sensors at the inlet of the air flotation separation system. These sensors include an online turbidimeter for measuring suspended solids concentration, a pH electrode for determining the acidity or alkalinity of the water, and a temperature sensor for monitoring the thermodynamic state of the water. This unit continuously acquires turbidity, pH, and temperature readings of the influent at a preset time frequency, such as per second or per minute. These discrete raw measurements are integrated into a structured dataset, namely the water quality parameter set. This water quality parameter set provides key environmental variables for subsequent model analysis, as the turbidity, pH, and temperature of the water directly affect the flocculation effect and separation efficiency. Simultaneously, the equipment parameter acquisition unit is responsible for monitoring the operating status of the core component of the air flotation separation system, namely the bubble generator. The bubble generator is the key device for generating microbubbles for solid-liquid separation. This unit monitors the internal pressure and flow parameters in real time through pressure sensors and flow meters installed on the air or water inlet pipes of the bubble generator. These parameters directly determine the size, number, and distribution of the generated bubbles, and are core process parameters affecting the air flotation effect. Similar to the water quality parameter acquisition unit, this unit also combines pressure and flow measurements at specific times into a set of equipment parameters. Finally, the parameter processing unit serves as the center for data fusion and preprocessing. Logically connected to both the water quality parameter acquisition unit and the equipment parameter acquisition unit, it receives the time-synchronized water quality parameter sets and equipment parameter sets generated by these two units. The core task of this unit is to perform the combination processing, integrating parameters with different physical meanings into a unified data format. Specifically, it combines all parameters from the two parameter sets—influent turbidity, pH, temperature, bubble generator pressure, and flow rate—into a single, multi-dimensional feature vector. This vector comprehensively describes the overall operating conditions of the flotation separation system at a specific moment. Subsequently, to eliminate the potential adverse effects of differences in dimensions and numerical ranges between different parameters on model training, this unit standardizes or normalizes this combined feature vector. After processing, the generated data becomes real-time monitoring data. This data has a unified format and complete dimensions, and can be directly used as input to the big data model processing module, thus ensuring the accuracy and reliability of the model analysis.

[0029] Furthermore, the step of combining the water quality parameter set and the equipment parameter set to generate real-time monitoring data includes: The water quality parameter set and the equipment parameter set are integrated to obtain a comprehensive parameter set; The comprehensive parameter set is standardized to generate real-time monitoring data.

[0030] Specifically, the first step is integration processing. This step merges the water quality parameter set obtained by the water quality parameter acquisition unit with the equipment parameter set obtained by the equipment parameter acquisition unit at the same sampling time. For example, at a specific time... The system collected influent turbidity, pH, temperature, and pressure and flow parameters of the bubble generator. Integration processing combines these time-aligned heterogeneous data into a multi-dimensional comprehensive parameter set. This comprehensive parameter set, in vector or record form, comprehensively describes the overall operating condition of the air flotation separation system at that moment, providing a complete state snapshot for subsequent analysis. The second step is standardization. Since the various parameters in the comprehensive parameter set, such as turbidity, temperature, and pressure, have different physical dimensions and numerical ranges, directly inputting them into a large data model may lead to parameters with larger numerical ranges dominating the model training, thus affecting the model's accuracy. To eliminate this negative impact caused by differences in dimensions and numerical scales, the comprehensive parameter set needs to be standardized. This invention uses the Z-score standardization method, and its calculation process is as follows: , in, Represents the original measured value of a parameter in the comprehensive parameter set; This represents the statistical average value of the parameter in historical operation data; This represents the statistical standard deviation of the parameter in historical operational data; These are dimensionless values ​​obtained after standardization. and These are statistical characteristics pre-calculated based on a large amount of historical operational data accumulated over a long period of system operation. They reflect the central tendency and dispersion of the parameter, respectively. By performing this standardization calculation on each parameter in the comprehensive parameter set, the final generated set of all parameters is obtained. The collection of data constitutes the real-time monitoring data.

[0031] Furthermore, the big data model processing module includes: a data acquisition unit, a predictive analysis unit, a parameter optimization unit, and an update processing unit; wherein, The data acquisition unit is used to acquire historical operation data; The predictive analysis unit is connected to the data acquisition unit and the parameter processing unit respectively, and is used to perform multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results. The parameter optimization unit is connected to the prediction analysis unit and is used to calculate the optimal bubble size, reagent dosage and airflow rate based on the intermediate prediction results to generate preliminary optimization parameters. The update processing unit is connected to the parameter optimization unit and is used to update the preliminary optimization parameters and generate an optimization parameter set.

[0032] Specifically, the data acquisition unit is the module's data input port. It is responsible for retrieving and acquiring historical operational data from the system database or historical archives. This historical operational data is a multi-dimensional time-series dataset, comprehensively recording the operating status of the air flotation separation system at different points in the past, including but not limited to the influent water quality parameter set, equipment parameter set, control parameters set by the operator, and the corresponding effluent water quality results. This data forms the basis for building and training the predictive model. Next, the predictive analysis unit executes the core intelligent predictive function. This unit is logically connected to both the parameter processing unit of the data acquisition module and the data acquisition unit of this module, enabling it to simultaneously receive real-time monitoring data and historical operational data. Its core task is to perform multivariate analysis using big data analytics. In implementation, this unit pre-builds and trains one or more predictive models, such as a deep learning-based neural network model or an ensemble learning-based gradient boosting decision tree model. During the training phase, the model learns using water quality parameters and equipment parameters from the historical operational data as input features and the corresponding effluent quality as the output label, thereby mastering the complex nonlinear mapping relationship between the system input and output. During the operation phase, this unit inputs the latest real-time monitoring data into a pre-trained model. Based on patterns learned from historical data, the model predicts the separation effect under current conditions if operations are maintained. This prediction result is the intermediate prediction result. The intermediate prediction result can be one or more quantitative indicators, such as predicted effluent turbidity or suspended solids removal rate, which quantifies the future performance of the system. Subsequently, the parameter optimization unit performs decision optimization based on the intermediate prediction results provided by the predictive analysis unit. The goal of this unit is to calculate a set of control parameters that optimize the system's separation effect. It integrates an optimization algorithm, such as a genetic algorithm or simulated annealing algorithm. This algorithm uses key process parameters such as optimal bubble size, reagent dosage, and airflow rate as optimization variables, with a preset objective function as the optimization objective. For example, the objective function can be set to minimize the predicted effluent turbidity while considering the consumption costs of reagents and energy. The optimization algorithm iteratively searches for different parameter combinations within a set feasible range and uses the predictive model to evaluate the intermediate prediction results under each combination, ultimately finding the parameter combination that optimizes the objective function. This set of parameters is the preliminary optimization parameter. Finally, the update processing unit performs a final verification and correction of the preliminary optimized parameters generated by the parameter optimization unit. Since the optimization algorithm may produce theoretically optimal parameters that are infeasible or unstable in actual operation, this unit adds a layer of safety and realistic constraints. It may fine-tune or filter the preliminary optimized parameters based on preset equipment safe operating ranges, parameter change rate limits, and expert experience rule bases to ensure they are stable, safe, and conform to actual operating conditions. For example, it prevents control parameters from undergoing drastic changes within a short period.After this processing step, the final combination of parameters generated is the optimized parameter set.

[0033] Furthermore, the step of performing multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results includes: Multivariate analysis was performed on the real-time monitoring data and the historical operation data to obtain the flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient. Intermediate prediction results are generated using the flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient.

[0034] Specifically, the predictive analysis unit first performs multivariate analysis on the received real-time monitoring data and historical operational data. This analysis doesn't simply input the data directly into a black-box model; instead, it extracts more physically meaningful characteristic variables from the raw data using pre-established mechanistic models or empirical formulas. Specifically, it calculates the following three core indicators. The first is the flocculation index. The flocculation index is a quantitative indicator evaluating the effect of suspended particles in water forming flocs after chemical dosing. It comprehensively considers factors such as influent turbidity, pH, temperature, and the type and amount of chemical added in historical data. Its calculation model can be established through regression analysis of experimental data or based on flocculation kinetics theory, aiming to quantify the size, density, and strength of flocs under current water quality conditions. A higher flocculation index indicates a good flocculation effect. The second is the gas-solid binding probability factor. This factor describes the probability that microbubbles and flocs will meet and successfully attach in water. It is mainly affected by multiple variables such as bubble size, floc characteristics, and hydraulic conditions. This factor can be derived through computational fluid dynamics simulation or based on statistical mechanics models, and its value directly relates to whether flocs can be effectively captured by bubbles and float. The third is the scum layer stability coefficient. This coefficient characterizes the stability of the scum layer formed by the floc-bubble aggregates floating to the water surface. A stable scum layer is less prone to breakage or re-entry into the water, which is beneficial for subsequent scum removal operations. The calculation of this coefficient comprehensively considers the rising water velocity determined by the pressure and flow parameters of the bubble generator, as well as the characteristics of the flocs, aiming to assess the scum layer's ability to resist hydraulic shear and disturbance. After calculating the three key intermediate indicators—flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient—the predictive analysis unit uses them as enhancing features, combining them with other parameters in the original real-time monitoring data, such as the original turbidity and pH value, to form a richer feature set. Then, the unit uses this enhanced feature set to generate the final intermediate prediction result through a pre-trained comprehensive prediction model, such as support vector machine, random forest, or neural network. This intermediate prediction result directly corresponds to the core performance indicators of the air flotation separation system, such as the predicted effluent turbidity value. Because the input feature set not only includes apparent operating data but also incorporates in-depth indicators that reflect the underlying mechanisms, the model is able to more accurately capture the causal relationship between system state and separation effect.

[0035] Furthermore, the workflow of the real-time control module is as follows: Figure 3 As shown, the real-time control module includes: a parameter setting unit, an execution control unit, and a feedback adjustment unit; wherein, The parameter setting unit is connected to the update processing unit and is used to set the target pressure value of the bubble generator and the target rate value of the reagent dosing pump according to the optimized parameter set, and generate a control command set. The execution control unit is connected to the parameter setting unit and is used to execute the control instruction set and adjust the control parameters of the air flotation separation system. The feedback adjustment unit is connected to the execution control unit and is used to collect effluent quality parameters in real time, and perform optimized separation operation in combination with the adjusted control parameters.

[0036] Specifically, the parameter setting unit first serves as the source of control command generation. It connects directly to the update processing unit in the big data model processing module, receiving its output set of optimized parameters. This set includes theoretically optimal process objectives such as the best bubble size, reagent dosage, and airflow rate under the current operating conditions. The parameter setting unit's task is to translate these abstract process objectives into control setpoints for specific equipment. For example, based on known equipment characteristic curves, it converts the objectives of "optimal bubble size" and "optimal airflow rate" into specific target pressure values ​​for the bubble generator and valve openings for the flow controller; simultaneously, it converts "optimal reagent dosage" into target rate values ​​or frequency setpoints for the reagent dosing pump. After conversion, these specific equipment setpoints are organized into a structured set of control commands, ready to be issued to the actuators. Next, the execution control unit is responsible for the physical implementation of the commands. This unit connects to the parameter setting unit, receiving and parsing the control command set. Internally, it integrates communication interfaces and drive logic with field equipment, such as the frequency converter of the bubble generator, pressure regulating valves, and the driver of the reagent dosing pump. Upon receiving the instruction, the execution control unit immediately sends control signals to the corresponding equipment, driving them to adjust their operating status until their actual operating parameters, such as actual pressure and actual pump speed, match the target pressure and target speed values ​​set in the control instruction set. Through this series of actions, the unit accurately applies the optimization decisions of the big data model to the physical equipment of the air flotation separation system, thereby adjusting the overall control parameters of the air flotation separation system. Finally, the feedback adjustment unit forms the closed loop of the control loop, ensuring that the system can cope with prediction model deviations and unmodeled dynamics. This unit is logically connected to the execution control unit and deploys online water quality monitoring sensors at the outlet of the air flotation separation system. After the execution control unit completes the parameter adjustment, the feedback adjustment unit begins to collect effluent quality parameters in real time, especially effluent turbidity. It compares these measured effluent quality values ​​with preset control targets, such as national or industry emission standards. If there is a deviation between the measured value and the target value, the unit activates a feedback controller to calculate the correction amount for the current control parameters based on the magnitude and direction of the deviation. Then, it adds this correction to the optimized setpoint given by the big data model, forming a new, real-time corrected control command, which drives the execution control unit to make fine adjustments again. This process is repeated continuously, combining the adjusted control parameters to continuously perform optimization separation operations, ensuring that even if the model prediction is not completely accurate or there are external disturbances, the effluent water quality can be stably maintained at the desired level.

[0037] Furthermore, the real-time acquisition of effluent quality parameters, combined with the adjusted control parameters, to perform optimized separation operations includes: Monitor the turbidity value of the effluent and generate effluent monitoring data; Obtain a quality standard for measuring the quality of effluent, compare the effluent monitoring data with the quality standard, and generate a deviation index. The deviation index is subjected to feedback adjustment processing to generate parameter correction instructions; Based on the parameter correction instructions, the target pressure value of the bubble generator and the target rate value of the reagent dosing pump are adjusted to perform optimized separation operation.

[0038] Specifically, the first step is data acquisition, which involves monitoring the turbidity value of the effluent and generating effluent monitoring data. The feedback control unit continuously collects turbidity readings of the effluent through an online turbidity sensor installed at the outlet of the air flotation separation system. These real-time measurements form the basis of closed-loop control, accurately reflecting the immediate treatment effect after optimized separation operations. The second step is deviation calculation. The feedback control unit first obtains the quality standard used to measure the effluent quality from the system configuration. This quality standard is usually a specific turbidity limit, such as the upper limit of effluent turbidity set according to environmental regulations or production process requirements. Then, the unit compares the real-time generated effluent monitoring data with this preset quality standard and calculates the difference between the two. This difference is the deviation index. The formula for calculating the deviation index can be expressed as: , in It is a deviation indicator; It is the preset quality standard target turbidity value; This refers to the actual turbidity value of the effluent, measured by sensors. Deviation index. The size and sign of the symbol visually reflect the gap between the current control effect and the expected target. A positive value indicates that the effluent water quality is better than the standard, while a negative value indicates that it is worse than the standard. Next, correction commands are generated. The feedback control unit processes the calculated deviation index using feedback control. This processing is usually performed by a built-in controller, most commonly a proportional-integral-derivative controller. The controller adjusts the current deviation accordingly. The system calculates a specific control adjustment amount based on the accumulation of historical deviations and the rate of change of these deviations. This adjustment amount aims to eliminate the deviation and bring the effluent turbidity back to near the quality standard. This adjustment amount output by the controller is the parameter correction command. Finally, the correction and optimization are executed. The parameter correction command is sent to the execution control unit to adjust the target pressure value of the bubble generator and the target rate value of the reagent dosing pump. Specifically, the parameter correction command is added as an increment to the target value previously given by the big data model. For example, if the deviation index is negative, the parameter correction command may instruct a slight increase in the target rate value of the reagent dosing pump to enhance flocculation, or adjust the target pressure value of the bubble generator to optimize bubble characteristics. Based on this corrected, new target value, the execution control unit drives the relevant equipment for fine-tuning again. This cycle of adjustment-monitoring-comparison-readjustment constitutes a continuous optimization separation operation, ensuring that the system always operates around the optimal performance point.

[0039] Furthermore, the monitoring of effluent turbidity values ​​and the generation of effluent monitoring data include: The turbidity value of the effluent is continuously and periodically collected to obtain the original turbidity sampling value sequence; The original turbidity sampling value sequence is subjected to sliding window mean filtering and data standardization to generate effluent monitoring data.

[0040] Specifically, the first step is data acquisition. An online turbidity sensor deployed at the outlet of the air flotation separation system continuously samples the effluent at a preset, fixed time interval, such as once per second. Each sample generates an instantaneous turbidity measurement. Over time, these chronologically ordered instantaneous measurements form a time series, i.e., the raw turbidity sampling value sequence. While this sequence reflects the dynamic changes in effluent turbidity, it typically includes random noise caused by minor fluctuations in water flow, bubble interference, or the sensor's own electrical characteristics. The second stage is data processing, which includes two steps: sliding window mean filtering and data standardization. First, sliding window mean filtering is performed to smooth the raw data and eliminate high-frequency noise. This method defines a fixed-size "window" that contains the most recent... Each initial turbidity sample value. At each new sampling time, calculate all values ​​within this window. The arithmetic mean of the sampled values ​​is used as the filtered turbidity value at the current moment. As new data arrives, the window "slides" forward one step on the time axis, incorporating the latest sampled values ​​and discarding the oldest, then recalculating the average. The calculation process can be represented as follows: , in At the current moment The resulting turbidity value after filtering; It is the size of the sliding window, which is a preset integer whose value is determined based on the actual process response time and noise level; It is in the past moment The raw turbidity samples are collected. Next, the filtered turbidity value sequence undergoes data standardization to eliminate the influence of dimensions and make it numerically comparable to other standardized system parameters. This step typically uses the Z-score standardization method to transform the filtered values ​​into a distribution with a mean of 0 and a standard deviation of 1. After these two steps, the final smoothed time series data is the effluent monitoring data that can be directly used by the feedback control unit.

[0041] Furthermore, updating the initial optimization parameters and generating the optimization parameter set includes: The initial optimization parameters are updated to obtain the corrected optimization parameters; The modified and optimized parameters are used to perform operating condition adaptation verification and generate an optimized parameter set.

[0042] Specifically, the first step is to update the initial optimization parameters. This step can be understood as a correction based on real-world constraints. The initial optimization parameters are calculated by the optimization algorithm under an ideal model environment, and may not fully consider equipment mechanical performance limitations, process response delays, or operational safety procedures. Therefore, the update processing unit reviews and adjusts the initial optimization parameters according to a series of preset rules. These rules may include limiting the rate of parameter change to prevent drastic changes in control quantities from impacting the equipment; imposing hard constraints on the parameter value range to ensure it always remains within the equipment's safe operating range; and logical judgments based on expert knowledge bases. For example, under certain water quality conditions, even if the model suggests reducing the dosage, the system will maintain a minimum dosage to ensure the robustness of the treatment effect. After this series of rule-based corrections, the resulting parameter set is the corrected optimization parameters. Next, the obtained corrected optimization parameters are used for operational condition adaptation verification. This is a simulation or pre-running step to evaluate the dynamic response that the corrected parameters may produce when applied to the actual system. This verification can be performed using a simplified transfer function model describing the dynamic behavior of the system or a digital twin simulation environment. Using the corrected and optimized parameters as input, the system's state evolution is simulated over the next control cycle to observe whether its response is stable and whether there are any unstable phenomena such as overshoot or oscillation. Simultaneously, this verification also checks the synergy between parameters to ensure that adjustments such as airflow rate and reagent dosage are logically matched and do not produce conflicting control effects. If the operating condition adaptation verification passes, meaning the simulation results show a good system response, then this set of corrected and optimized parameters is confirmed as the final optimized parameter set and prepared for output to the downstream real-time control module. If the verification fails, the system may backtrack, fine-tune the correction rules, and perform verification again, or make small adjustments to the corrected and optimized parameters within permissible limits until the verification passes.

[0043] To verify the feasibility of this invention in practice, it was applied to a wastewater treatment plant in a large petrochemical industrial park. This wastewater treatment plant is responsible for treating oily wastewater generated during the production process. The influent water quality is characterized by high turbidity, complex composition, and drastic fluctuations. Traditional air flotation separation systems mainly rely on manual experience to set parameters such as reagent dosage and flotation pressure, which is difficult to adapt to rapid changes in influent water quality, often leading to unstable effluent quality, reagent waste, and excessive energy consumption. This wastewater treatment plant aims to use this invention to achieve intelligent and automated control of the air flotation separation process, thereby improving treatment efficiency and operational economy.

[0044] In this embodiment, the wastewater treatment plant deploys the air flotation separation system based on a big data model as described in this invention. First, an online turbidity meter, pH meter, and temperature sensor are installed at the inlet of the air flotation tank to form a water quality parameter acquisition unit. Simultaneously, pressure and flow sensors are installed on the microbubble generator as equipment parameter acquisition units. The data collected by these sensors is integrated and Z-score standardized by the parameter processing unit to generate real-time monitoring data. The big data model processing module pre-loads historical operating data from the past two years of the treatment plant, and its internal predictive analysis unit, based on a deep neural network model, processes the real-time monitoring data. The real-time control module is directly connected to the frequency converters of the on-site reagent dosing pump and the bubble generator, and is responsible for executing optimization commands.

[0045] To verify the effectiveness of the present invention, the system underwent a three-month continuous operation test from July to September 2024, and its performance was compared with that of a control group using a traditional fixed parameter control method.

[0046] In a typical scenario of fluctuating operating conditions, the data acquisition module detected a rapid increase in influent turbidity from the usual 150 NTU to 450 NTU within a short period. The system responded immediately. Upon receiving this real-time monitoring data, the big data model processing module's predictive analysis unit quickly calculated the flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient. Analysis results indicated that maintaining the current reagent dosage of 20 mg / L and the bubble generator pressure of 0.3 MPa would result in intermediate predictions showing that the turbidity would far exceed the 50 NTU discharge standard. Subsequently, the parameter optimization unit was activated, using a genetic algorithm to calculate preliminary optimized parameters, recommending increasing the reagent dosage to 45 mg / L and the bubble generator pressure to 0.45 MPa. The update processing unit then performed operating condition adaptation verification on these preliminary parameters, generating a smoothly transitioning optimized parameter set, which was then sent to the real-time control module.

[0047] The parameter setting unit of the real-time control module converts the optimized parameter set into the target rate value of the reagent dosing pump and the target pressure value of the bubble generator, which is then precisely executed by the execution control unit. Within 5 minutes of parameter adjustment, the feedback adjustment unit generates effluent monitoring data by applying a sliding window mean filter to the original turbidity sampling value sequence at the outlet. The data shows that the effluent turbidity is stable at around 48 NTU. The system compares this data with the internal quality standard of 45 NTU, generating a deviation index of -3 NTU. The feedback adjustment unit processes this deviation index, generates a parameter correction command, and fine-tunes the target pressure value of the bubble generator to 0.46 MPa. After this round of feedback adjustment, the effluent turbidity finally stabilizes at 44 NTU, achieving a rapid and accurate response to sudden high load shocks.

[0048] Data shows that the wastewater treatment plant using the system of this invention exhibits adaptive capabilities in response to fluctuations in influent water quality. Compared to traditional control methods, which suffer from effluent exceeding standards and subsequent excessive chemical dosing due to response lag, this invention not only ensures the continuous and stable compliance of effluent quality but also achieves savings in reagents and energy consumption.

[0049] It should be noted that the electrical connections between the various units described above do not necessarily represent direct or indirect connections. Any indirect connection method can be applied to the embodiments of the present invention as long as it achieves the purpose of the present invention. The above descriptions are merely exemplary embodiments of the present invention and should not be construed as limiting the scope of the present invention.

[0050] All equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of other embodiments of this invention upon considering the specification and the disclosure of practical truth. This application is intended to cover any variations, uses, or adaptations of this invention that follow the general principles of this invention and include common knowledge or conventional techniques in the art not described herein.

Claims

1. An air flotation separation system based on a big data model, characterized in that, The system includes: a data acquisition module, a big data model processing module, and a real-time control module; wherein... The data acquisition module is used to acquire the influent water quality parameters and equipment operating parameters of the air flotation separation system and generate real-time monitoring data; The big data model processing module is used to acquire historical operation data, perform predictive processing on the real-time monitoring data and the historical operation data, and generate an optimized parameter set. The real-time control module is used to adjust the control parameters of the air flotation separation system according to the optimized parameter set and to perform optimized separation operations.

2. The air flotation separation system based on a big data model according to claim 1, characterized in that, The data acquisition module includes: a water quality parameter acquisition unit, an equipment parameter acquisition unit, and a parameter processing unit; wherein... The water quality parameter acquisition unit is used to monitor the turbidity, pH value and temperature of the influent and generate a water quality parameter set; The equipment parameter acquisition unit is used to monitor the pressure and flow parameters of the bubble generator and generate a set of equipment parameters; The parameter processing unit is connected to both the water quality parameter acquisition unit and the equipment parameter acquisition unit, and is used to combine the water quality parameter set and the equipment parameter set to generate real-time monitoring data.

3. The air flotation separation system based on a big data model according to claim 2, characterized in that, The process of generating real-time monitoring data by combining the water quality parameter set and the equipment parameter set includes: The water quality parameter set and the equipment parameter set are integrated to obtain a comprehensive parameter set; The comprehensive parameter set is standardized to generate real-time monitoring data.

4. The air flotation separation system based on a big data model according to claim 2, characterized in that, The big data model processing module includes: a data acquisition unit, a predictive analysis unit, a parameter optimization unit, and an update processing unit; wherein... The data acquisition unit is used to acquire historical operation data; The predictive analysis unit is connected to the data acquisition unit and the parameter processing unit respectively, and is used to perform multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results. The parameter optimization unit is connected to the prediction analysis unit and is used to calculate the optimal bubble size, reagent dosage and airflow rate based on the intermediate prediction results to generate preliminary optimization parameters. The update processing unit is connected to the parameter optimization unit and is used to update the preliminary optimization parameters and generate an optimization parameter set.

5. The air flotation separation system based on a big data model according to claim 4, characterized in that, The step of performing multivariate analysis on the real-time monitoring data and the historical operation data to generate intermediate prediction results includes: Multivariate analysis was performed on the real-time monitoring data and the historical operation data to obtain the flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient. Intermediate prediction results are generated using the flocculation index, gas-solid bonding probability factor, and scum layer stability coefficient.

6. The air flotation separation system based on a big data model according to claim 4, characterized in that, The real-time control module includes: a parameter setting unit, an execution control unit, and a feedback adjustment unit; wherein... The parameter setting unit is connected to the update processing unit and is used to set the target pressure value of the bubble generator and the target rate value of the reagent dosing pump according to the optimized parameter set, and generate a control command set. The execution control unit is connected to the parameter setting unit and is used to execute the control instruction set and adjust the control parameters of the air flotation separation system. The feedback adjustment unit is connected to the execution control unit and is used to collect effluent quality parameters in real time, and perform optimized separation operation in combination with the adjusted control parameters.

7. The air flotation separation system based on a big data model according to claim 6, characterized in that, The real-time acquisition of effluent quality parameters, combined with the adjusted control parameters, and the execution of optimized separation operations include: Monitor the turbidity value of the effluent and generate effluent monitoring data; Obtain a quality standard for measuring the quality of effluent, compare the effluent monitoring data with the quality standard, and generate a deviation index. The deviation index is subjected to feedback adjustment processing to generate parameter correction instructions; Based on the parameter correction instructions, the target pressure value of the bubble generator and the target rate value of the reagent dosing pump are adjusted to perform optimized separation operation.

8. The air flotation separation system based on a big data model according to claim 7, characterized in that, The monitoring of effluent turbidity values ​​generates effluent monitoring data including: The turbidity value of the effluent is continuously and periodically collected to obtain the original turbidity sampling value sequence; The original turbidity sampling value sequence is subjected to sliding window mean filtering and data standardization to generate effluent monitoring data.

9. The air flotation separation system based on a big data model according to claim 4, characterized in that, The step of updating the preliminary optimization parameters and generating an optimization parameter set includes: The initial optimization parameters are updated to obtain the corrected optimization parameters; The modified and optimized parameters are used to perform operating condition adaptation verification and generate an optimized parameter set.

Images