Electroplating wastewater dosing feedback regulation method based on online water quality detection

Abstract

The invention provides an electroplating wastewater dosing feedback regulation method based on online water quality detection, and relates to the technical field of environmental chemical analysis and water quality monitoring, the electroplating wastewater dosing feedback regulation method comprises the following steps: collecting chromium ion concentration parameters, pH value parameters and oxidation-reduction potential parameters of electroplating wastewater, and combining a trained chromium ion reduction model to carry out real-time calculation and analysis; and dynamic correction and closed-loop feedback adjustment of the dosage are realized. Therefore, the accuracy and the response speed of dosing control can be remarkably improved, the problem of excessive dosing or insufficient dosing of the medicament is effectively reduced, and meanwhile, the stability and the standard-reaching rate of wastewater treatment are ensured, so that the economical efficiency and the environmental protection property of wastewater treatment are improved.

Description

Technical Field

[0001] This invention relates to the fields of environmental chemical analysis and water quality monitoring technology, and more specifically, to a feedback regulation method for chemical dosing of electroplating wastewater based on online water quality monitoring. Background Technology

[0002] Electroplating wastewater treatment technology is an important research direction in the field of environmental protection. In recent years, with the acceleration of industrialization, the rapid development of the electroplating industry has led to the discharge of large amounts of electroplating wastewater. This wastewater typically contains high levels of heavy metal ions (such as chromium, nickel, and zinc). If the concentration of these ions is not effectively controlled, it will cause serious harm to the ecological environment and human health. In particular, hexavalent chromium ions (Cr...) are a significant concern. 6+ Chromium ions in electroplating wastewater are highly toxic and carcinogenic, making efficient treatment of these ions a key research focus. Traditional methods for treating electroplating wastewater include chemical precipitation, ion exchange, electrolysis, and membrane separation. While these methods have achieved varying degrees of removal of heavy metal ions from wastewater, they still have limitations in practical applications. For example, in chemical precipitation, the dosage often relies on empirical values, making precise dynamic adjustment difficult, leading to reagent waste or unstable treatment results. Ion exchange and membrane separation, while achieving high separation efficiency, are costly and require complex equipment maintenance. Furthermore, the lack of real-time monitoring and feedback control technologies during the treatment process makes it difficult to simultaneously achieve both stability and efficiency in wastewater treatment.

[0003] Existing electroplating wastewater treatment technologies have significant shortcomings in chemical dosing control. Traditional methods typically employ fixed dosages or simple open-loop control, failing to fully utilize online water quality monitoring technology and intelligent algorithms for precise dosing adjustments. This approach often yields unsatisfactory treatment results when faced with complex wastewater compositions and dynamically changing water quality conditions. On the one hand, insufficient dosing prevents the complete reduction reaction of hexavalent chromium ions, leading to effluent exceeding quality standards; on the other hand, excessive dosing not only causes economic losses but may also trigger secondary pollution. Furthermore, most existing technologies lack closed-loop control mechanisms, failing to dynamically adjust the dosing based on real-time changes in water quality, resulting in slow response times and difficulty meeting increasingly stringent modern industrial emission standards. Therefore, how to achieve real-time dynamic adjustment of electroplating wastewater chemical dosing based on online water quality monitoring technology combined with intelligent algorithms is a pressing technical challenge that needs to be addressed. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention is proposed. This invention provides a feedback adjustment method for chemical dosing in electroplating wastewater based on online water quality monitoring. This method addresses, to some extent, the technical problem in electroplating wastewater treatment where fluctuations in the concentration of heavy metals in the incoming water cause a fixed dosage that cannot adapt to changes in water quality, resulting in wasted chemicals or unstable treatment effects.

[0005] According to one aspect of the present invention, a feedback regulation method for chemical dosing of electroplating wastewater based on online water quality detection is provided, comprising:

[0006] Collect chromium ion concentration parameters, pH value parameters, and redox potential parameters of electroplating wastewater, and set the initial dosage benchmark value based on the chromium ion concentration parameters;

[0007] The pH value parameter and redox potential parameter are input into the trained chromium ion reduction model to obtain the chromium ion reduction index;

[0008] A fuzzy neural network is used to calculate the rate of change and acceleration of the chromium ion reduction index per unit time. A dosage correction coefficient is generated based on the rate of change and acceleration. The product of the dosage correction coefficient and the initial dosage reference value is used as the real-time dosage.

[0009] The dosing device is controlled to administer the drug according to the real-time dosage, and the pH value and redox potential parameters are continuously collected during the dosing process. The chromium ion reduction index is updated in real time according to the parameter change trend to form a closed-loop feedback regulation.

[0010] Furthermore, in the chromium ion reduction model, a mapping relationship between the conversion rate of hexavalent chromium to trivalent chromium and pH value and redox potential is established based on historical processing data; by calculating the dynamic response coefficient of pH value parameter and the dynamic response coefficient of redox potential parameter, and based on the dynamic response coefficient, the comprehensive operating condition coefficient is calculated, and then the chromium ion reduction index is calculated based on the time response characteristics.

[0011] Furthermore, the calculation of the comprehensive operating condition coefficient is shown in the following formula:

[0012]

[0013] in, and These represent the rates of change of pH value and redox potential response coefficients, respectively, β is the overall operating condition coefficient, and α pH For optimal reducing pH, α orp This is the optimal reduction potential.

[0014] Furthermore, the fuzzy neural network includes an input layer, a fuzzing layer, a rule layer, a defuzzing layer, and an output layer;

[0015] The rate of change and acceleration of the chromium ion reduction index are calculated in the input layer.

[0016] The fuzzy layer performs fuzzification processing on the calculated rate of change and acceleration;

[0017] The rule layer establishes multiple rules based on expert experience and historical data, and sets corresponding correction coefficients for different combinations of working conditions;

[0018] The defuzzing layer uses an improved centroid method to calculate the dosage correction coefficient and achieves adaptive adjustment of control parameters through a rule-based competition mechanism.

[0019] Furthermore, the fuzzy layer divides the rate of change and acceleration into multiple continuous and partially overlapping intervals according to their value ranges;

[0020] The fuzzy layer is divided according to the degree of membership of each interval based on the actual measured rate of change and acceleration value;

[0021] The multiple consecutive and partially overlapping intervals include the "negative large" interval, the "negative medium" interval, the "negative small" interval, the "zero" interval, the "positive small" interval, the "positive medium" interval, and the "positive large" interval.

[0022] Furthermore, the rule layer determines the specific operating conditions based on the intervals where the acceleration and rate of change are located, outputs different correction coefficients, and determines the degree of influence of each rule on the final control output by the activation intensity.

[0023] Furthermore, the activation intensity is calculated as follows:

[0024]

[0025] in,

[0026]

[0027] Where, ω j Let μ be the rule strength of the j-th rule. Aj μ is the membership function of the rate. Bj Let v be the membership function of acceleration. smooth Here, 'a' is the velocity value, 'a' is the acceleration value, and 'v' is the velocity value. max For the maximum speed, a max For the maximum acceleration, μ i (x) represents the membership degree of input x to the i-th fuzzy set, where x is the input variable and c i Let σ be the center value of the i-th fuzzy set. i Let be the width parameter of the i-th fuzzy set, and k be the shape adjustment coefficient.

[0028] Furthermore, the deblurring layer uses an improved centroid method to calculate the dosage correction coefficient, and controls the parameters to be adaptively adjusted through a rule-based competition mechanism.

[0029] Furthermore, the improved centroid method calculates the strength value of each rule and combines it with a dynamic weight adjustment factor to perform a weighted average of the rule output. At the same time, it uses an interval correction mechanism to prevent the result from falling into local extrema, and finally obtains the dosage correction coefficient.

[0030] Furthermore, the calculation of the real-time dosage is shown in the following formula:

[0031]

[0032] Among them, Q real Q represents the actual dosage. base δ represents the initial dosage baseline value, and δ is the dynamic correction coefficient for the dosage. K is the rate of change of the correction coefficient. cor This is the correction factor.

[0033] Compared with existing technologies, the electroplating wastewater dosing feedback regulation method based on online water quality monitoring provided by this invention collects chromium ion concentration, pH value, and redox potential parameters of the electroplating wastewater, and performs real-time calculations and analyses using a trained chromium ion reduction model. This achieves dynamic correction and closed-loop feedback regulation of the dosing dosage. This significantly improves the accuracy and response speed of dosing control, effectively reduces the problem of overdosing or underdosing of chemicals, and ensures the stability and compliance rate of wastewater treatment, thereby enhancing the economic and environmental benefits of wastewater treatment. Attached Figure Description

[0034] 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. In the drawings:

[0035] Figure 1 This is a flowchart of a feedback adjustment method for chemical dosing in electroplating wastewater based on online water quality monitoring, according to an embodiment of the present invention.

[0036] Figure 2 This is a schematic diagram illustrating the relationship between the chromium ion reduction index and the dosage correction coefficient based on online water quality detection according to an embodiment of the present invention. Detailed Implementation

[0037] Hereinafter, exemplary embodiments according to the present invention will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein.

[0038] Figure 1 This is a flowchart of a feedback adjustment method for chemical dosing in electroplating wastewater based on online water quality monitoring, according to an embodiment of the present invention. Figure 1 As shown, the feedback adjustment method for chemical dosing in electroplating wastewater based on online water quality monitoring includes:

[0039] S1: Collect the chromium ion concentration parameters, pH value parameters, and redox potential parameters of the electroplating wastewater, and set the initial dosage benchmark value based on the chromium ion concentration parameters;

[0040] An online water quality monitor was used to collect chromium ion concentration parameters in electroplating wastewater in real time. These parameters included hexavalent chromium concentration and total chromium concentration. Simultaneously, pH and oxidation-reduction potential (ORP) parameters, reflecting the oxidation-reduction state of the wastewater, were also collected. The pH value was collected every 10 minutes using a pH electrode, and the ORP parameter was collected every 10 minutes using an ORP electrode. The trivalent chromium concentration was determined based on the difference between the hexavalent and total chromium concentrations. An initial dosage reference value for the reducing agent was set at a mass ratio of 2:1 (hexavalent chromium to trivalent chromium concentration). The initial dosage reference value for the flocculant was also set based on the trivalent chromium concentration. Sodium bisulfite was selected as the reducing agent, and polyaluminum chloride was selected as the flocculant. The unit of the initial dosage reference value was mg / L.

[0041] S2: Input the pH value parameter and redox potential parameter into the trained chromium ion reduction model to obtain the chromium ion reduction index, wherein the chromium ion reduction model establishes a mapping relationship between the conversion rate of hexavalent chromium to trivalent chromium and pH value and redox potential based on historical processing data.

[0042] The real-time collected pH and redox potential parameters are input into a trained chromium ion reduction model. The chromium ion reduction model is constructed using an improved multilayer perceptron structure. The input layer contains two neurons for pH and redox potential parameters, the hidden layer has a three-layer structure with eight neurons in each layer, and the output layer has a single neuron. The neurons use an improved ReLU activation function. The training data for the chromium ion reduction model comes from historical processing data from the past year, including 10,000 sets of corresponding records of pH, redox potential, and hexavalent chromium reduction conversion rate. The historical processing data has undergone outlier detection and data standardization preprocessing.

[0043] In the chromium ion reduction model, the dynamic response coefficient of the pH parameter is first calculated. This dynamic response coefficient reflects the degree to which the pH value deviates from the optimal reduction range, and is calculated using a modified Gaussian function.

[0044]

[0045] Where, α pH The optimal reduction pH value is the pH value determined through batch testing for the reduction of hexavalent chromium. max pH is the upper limit. min These are the lower limits, which are the acceptable pH ranges for the reduction reaction determined by orthogonal experiments; when the pH is within the optimal reduction range, the dynamic response coefficient is close to 1, and the response coefficient decays exponentially as the pH deviates from the optimal range.

[0046] The dynamic response coefficient of the redox potential parameter is then calculated. This dynamic response coefficient characterizes the degree of influence of the redox potential on the reduction reaction, and is also calculated using a modified Gaussian function.

[0047]

[0048] Where, α orp The optimal reduction potential, determined through dynamic monitoring, represents the optimal reduction conditions for hexavalent chromium; ORP max Upper limit, ORP min These two limits define the effective potential range of the reduction reaction; when the redox potential is in the optimal reduction range, the dynamic response coefficient is close to 1, and as the potential deviates from the optimal range, the response coefficient decays exponentially.

[0049] Based on the two dynamic response coefficients mentioned above, an innovative coupling effect of parameter change rate is introduced to calculate the comprehensive operating condition coefficient. When both pH and redox potential parameters change towards their optimal range, it indicates that the reduction reaction is progressing in a favorable direction. In this case, the comprehensive operating condition coefficient is increased through the positive correlation of parameter change rates. When the two parameters show opposite trends, it indicates that the reduction reaction may be abnormal. In this case, the comprehensive operating condition coefficient β is decreased. The comprehensive operating condition coefficient β is shown in the following formula:

[0050]

[0051] in, and These represent the rate of change of pH value and redox potential response coefficient, respectively, calculated through a 5-minute sliding time window. The comprehensive operating coefficient not only considers the absolute value of the parameters but also introduces the mutual influence of the parameter change trends. When both parameters change towards the optimal range simultaneously, the comprehensive operating coefficient will be improved.

[0052] It should be noted that when the pH value gradually decreases from the slightly alkaline range towards the optimal value, and the redox potential gradually decreases from the slightly high range towards the optimal value, although the instantaneous values ​​of both parameters have not reached the optimal range, their changing trends are synergistic (the product of the rates of change of the two parameters is positive and relatively large). In this case, the calculated comprehensive operating condition coefficient will be higher than that when only the absolute values ​​of the parameters are considered. This early prediction mechanism allows the model to identify the trend that the reduction reaction is about to enter the optimal state in advance, so that the dosing strategy can be adjusted in advance to avoid fluctuations in the reduction effect caused by parameter overshoot. Conversely, when the changing trends of pH value and redox potential are opposite (such as pH value rising while redox potential falls), even if the absolute values ​​of both parameters are at a good level, the model will reduce the comprehensive operating condition coefficient through the negative correlation of the rates of change and issue an early warning signal in time. In production practice, this mechanism can detect potential process fluctuations in advance and reserve sufficient response time for timely adjustment of treatment parameters, thereby effectively avoiding drastic fluctuations in treatment effect.

[0053] The final chromium ion reduction index I red The calculation incorporates time response characteristics. At the initial stage of the reaction, the reduction exponent increases exponentially due to the start-up time required. When the reaction duration exceeds the characteristic time constant, the reduction exponent is primarily controlled by the comprehensive operating condition coefficient, as shown in the following formula:

[0054]

[0055] Where t is the reduction reaction time in seconds; τ is the system characteristic time constant, which is 300s. This value was determined through a step response experiment and reflects the time required for the reduction reaction to reach a steady state.

[0056] The chromium ion reduction model updates its calculations every minute under normal operating conditions. When a sudden change in influent water quality is detected (pH change exceeding 0.5 units within 1 minute or redox potential change exceeding 50 mV within 1 minute), the model automatically increases the calculation frequency to once every 30 seconds until the water quality stabilizes. All calculation results are stored in the database in real time, along with the operating conditions at the time of calculation. Every 24 hours of operation, the system automatically performs online calibration of the model, fine-tuning the model parameters by comparing the error between predicted and actual monitored values ​​to ensure the model's prediction accuracy.

[0057] For example, when the online monitoring instrument collects real-time data showing that the influent pH value is slightly higher than the median of the optimal reduction range, and the oxidation-reduction potential is also higher than the median of the optimal reduction range, the model first calculates the dynamic response coefficients of the two parameters separately. At this time, since neither parameter is at its optimal value but is still within an acceptable range, the calculated dynamic response coefficients are close to but have not reached their maximum values. By comparing the historical data from the previous time window, it can be found that both the pH value and the oxidation-reduction potential show a trend of moving closer to the optimal range, that is, the pH value is slowly decreasing and the oxidation-reduction potential is also gradually decreasing, indicating that the trends of the two parameters are consistent and in the correct direction. Based on this, the comprehensive operating condition coefficient calculated is at a high level. At the same time, since the reaction duration at this moment has significantly exceeded the system characteristic time constant, the time response factor is close to the steady-state value. The chromium ion reduction index finally calculated by the model is in the good range but has not reached the optimal range. Based on this, the system suggests that the dosage of the reducing agent be appropriately adjusted to further optimize the reduction effect.

[0058] S3: Use a fuzzy neural network to calculate the rate of change and acceleration of the chromium ion reduction index per unit time, generate a dosage correction coefficient based on the rate of change and acceleration, and use the product of the dosage correction coefficient and the initial dosage benchmark value as the real-time dosage.

[0059] The dynamic characteristics of the chromium ion reduction index are calculated using an improved Takagi-Sugeno fuzzy neural network. This fuzzy neural network comprises five layers: an input layer, a fuzzy layer, a regularization layer, a defuzzification layer, and an output layer. When calculating the dynamic characteristics of the chromium ion reduction index, the rate of change and acceleration of the index are first calculated in the input layer using a 10-minute sliding time window. The rate and acceleration values ​​are then calculated sequentially at each time point, with a 1-minute interval. When wastewater quality fluctuates, the system adapts to changes in operating conditions by adjusting the sampling frequency of the sliding time window. The sampling frequency is increased to once every 30 seconds when water quality fluctuations are severe, and returns to once every 1 minute when water quality is relatively stable. The calculation of the dynamic characteristics is shown in the following formula:

[0060]

[0061] Where v(t) is the rate of change at time t, I red The chromium ion reduction index is given by I, where t is the current time point and Δt is the time interval. red (t) represents the restoration exponent at the current time, I red (t-Δt) is the restoration exponent at the previous moment, a(t) is the acceleration at moment t, v(t) is the velocity at the current moment, v(t-Δt) is the velocity at the previous moment, and Δt is the time interval.

[0062] In the fuzzy layer, the calculated rate of change and acceleration are fuzzified. First, the range of the rate of change is divided into seven consecutive, partially overlapping intervals, from smallest to largest. When the rate of change is in the smallest interval, it is classified as a "negative large" interval, indicating that the restoration index is rapidly decreasing; when the rate of change gradually increases into the second smallest interval, it is classified as a "negative medium" interval, indicating that the restoration index is decreasing at a moderate rate; when the rate of change continues to increase into the medium-small interval, it is classified as a "negative small" interval, indicating that the restoration index is slowly decreasing; when the rate of change is in the middle interval, it is classified as a "zero" interval, indicating that the restoration index is basically stable; when the rate of change enters the medium-large interval, it is classified as a "positive small" interval, indicating that the restoration index is slowly increasing; when the rate of change enters the second largest interval, it is classified as a "positive medium" interval, at which point the dosing strategy needs to be adjusted promptly; when the rate of change reaches the largest interval, it is classified as a "positive large" interval, indicating that the restoration index is rapidly increasing.

[0063] Similarly, when the system fuzzifies the acceleration, it divides the acceleration value range into seven consecutive and partially overlapping intervals. When the acceleration is in the minimum interval, it is determined to be in the "negative large" interval, indicating that the downward trend of the restoration index is rapidly increasing; when the acceleration enters the second smallest interval, it is determined to be in the "negative medium" interval, indicating that the downward trend of the restoration index is increasing at a moderate speed; when the acceleration is in the small-medium interval, it is determined to be in the "negative small" interval, indicating that the downward trend of the restoration index is slowly increasing; when the acceleration is in the middle interval, it is determined to be in the "zero" interval, indicating that the trend of the restoration index is basically stable; when the acceleration enters the large-medium interval, it is determined to be in the "positive small" interval, indicating that the upward trend of the restoration index is slowly increasing; when the acceleration reaches the second largest interval, it is determined to be in the "positive medium" interval, indicating that the upward trend of the restoration index is increasing at a moderate speed, and the system needs to adjust the control strategy in time; when the acceleration is in the maximum interval, it is determined to be in the "positive large" interval, indicating that the upward trend of the restoration index is rapidly increasing.

[0064] During the fuzzification process, the system calculates the membership degree of each interval based on the measured rate of change and acceleration values. The membership degree is highest when a parameter value is exactly at the center of an interval; as the parameter value deviates from the center, the membership degree decreases, while the membership degree of adjacent intervals increases. When the rate of change or acceleration value overlaps with two adjacent intervals, the system calculates the membership degree of that value to both intervals and determines which interval the value leans towards based on the membership degree. For example, when the rate of change overlaps with both the "smallest" and "middle" intervals, its membership degree is calculated to determine the appropriate weights for different rules in subsequent rule reasoning.

[0065] Upon entering the rule layer, the system establishes a fuzzy rule base containing 49 rules based on expert experience and historical data. Specifically, it determines the specific operating condition based on the intervals where the acceleration and rate of change fall, outputting different correction coefficients. Furthermore, the activation intensity determines the degree of influence of each rule on the final control output. For example, when the chromium ion reduction index shows a rapid downward trend (both the rate of change and acceleration are negative), the activation intensity of the corresponding rule will increase. This increases the weight of the rule in the final dosage correction, thereby achieving a rapid increase in dosage. More specifically, the activation intensity is calculated as follows:

[0066]

[0067] in,

[0068]

[0069] Where, ω j Let μ be the rule strength of the j-th rule. Aj μ is the membership function of the rate. Bj Let v be the membership function of acceleration. smooth Here, 'a' is the velocity value, 'a' is the acceleration value, and 'v' is the velocity value. max For the maximum speed, a max For the maximum acceleration, μ i (x) represents the membership degree of input x to the i-th fuzzy set, where x is the input variable and c i Let σ be the center value of the i-th fuzzy set. i Let be the width parameter of the i-th fuzzy set, and k be the shape adjustment coefficient.

[0070] The defuzzing layer employs an improved centroid method to calculate the dosage correction coefficient, and an adaptive adjustment of control parameters is achieved through a rule-based competition mechanism. The final real-time dosage is determined by the dynamic product of the correction coefficient and the initial dosage reference value. The dosage correction coefficient K... cor The calculation is as follows:

[0071]

[0072] Among them, y j Let γ be the output value corresponding to the j-th rule, and γ be the dynamic adjustment coefficient. ω represents the rate of change of the activation intensity of the rule. j Let be the rule strength of the j-th rule.

[0073] It should be noted that the rule competition mechanism is a dynamic rule selection method established between the rule layer and the defuzzification layer of the fuzzy neural network. When the system receives new change rate and acceleration data, it first calculates the competition intensity of each rule. The competition intensity is determined by three factors: the rule's activation level, historical execution effect, and rule confidence. When a rule demonstrates good control effect in multiple consecutive calculations, its confidence gradually increases, giving it an advantage in subsequent competitions. If a rule's control effect is poor or causes increased system fluctuations, its confidence is reduced. The system sets a dynamic competition threshold to filter out rules with high competition intensity to participate in the final control calculation. The competition threshold is automatically adjusted according to the system's operating state: when the system is in a stable state, the competition threshold is increased to reduce the number of rules participating in the calculation; when the system is in a fluctuating state, the competition threshold is decreased to increase rule diversity. If a rule is detected as failing to win in the competition for a long period, the system will adaptively adjust or temporarily freeze that rule to ensure that the rule base always maintains efficient decision-making capabilities.

[0074] The calculation logic of the improved centroid method is as follows:

[0075] First, the system calculates the rule strength for each rule. When the antecedent of a rule includes linguistic descriptions of the rate of change and acceleration, the rule strength is determined according to the degree of matching between the actual values ​​of the parameters and the linguistic variables: when both the rate of change and acceleration are within the core range of the corresponding linguistic variables, the rule strength is at its maximum; when a parameter of the rate of change or acceleration deviates from the core range but is still within the transition range, the rule strength gradually decreases according to the degree of deviation; when any parameter completely deviates from the set range, the rule strength is at its minimum.

[0076] Then, when the operating conditions corresponding to a certain rule remain stable, its dynamic weight factor gradually increases, but does not exceed the preset upper limit; when the operating conditions fluctuate frequently, the dynamic weight factor gradually decreases, but does not fall below the preset lower limit. This dynamic weight adjustment mechanism can highlight the control effect under stable operating conditions.

[0077] Next, the overall weight of the rule is calculated. The overall weight is equal to the product of the rule strength and the dynamic weight factor, and then multiplied by the rule's historical performance score. The historical performance score is determined based on the degree of improvement in the restoration index after the rule's most recent executions, with upper and lower limits set for the score range.

[0078] For the dosage correction coefficient output by each rule, an adaptive range limit is set: when the operating conditions are within the normal range, a narrower range is used; when the operating conditions fluctuate significantly, the range is appropriately expanded; when the operating conditions change drastically, the widest range is used.

[0079] The final dosage correction coefficient is obtained by weighted averaging based on rule intensity, while also incorporating an interval correction coefficient. The interval correction coefficient is used to prevent the correction coefficient from falling into local extrema: when the correction coefficient changes very little after multiple consecutive calculations and the restoration index still has not reached the target value, the current equilibrium is broken by adjusting the interval correction coefficient, as shown in the following formula:

[0080]

[0081] Among them, Q real Q represents the actual dosage. base δ represents the initial dosage baseline value, and δ is the dynamic correction coefficient for the dosage. K is the rate of change of the correction coefficient. cor This is the correction factor.

[0082] S4: Control the dosing device to administer the drug according to the real-time dosage, and continuously collect the pH value parameter and redox potential parameter during the dosing process. Update the chromium ion reduction index in real time according to the parameter change trend to form a closed-loop feedback regulation.

[0083] First, the dosing device executes the dosing operation based on the calculated real-time dosage. When the system sends a dosing command to the dosing device, the device automatically adjusts the output frequency and operating power of the dosing pump to ensure that the actual dosage matches the given value. The system monitors the operating status of the dosing pump in real time, including parameters such as pump speed, flow rate, and pipeline pressure, to ensure the accuracy and reliability of the dosing process. When the actual operating parameters of the dosing pump deviate, the system automatically compensates and adjusts, ensuring that the actual dosage always dynamically changes to follow the given value.

[0084] During the dosing process, the system continuously collects pH and redox potential parameters using multi-point online monitoring instruments. Each sampling point is equipped with an independent pH electrode and redox potential electrode to achieve real-time parameter monitoring. The system preprocesses the collected raw signals, including signal filtering, temperature compensation, and data standardization, to ensure data accuracy and comparability. When data from a certain monitoring point becomes abnormal, the system automatically switches to a backup monitoring point to ensure the continuity of parameter acquisition.

[0085] The system dynamically calculates the chromium ion reduction index based on the processed pH value and redox potential parameters. When new parameter data arrives, the system immediately updates the calculated reduction index. During the calculation process, the system considers the rate and acceleration of parameter change to accurately reflect the dynamic characteristics of the reduction process. If abnormal fluctuations occur in the parameters, the system activates a data verification mechanism, cross-validating data from multiple detection points to ensure the reliability of the calculated reduction index.

[0086] Based on the updated restoration index, the system evaluates the current restoration effect in real time. When the restoration index deviates from the target value, the system analyzes the direction and degree of the deviation, and predicts the development trend of the restoration process by combining the changing trends of parameters. The system determines the next control strategy by establishing a mapping relationship between parameter change patterns and dosage adjustment needs. During the evaluation process, the system also considers the statistical characteristics of historical data to improve the accuracy of predictions.

[0087] Based on the evaluation results, the system executes closed-loop feedback regulation. When the reduction effect is insufficient, the system gradually increases the dosage according to the degree of deviation; when the reduction is excessive, the system reduces the dosage accordingly. During the regulation process, the system adopts a gradual regulation strategy to avoid abrupt changes in the dosage. After each regulation, the system closely monitors the parameter change response and dynamically adjusts the control parameters according to the response characteristics to achieve precise regulation.

[0088] In summary, the feedback regulation method for chemical dosing in electroplating wastewater based on online water quality monitoring, as described in this invention, is elucidated. It collects parameters such as chromium ion concentration, pH value, and redox potential of the electroplating wastewater and performs real-time calculations and analyses using a trained chromium ion reduction model, achieving dynamic correction and closed-loop feedback regulation of the dosing dosage. This significantly improves the accuracy and response speed of dosing control, effectively reduces the problem of overdosing or underdosing of chemicals, and ensures the stability and compliance rate of wastewater treatment, thereby enhancing the economic and environmental benefits of wastewater treatment.

[0089] Here, those skilled in the art will understand that the specific operations of each step in the above-described feedback adjustment method for chemical dosing of electroplating wastewater based on online water quality monitoring have been referenced above. Figure 1 and Figure 2 The method for chemical dosing and feedback regulation of electroplating wastewater based on online water quality monitoring has been described in detail, and therefore, its repeated description will be omitted.

Claims

1. A feedback adjustment method for chemical dosing in electroplating wastewater based on online water quality monitoring, characterized in that, include: Collect chromium ion concentration parameters, pH value parameters, and redox potential parameters of electroplating wastewater, and set the initial dosage benchmark value based on the chromium ion concentration parameters; The pH value parameter and redox potential parameter are input into the trained chromium ion reduction model to obtain the chromium ion reduction index; A fuzzy neural network is used to calculate the rate of change and acceleration of the chromium ion reduction index per unit time. A dosage correction coefficient is generated based on the rate of change and acceleration. The product of the dosage correction coefficient and the initial dosage reference value is used as the real-time dosage. The dosing device is controlled to administer the drug according to the real-time dosage, and the pH value and redox potential parameters are continuously collected during the dosing process. The chromium ion reduction index is updated in real time according to the parameter change trend to form a closed-loop feedback regulation.

2. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 1, characterized in that, The chromium ion reduction model establishes a mapping relationship between the conversion rate of hexavalent chromium to trivalent chromium and pH value and redox potential based on historical processing data. The dynamic response coefficients of pH value parameter and redox potential parameter are calculated, and the comprehensive operating condition coefficient is calculated based on the dynamic response coefficients. Then, the chromium ion reduction index is calculated based on the time response characteristics.

3. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 2, characterized in that, The comprehensive operating condition coefficient is calculated as follows: in, and These represent the rates of change of pH value and redox potential response coefficients, respectively, β is the overall operating condition coefficient, and α pH For the optimal reducing pH, α orp This is the optimal reduction potential.

4. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 3, characterized in that, The fuzzy neural network includes an input layer, a fuzzing layer, a rule layer, a defuzzing layer, and an output layer; The rate of change and acceleration of the chromium ion reduction index are calculated in the input layer. The fuzzy layer performs fuzzification processing on the calculated rate of change and acceleration; The rule layer establishes multiple rules based on expert experience and historical data, and sets corresponding correction coefficients for different combinations of working conditions; The defuzzing layer uses an improved centroid method to calculate the dosage correction coefficient and achieves adaptive adjustment of control parameters through a rule-based competition mechanism.

5. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 4, characterized in that, The fuzzy layer divides the rate of change and acceleration into multiple continuous and partially overlapping intervals according to their value ranges; The fuzzy layer is divided according to the degree of membership of each interval based on the actual measured rate of change and acceleration value; The multiple consecutive and partially overlapping intervals include the "negative large" interval, the "negative medium" interval, the "negative small" interval, the "zero" interval, the "positive small" interval, the "positive medium" interval, and the "positive large" interval.

6. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 3, characterized in that, The rule layer determines the specific operating conditions based on the intervals in which the acceleration and rate of change fall, outputs different correction coefficients, and determines the degree of influence of each rule on the final control output by the activation intensity.

7. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 6, characterized in that, The activation intensity is calculated as follows: in, Where, ω j Let μ be the rule strength of the j-th rule. Aj μ is the membership function of the rate. Bj Let v be the membership function of acceleration. smooth Here, 'a' is the velocity value, 'a' is the acceleration value, and 'v' is the velocity value. max For the maximum speed, a max For the maximum acceleration, μ i (x) represents the membership degree of input x to the i-th fuzzy set, where x is the input variable and c i Let σ be the center value of the i-th fuzzy set. i Let be the width parameter of the i-th fuzzy set, and k be the shape adjustment coefficient.

8. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 3, characterized in that, include: The deblurring layer uses an improved centroid method to calculate the dosage correction coefficient, and the control parameters are adaptively adjusted through a rule-based competition mechanism.

9. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 8, characterized in that, The improved centroid method calculates the strength value of each rule and combines it with a dynamic weight adjustment factor to perform a weighted average of the rule output. At the same time, it uses an interval correction mechanism to prevent the result from getting trapped in local extrema, and finally obtains the dosage correction coefficient.

10. The electroplating wastewater dosing feedback adjustment method based on online water quality detection according to claim 9, characterized in that, The calculation of the real-time dosage is shown in the following formula: Among them, Q real Q represents the actual dosage. base δ represents the initial dosage baseline value, and δ is the dynamic correction coefficient for the dosage. K is the rate of change of the correction coefficient. cor This is the correction factor.

Images