A Control Method for Industrial Wastewater Aeration Systems Based on Multi-Source Data Fusion

Abstract

This invention relates to the field of automatic control technology, specifically to a control method for an industrial wastewater aeration system based on multi-source data fusion. The method includes: acquiring multi-source operating condition data collected from multiple preset monitoring points in the aerobic tank; fusing the multi-source operating condition data to construct an operating condition feature vector that characterizes the overall operating state of the current system; optimizing the control parameters of the aeration system using a differential evolution algorithm, and guiding the spatial rotation of the randomly generated differential vector using the operating condition feature vector during mutation operations; and controlling the aeration system according to the optimized control parameters. The solution of this invention enables the control system to respond quickly and proactively to changes in operating conditions, significantly improving the robustness of control and energy-saving effects.

Description

Technical Field

[0001] This invention relates to the field of automatic control technology, and more specifically, to a control method for an industrial wastewater aeration system based on multi-source data fusion. Background Technology

[0002] Within the framework of sustainable development in modern industry, efficient treatment of industrial wastewater is an indispensable key link. Activated sludge process, as a mainstream treatment technology, relies on an aeration system to provide sufficient dissolved oxygen for aerobic microorganisms to degrade pollutants. However, the aeration system is the largest energy-consuming unit in a wastewater treatment plant, typically accounting for more than 50% of the total energy consumption. Currently, the commonly used control method in the industry is single-parameter feedback control based on online dissolved oxygen sensors. This involves a PID controller passively adjusting the blower volume based on the deviation between the measured dissolved oxygen value and a fixed setpoint. This method can basically meet the requirements when treating municipal wastewater with stable loads, but its inherent limitations become apparent when dealing with the complex and variable industrial wastewater.

[0003] The core challenge in industrial wastewater treatment lies in the drastic, frequent, and unpredictable fluctuations in water quality and quantity. When upstream production processes are adjusted or malfunctions occur, and high concentrations of pollutants surge in within a short period, traditional PID control methods prove inadequate due to their inherent response lag. The controller only begins increasing the aeration rate after detecting a sharp drop in dissolved oxygen concentration; this adjustment is often too late, easily leading to partial or overall anoxic conditions in the aeration tank, severely inhibiting microbial activity, and even causing process problems such as sludge bulking. Furthermore, the aeration process is a complex, nonlinear, and time-varying system; fixed PID parameters cannot adapt to drastic changes in operating conditions, resulting in poor system robustness and difficulty in ensuring the continuous stability of effluent quality. Therefore, the key technical challenge facing existing technologies is: how to break through the passive and lagging single-parameter feedback control mode and establish an intelligent optimization control method that can actively sense and quickly adapt to shock loads, achieving precise matching between aeration volume and real-time oxygen demand, thereby fundamentally solving the problems of slow control response, poor adaptability, and insufficient control robustness. Summary of the Invention

[0004] The purpose of this invention is to propose an optimized control method and system for industrial wastewater aeration systems based on multi-source data fusion, in order to solve the problems of slow response and poor control robustness of existing aeration control methods when dealing with drastic fluctuations in water quality and quantity.

[0005] To address this, the present invention provides a control method for an industrial wastewater aeration system based on multi-source data fusion. The control method includes: acquiring multi-source operating condition data collected from multiple preset monitoring points in the aerobic tank; standardizing the multi-source operating condition data to obtain a condition feature vector characterizing the overall operating state of the current aeration system; executing an improved differential evolution algorithm to search for and determine optimal control parameters with the goal of minimizing the comprehensive cost function; generating control commands based on the optimal control parameters to drive the aeration system to complete the aeration control task. The improved differential evolution algorithm includes: each individual in the population contains a control parameter gene and a rotation angle gene; the mutation operation of the differential evolution algorithm includes: selecting at least two individuals from the differential evolution algorithm population, generating an initial difference vector by performing vector subtraction on the two individuals; rotating the initial difference vector by an angle determined by the rotation angle gene in the individual to obtain a guiding difference vector; and generating mutated individuals using the guiding difference vector.

[0006] This invention achieves intelligent and precise control of industrial wastewater aeration systems by integrating multi-source operating condition data and employing an improved differential evolution algorithm, effectively reducing aeration energy consumption and operating costs while improving the stability and treatment efficiency of the wastewater treatment process.

[0007] Optionally, the multi-source operating condition data includes: influent chemical oxygen demand concentration, influent ammonia nitrogen concentration, influent flow rate, dissolved oxygen concentration at the front end of the aerobic tank, dissolved oxygen concentration in the middle section of the aerobic tank, dissolved oxygen concentration at the end of the aerobic tank, and pH value.

[0008] Optionally, the preprocessing steps of the multi-source operating condition data include: collecting multi-source operating condition data within the most recent N time windows; removing outliers based on the 3σ criterion, whereby an outlier is identified when the difference between the collected data point and the historical data mean is greater than 3 times the standard deviation and replaced by the previous valid data point; and standardizing the preprocessed data.

[0009] Preprocessing multi-source operating condition data can effectively eliminate the effects of sensor noise, instantaneous interference, and outliers, thereby improving data quality and reliability.

[0010] Optionally, the steps for constructing the operating condition feature vector include: calculating the rate of change of influent chemical oxygen demand concentration and influent flow rate to obtain the time rate of change feature; mapping each index and the time rate of change feature to the [0,1] interval using the maximum-minimum normalization formula; and constructing a multidimensional operating condition feature vector by combining multiple normalized index values ​​in a predetermined order.

[0011] Optionally, the step of constructing the time change rate feature includes: the time change rate feature includes the change rate of influent chemical oxygen demand concentration over the past N minutes and the change rate of influent flow rate over the past N minutes, and these two change rates are obtained by calculating the difference between the current time and N minutes ago.

[0012] By leveraging the time rate of change characteristic, the hysteresis characteristic of traditional passive feedback control is fundamentally altered, enabling the system to possess proactive adaptive capabilities.

[0013] Optionally, the differential evolution algorithm includes: performing mutation operations using the DE / rand / 1 strategy, and implementing spatial rotation transformation of the initial difference vector using the Rodrigues rotation formula.

[0014] Optionally, the spatial rotation operation is implemented through the following steps: performing a vector outer product operation to determine a rotation axis that is simultaneously perpendicular to the initial difference vector and the working condition feature vector; normalizing the rotation axis vector; decomposing the initial difference vector to be rotated into a parallel component parallel to the rotation axis and a perpendicular component perpendicular to the rotation axis; keeping the magnitude and direction of the parallel component unchanged, rotating the perpendicular component according to a preset rotation angle; and adding the unchanged parallel component to the rotated perpendicular component to obtain the final guiding difference vector.

[0015] The improved differential evolution algorithm not only retains the global optimization capability, but more importantly, it couples its search direction with the actual needs of the physical system, which greatly improves the convergence speed and the quality of the solution.

[0016] Optionally, the calculation step of the rotation angle includes: taking the rotation angle and the blower output frequency in the spatial rotation operation as individual genes, and using an improved differential evolution algorithm for adaptive optimization to obtain the rotation angle.

[0017] Optionally, the comprehensive cost function includes a weighted combination of an operating cost term, a dissolved oxygen concentration deviation term, and a system safety assurance term; wherein, the operating cost term is calculated using a pre-established cubic polynomial model of blower frequency-power; the dissolved oxygen concentration deviation term is obtained by calculating the difference between the dissolved oxygen concentration at the front end and the dissolved oxygen concentration at the end of the aerobic pool measured at the current moment, and obtaining the absolute difference between this difference and a preset value; the system safety assurance term is a boundary constraint penalty term, which is obtained by assigning a maximum penalty value when any preset monitoring indicator exceeds its safety boundary.

[0018] Optionally, the triggering conditions for the boundary constraint penalty include: when the dissolved oxygen concentration at the front end of the aerobic tank is lower than the minimum dissolved oxygen concentration and a preset threshold set to ensure the survival of microorganisms, a penalty is triggered; when the dissolved oxygen concentration at the end of the aerobic tank is higher than the preset threshold set for the maximum dissolved oxygen concentration, a penalty is triggered; when the pH value in the aerobic tank is not within the suitable pH range set for microorganisms, a penalty is triggered.

[0019] Optionally, during the execution of the aeration control task, multi-source operating condition data are continuously monitored. When the change in the time rate of change characteristic exceeds a preset threshold, that is, the difference between the reconstructed feature vector and the original feature vector exceeds a preset threshold, the improved differential evolution algorithm is restarted for online optimization, and the aeration system is driven to complete the control task based on the optimized new control parameters.

[0020] The online optimization mechanism enables the system to dynamically adapt to changes in operating conditions and adjust control strategies in a timely manner, avoiding the efficiency decline caused by long-term fixed control parameters and significantly improving the system's adaptability and robustness.

[0021] The beneficial effects of this invention are as follows: The solution of this invention deeply integrates real-time multi-source operating condition data perception with an improved differential evolution algorithm. By constructing operating condition feature vectors, it achieves a comprehensive representation of the overall system operating state. Through the dynamic guidance of the algorithm's mutation direction via these feature vectors, it achieves deep coupling between the search mechanism and the physical system, thereby realizing condition-aware intelligent optimization of aeration system control parameters. This method solves the problems of oxygen deficiency caused by response lag in traditional methods, low energy efficiency due to fixed parameters, and unstable treatment effects due to a lack of foresight. Simultaneously, by continuously monitoring changes in operating conditions and performing online re-optimization, it further enhances the system's adaptive capability to dynamic operating conditions. The solution of this invention enables the control system to respond quickly and proactively to changes in operating conditions, significantly improving control robustness and energy-saving effects. Attached Figure Description

[0022] Figure 1 This is a flowchart of a control method for an industrial wastewater aeration system based on multi-source data fusion according to an embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will now be clearly and completely described with reference to the accompanying drawings. Figure 1 The diagram shown is a flowchart of a control method for an industrial wastewater aeration system based on multi-source data fusion according to an embodiment of the present invention.

[0024] This invention provides a control method for an industrial wastewater aeration system based on multi-source data fusion, applicable to a system employing A... 2The plant is a comprehensive wastewater treatment plant in a chemical industrial park using the / O process. Its aeration system uses a variable frequency centrifugal blower to aerate the aerobic tanks, and the total aeration volume is controlled by adjusting the output frequency of the blower's frequency converter.

[0025] This embodiment of the method for controlling an industrial wastewater aeration system based on multi-source data fusion includes the following steps:

[0026] Step S1: Obtain multi-source operating condition data collected from multiple preset monitoring points in the aerobic tank.

[0027] Multi-source operating condition data is collected by sensors at multiple preset monitoring points in the aerobic tank. The multi-source operating condition data includes: influent chemical oxygen demand concentration, influent ammonia nitrogen concentration, influent flow rate, dissolved oxygen concentration at the front end of the aerobic tank, dissolved oxygen concentration in the middle section of the aerobic tank, dissolved oxygen concentration at the end of the aerobic tank, and pH value.

[0028] Specifically, the influent chemical oxygen demand (COD) concentration is obtained through an online analytical instrument installed at the influent, in mg / L; the influent ammonia nitrogen concentration is monitored in real time through an online analytical instrument, in mg / L; and the influent flow rate is measured by a flow sensor, in m³ / s. 3 / h; Dissolved oxygen concentrations at the front, middle, and end of the aerobic tank are collected by dissolved oxygen sensors, in mg / L; pH values ​​are detected by pH sensors, in dimensionless units. All sensor data are aggregated via an industrial fieldbus and uploaded to the lower-level controller for preprocessing and storage. The data acquisition frequency is set to 1Hz, and the sampling period is 1 minute to ensure that the control system can promptly detect changes in operating conditions. The system maintains a sliding data window containing the most recent 300 data points for subsequent construction of operating condition feature vectors.

[0029] Step S2: After standardizing the multi-source operating condition data, a condition feature vector representing the overall operating status of the current aeration system is obtained.

[0030] First, the collected multi-source operating condition data is preprocessed. The preprocessing includes outlier removal and standardization. Outlier removal is based on the 3σ criterion; the mean and standard deviation of each multi-source operating condition data point are calculated based on historical data; if the difference between each multi-source operating condition data point and the mean is greater than three times the standard deviation, the data point is considered an outlier, discarded, and replaced with multi-source operating condition data from the previous valid sampling time; the preprocessed data is then standardized to ensure uniformity of measurement and comparability.

[0031] Secondly, construct the operating condition feature vector. The construction of the operating condition feature vector involves three main steps:

[0032] i. Calculate the time-varying rate characteristics; calculate the time-varying rate of change of two preset key load indicators within the past 5-minute time window. The preset key load indicators are the influent chemical oxygen demand concentration and the influent flow rate. The time-varying rate is obtained by calculating the difference between the effective value at the current moment and the effective value 5 minutes ago.

[0033] ii. Data normalization: Nine indicators were selected, including chemical oxygen demand (COD) concentration, ammonia nitrogen concentration, flow rate, COD concentration over time, influent flow rate over time, upstream dissolved oxygen concentration, midstream dissolved oxygen concentration, downstream dissolved oxygen concentration, and pH value. Based on long-term historical operating data, the maximum and minimum values ​​of each indicator were statistically determined. Then, the current value x of each indicator was linearly mapped to the [0,1] interval using the maximum-minimum normalization formula.

[0034] iii. Vector synthesis; The above nine normalized index values ​​are constructed into a nine-dimensional working condition feature vector in a predetermined order. The direction and magnitude of this vector in 9-dimensional space uniquely characterize the current operating condition.

[0035] Step S3: Execute the improved differential evolution algorithm to search for and determine the optimal control parameters with the goal of minimizing the overall cost function.

[0036] First, define a fitness function to evaluate the quality of each individual. The calculation formula is as follows:

[0037] ;

[0038] in, This is the operating cost item, represented by instantaneous operating power in this embodiment, with the unit being kW; This is the dissolved oxygen concentration deviation item, in mg / L; This is a penalty item, and the unit is dimensionless. The preset maximum operating power, in kW; This is the preset baseline deviation value, in mg / L; , These are dimensionless weighting coefficients; therefore, Divide by and will Divide by Normalization is performed to make all terms on the right-hand side of the formula dimensionless, thus achieving a weighted summation of indices with different physical dimensions. This is a dimensionless fitness value.

[0039] First item: Operating costs The calculation of the applied frequency is performed using a pre-established cubic polynomial model of the blower's "frequency-power" relationship. The instantaneous operating power is used in this embodiment to quantify the operating cost item:

[0040] ;

[0041] Among them, coefficient , , , This value was obtained by performing least squares regression fitting on the measured power data of the blower operating at different frequencies. The value is entirely determined by the candidate frequencies. Confirmed, unit is Hz; where The unit is kW. The unit is kW / Hz 3 , The unit is kW / Hz 2 , The unit is kW / Hz. The unit is kW.

[0042] Second item: Dissolved oxygen concentration deviation This is obtained by calculating the difference between the dissolved oxygen concentration at the front end and the dissolved oxygen concentration at the end of the aerobic tank at the current moment, and then taking the absolute difference between this difference and a preset value. A small gradient deviation value means that the aeration rate meets the high oxygen demand at the front end while reserving an appropriate amount of oxygen for the deep treatment at the end, indicating a balanced aeration distribution; conversely, a large deviation value indicates an unreasonable aeration distribution, with problems such as oxygen deficiency at the front end or over-aeration at the end.

[0043] Third boundary constraint penalty item In this embodiment, the penalty value is 1000. The triggering conditions for this item are multi-dimensional. The penalty will take effect as long as any of the following conditions are met: The first case is to prevent severe hypoxia in the core area of ​​the aerobic tank, that is, when the dissolved oxygen concentration at the front end of the aerobic tank measured in real time is lower than the preset threshold for the minimum dissolved oxygen concentration; the second case is to prevent over-aeration at the end of the aerobic tank, that is, when the dissolved oxygen concentration at the end of the aerobic tank measured in real time is higher than the maximum preset threshold; the third case is to maintain the stability of the biochemical environment, that is, when the pH value in the aerobic tank measured in real time is not within the preset pH value range.

[0044] Then, the iterative optimization process of the differential evolution algorithm is executed, as follows:

[0045] a. Set the basic parameters of the differential evolution algorithm: population size NP = 20, maximum number of generations G = 50. Each individual is represented by a vector. The blower's output frequency is constrained within the range of [30Hz, 50Hz], and the spatial rotation angle ranges from [0, 2π]. Individuals are initialized by randomly generating initial values ​​within the constraints of each gene, ensuring sufficient diversity in the initial population, which is beneficial for global optimization search.

[0046] b. Mutation operation: Following the DE / rand / 1 strategy of the differential evolution algorithm, randomly select three other individuals from the population, denoted as... , and Among them, individuals The vector is directly designated as the basis vector. The other two individuals and The vector is then used to generate the initial difference vector. Its generation method is to perform vector subtraction operation, that is = - Due to individual and The initial difference vector is randomly selected from the population, therefore it is generated from the random selection. The direction is completely random, representing a random, exploratory search direction.

[0047] Obtain the working condition feature vector constructed in step S2 .

[0048] The initial difference vector Towards the working condition feature vector Rotate by an angle in the direction. This yields a guiding difference vector. .

[0049] In this embodiment, the rotation operation is implemented through a series of vector algebra operations based on the geometric principles of Rodrigues' rotation formula. The process first determines a unique rotation axis perpendicular to both vectors by performing a cross product operation on the initial difference vector and the operating condition characteristic vector, and then normalizes this rotation axis vector. Next, the rotation operation is accomplished by decomposing the initial difference vector to be rotated into a parallel component parallel to the unit rotation axis and a perpendicular component perpendicular to the unit rotation axis. During the rotation operation, the magnitude and direction of the parallel component remain unchanged, while the perpendicular component rotates within the plane formed by itself and the rotation axis according to a preset rotation angle. Finally, by adding the unchanged parallel component and the rotated perpendicular component, the final guided difference vector, whose direction has been successfully guided, is obtained.

[0050] In this embodiment, the rotation angle and the blower output frequency are used as individual genes, and an improved differential evolution algorithm is used for adaptive optimization.

[0051] Finally, new mutant individuals are generated. ,in For mutation operators (in this embodiment) =0.8).

[0052] c. Subsequent crossover and selection operations employ the standard binomial crossover and greedy selection strategy of the differential evolution algorithm. For each candidate solution in the feasible candidate solution set, its comprehensive cost function value is calculated, and the candidate solution with the smallest comprehensive cost function value is selected as the new current solution.

[0053] Check if any of the following termination conditions are met: reaching the maximum number of iterations (50) or exceeding a preset threshold for consecutive iterations without improvement. If either termination condition is met, output the global optimal solution; otherwise, increment the iteration counter and return to the candidate solution generation step to continue the search.

[0054] Step S4: Generate control commands based on the optimal control parameters and drive the aerator of the aeration system to complete the aeration control task.

[0055] After the differential evolution algorithm completes its iterations, the blower frequency corresponding to the individual with the lowest fitness value in the population becomes the optimal control parameter under the current operating conditions. The control program sends the optimal control parameter to the lower-level PLC via an industrial communication protocol. The PLC then writes this value into the frequency setting register of the blower inverter, completing one control cycle adjustment.

[0056] During the aeration control process, the system continuously monitors multi-source operating condition data. When a significant change in the operating condition feature vector is detected—that is, the Euclidean distance between the reconstructed operating condition feature vector and the current feature vector exceeds a preset threshold—the system restarts the improved differential evolution algorithm for online optimization. Based on the new operating condition feature vector and the new multi-source operating condition data, step S3 is re-executed to obtain new optimal control parameters, and the aeration system is immediately driven to adjust the blower frequency to adapt to the new operating conditions. This online re-optimization mechanism ensures the system's rapid dynamic response to changes in operating conditions.

[0057] Through the above method, the present invention realizes the control of aeration system based on multi-source data fusion, and deeply integrates the search mechanism of optimization algorithm with the real-time state of physical system, which significantly improves the treatment effect, control performance and operating efficiency of industrial wastewater treatment system under complex and variable working conditions.

[0058] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be within the scope of protection of the present invention.

Claims

1. An industrial sewage aeration system control method based on multi-source data fusion, characterized in that, The control method comprises: acquiring multi-source operation condition data collected by multiple preset monitoring points of the aerobic tank; After standardizing the multi-source operation condition data, an operation condition feature vector representing the overall operation state of the current aeration system is obtained; An improved differential evolution algorithm is executed to search and determine optimal control parameters with the minimum comprehensive cost function as the target; Control instructions are generated according to the optimal control parameters to drive the aeration system to complete the aeration control task; The improved differential evolution algorithm comprises: each individual in the population comprises a control parameter gene and a rotation angle gene; the mutation operation of the differential evolution algorithm comprises: selecting at least two individuals from the differential evolution algorithm population, generating an initial differential vector by performing vector subtraction operation on the two individuals; rotating the initial differential vector by an angle determined by the rotation angle gene in the individual to obtain a guide differential vector; and generating a mutated individual by using the guide differential vector; The improved differential evolution algorithm not only retains the global optimization capability, but more importantly, the search direction is coupled with the actual demand of the physical system, greatly improving the convergence speed and solution quality; The differential evolution algorithm comprises: adopting DE / rand / 1 strategy for mutation operation, and realizing spatial rotation transformation of the initial differential vector by using the Rodrigues rotation formula; The steps of the spatial rotation operation are: performing vector outer product operation to determine the rotation axis perpendicular to the initial differential vector and the operation condition feature vector; performing unitization processing on the rotation axis vector; decomposing the initial differential vector to be rotated into a parallel component parallel to the rotation axis and a vertical component perpendicular to the rotation axis; keeping the size and direction of the parallel component unchanged, and rotating the vertical component according to the preset rotation angle; performing vector addition on the unchanged parallel component and the rotated vertical component to obtain the final guide differential vector; The steps of obtaining the rotation angle comprise: taking the rotation angle and the blower output frequency in the spatial rotation operation as individual genes, and performing adaptive optimization by using the improved differential evolution algorithm to obtain the rotation angle.

2. The industrial sewage aeration system control method based on multi-source data fusion according to claim 1, characterized in that, The multi-source operation condition data comprises: influent chemical oxygen demand concentration, influent ammonia nitrogen concentration, influent flow rate, dissolved oxygen concentration at the front end of the aerobic tank, dissolved oxygen concentration in the middle section of the aerobic tank, dissolved oxygen concentration at the end of the aerobic tank, and pH value.

3. The industrial wastewater aeration system control method based on multi-source data fusion according to claim 1, characterized in that, The preprocessing steps of the multi-source operation condition data comprise: collecting multi-source operation condition data in the last N time windows; performing abnormal value elimination based on the 3σ criterion, when the difference between the collected data point and the historical data mean value is greater than 3 times the standard deviation, the data point is determined as an abnormal value and is replaced by the last valid data point; and performing standardization processing on the preprocessed data.

4. The industrial sewage aeration system control method based on multi-source data fusion according to claim 1, characterized in that, The construction steps of the operation condition feature vector comprise: calculating the change rates of the influent chemical oxygen demand concentration and the influent flow rate to obtain time change rate features; mapping each index and time change rate feature to the [0, 1] interval by using the maximum and minimum normalization formula; and constructing multiple normalized index values into a multi-dimensional operation condition feature vector in a predetermined order.

5. The industrial sewage aeration system control method based on multi-source data fusion according to claim 4, characterized in that, The time change rate feature comprises a change rate of the influent chemical oxygen demand concentration in the past N minutes and a change rate of the influent flow in the past N minutes, and the two change rates are obtained by calculating the difference between the current time and N minutes ago.

6. The industrial sewage aeration system control method based on multi-source data fusion according to claim 1, characterized in that, The comprehensive cost function is a weighted combination of an operation cost term, a dissolved oxygen concentration deviation term and a system security term; wherein the operation cost term is calculated by a pre-established blower frequency-power cubic polynomial model; the dissolved oxygen concentration deviation term is obtained by calculating the difference between the measured dissolved oxygen concentration at the front end and the end of the aerobic tank at the current time, and obtaining the absolute difference between the difference and a preset value; and the system security term is a boundary constraint penalty term, which is obtained by giving a maximum penalty value when any preset monitoring index exceeds its safety boundary.

7. The industrial sewage aeration system control method based on multi-source data fusion according to claim 6, characterized in that, The triggering conditions of the system security term include triggering a penalty when the dissolved oxygen concentration at the front end of the aerobic tank is lower than a preset minimum dissolved oxygen concentration threshold set for guaranteeing the survival of microorganisms, triggering a penalty when the dissolved oxygen concentration at the end of the aerobic tank is higher than a preset maximum dissolved oxygen concentration threshold, and triggering a penalty when the pH value in the aerobic tank is not within the suitable pH value range set for microorganisms.

Images