A pure water pressure self-adaptive control method based on multi-sensor fusion

Abstract

This invention relates to the field of water pressure control technology, and more particularly to a pure water pressure adaptive control method based on multi-sensor fusion. The method includes: acquiring real-time liquid level values ​​from a pure water storage tank and real-time pressure values ​​from a water supply network using level and pressure sensors, respectively; determining a feedforward pressure compensation amount based on changes in the real-time liquid level values, and correcting a fixed pressure setpoint accordingly to obtain a dynamic pressure setpoint; determining an adaptive gain coefficient negatively correlated with the real-time liquid level values; calculating an initial power adjustment amount for a booster pump based on the dynamic pressure setpoint and real-time pressure values, and correcting this initial adjustment amount using the adaptive gain coefficient to obtain the actual power adjustment amount; and adjusting the power of the booster pump according to the actual power adjustment amount to regulate the water supply pressure. This invention improves the pressure stability and adaptive adjustment capability of a pure water system when facing flow fluctuations.

Description

Technical Field

[0001] This invention relates to the field of water pressure control technology, and in particular to a pure water pressure adaptive control method based on multi-sensor fusion. Background Technology

[0002] Electrolytic copper foil, as a key basic material for the negative electrode current collector of lithium-ion batteries and printed circuit boards, places extremely high demands on process stability and quality control in its production process. The manufacturing of raw foil and surface treatment are the core processes that determine the final performance of the copper foil. The surface of the raw foil produced by the electrolytic foil machine is coated with electrolyte containing copper ions, sulfate ions, and other components. This must be thoroughly removed through a subsequent water washing process to prevent impurities from crystallizing or interfering with subsequent electroplating reactions. The surface treatment process imparts specific surface morphology and properties to the copper foil through a series of precision electroplating processes. The water washing after this process is equally crucial, used to remove residual chemicals, prevent cross-contamination of the plating bath, and provide a clean base for obtaining consistent product performance. In modern production lines, the above-mentioned water washing process typically relies on high-purity water, and the stability of its water supply pressure is a prerequisite for ensuring uniform and thorough cleaning. Current water supply systems typically consist of a front-end pure water treatment unit, intermediate storage tanks, transfer pumps, and piping networks, and generally employ a distributed control system (DCS) for automated monitoring. Under this architecture, the liquid level in the storage tank is monitored by a liquid level sensor and the water supply valve is activated to maintain a stable liquid level. At the same time, the water pressure is monitored by a pipeline pressure sensor, and the operation of the booster pump is adjusted accordingly, thus forming a basic pressure and liquid level control loop.

[0003] However, in actual operation, especially in continuous electrolytic copper foil production scenarios, existing control methods face significant challenges. On one hand, the permeate flow rate of the pure water treatment system changes due to variations in its own operating state, such as reverse osmosis membrane cleaning, raw water quality fluctuations, and equipment switching. These upstream flow changes are directly transmitted to the water supply network, causing water pressure fluctuations. Existing pressure control loops mostly employ feedback regulation based on the current pressure deviation, which has a lag in response. On the other hand, the tank level itself is an important system state variable. Existing methods typically fail to deeply integrate this critical boundary condition into the pressure control logic, resulting in a fixed adjustment intensity (gain) of the control system. This prevents adaptive adjustment based on the tank's real-time buffering capacity, leading to decreased control performance under extreme liquid level conditions. These insufficient pressure fluctuation suppression capabilities and lack of system adaptive adjustment capabilities are significant reasons for quality defects in copper foil, such as incomplete washing, surface dents, and color differences.

[0004] Chinese Patent Publication No. CN106194779A discloses a water pump control method and controller based on a pressure sensor. This method automatically sets the pump's stop pressure by continuously reading the pump's operating pressure as the start-up pressure rises. This eliminates the need for manual setting of the pump's stop pressure, improving automation and accuracy, and reducing the frequency of pump start-ups and shutdowns. Furthermore, it provides a control method for automatically setting the pump's start-up pressure, further enhancing automation and reducing user intervention. However, the pressure sensor-based water pump control method and controller suffer from the following problems: insufficient pressure fluctuation suppression and a lack of adaptive adjustment capabilities. Summary of the Invention

[0005] Therefore, this invention provides a pure water pressure adaptive control method based on multi-sensor fusion to overcome the problems of insufficient pressure fluctuation suppression capability and lack of adaptive adjustment capability in the existing pure water system.

[0006] To achieve the above objectives, the present invention provides a pure water pressure adaptive control method based on multi-sensor fusion, comprising:

[0007] Step S1: Obtain the real-time liquid level value of the pure water storage tank through the liquid level sensor and obtain the real-time pressure value of the water supply network through the pressure sensor.

[0008] Step S2: Determine the feedforward pressure compensation amount based on the change in the real-time liquid level value;

[0009] Step S3: Based on the feedforward pressure compensation amount, the preset fixed pressure setting value is corrected to obtain the dynamic pressure setting value.

[0010] Step S4: Determine the adaptive gain coefficient based on the real-time liquid level value, wherein the adaptive gain coefficient is negatively correlated with the real-time liquid level value;

[0011] Step S5: Based on the dynamic pressure setpoint and the real-time pressure value, calculate the initial power adjustment amount of the booster pump, and correct the initial power adjustment amount according to the adaptive gain coefficient to determine the actual power adjustment amount.

[0012] Step S6: Adjust the power of the booster pump according to the actual power output to regulate the water supply pressure.

[0013] Furthermore, in step S1, when the absolute value of the difference between two adjacent real-time liquid level values ​​is greater than or equal to a first preset threshold, the frequency of the liquid level sensor acquiring the real-time liquid level value of the pure water storage tank and the frequency of the pressure sensor acquiring the real-time pressure value of the water supply network are increased.

[0014] Further, step S2 includes:

[0015] Step S21: Based on the real-time liquid level value and the previously acquired historical liquid level value, calculate the rate of change of liquid level per unit time;

[0016] Step S22: Query the preset compensation relationship table according to the liquid level change rate to determine the corresponding feedforward pressure compensation amount;

[0017] Step S23: Adjust the direction of the feedforward pressure compensation amount based on the sign of the liquid level change rate.

[0018] Furthermore, in step S23, when the liquid level change rate is negative, the feedforward pressure compensation amount is positive, and when the liquid level change rate is positive, the feedforward pressure compensation amount is negative.

[0019] Further, step S3 includes:

[0020] Step S31: Obtain the feedforward pressure compensation amount determined in step S2;

[0021] Step S32: The feedforward pressure compensation amount is algebraically superimposed with the preset fixed pressure setting value to generate an initial corrected pressure setting value.

[0022] Step S33: Obtain the preset upper limit and lower limit of the pressure setting value;

[0023] Step S34: When the initial corrected pressure setting value is greater than the upper limit of the pressure setting value, the upper limit of the pressure setting value is determined as the dynamic pressure setting value; when the initial corrected pressure setting value is less than the lower limit of the pressure setting value, the lower limit of the pressure setting value is determined as the dynamic pressure setting value.

[0024] Furthermore, step S3 also includes:

[0025] Step S35: Record each determined dynamic pressure setpoint and its corresponding acquisition time to obtain a dynamic pressure setpoint sequence;

[0026] Step S36: Based on the dynamic pressure setpoint sequence, calculate the difference between the maximum and minimum values ​​of the dynamic pressure setpoint within a predefined time window to obtain the setpoint fluctuation amplitude;

[0027] Step S37: When the fluctuation amplitude of the set value is greater than or equal to a second preset threshold, the feedforward pressure compensation amount determined in step S2 is multiplied by a first correction coefficient, the first correction coefficient being less than 1, and the value after multiplying by the first correction coefficient is used as a new feedforward pressure compensation amount for the algebraic superposition in step S32.

[0028] Furthermore, in step S34, when the initial corrected pressure setting value is between the upper limit of the pressure setting value and the lower limit of the pressure setting value, the initial corrected pressure setting value is determined as the dynamic pressure setting value.

[0029] Further, step S4 includes:

[0030] Step S41: Obtain the real-time liquid level value;

[0031] Step S42: Compare the real-time liquid level value with a number of preset liquid level thresholds to determine the liquid level interval to which the real-time liquid level value belongs. The number of liquid level thresholds divide the entire liquid level measurement range into a number of continuous and non-overlapping liquid level intervals.

[0032] Step S43: Based on the liquid level range to which the real-time liquid level value belongs, query a preset gain mapping table to determine the corresponding adaptive gain coefficient. The gain mapping table defines the mapping relationship between different liquid level ranges and the adaptive gain coefficient, and the mapping relationship satisfies that the adaptive gain coefficient is negatively correlated with the real-time liquid level value.

[0033] Further, step S5 includes:

[0034] Step S51: Obtain the dynamic pressure setpoint, the real-time pressure value, and the adaptive gain coefficient determined in step S4;

[0035] Step S52: Calculate the difference between the dynamic pressure setpoint and the real-time pressure value to obtain the current pressure deviation;

[0036] Step S53: Calculate the initial power adjustment amount of the booster pump based on the current pressure deviation;

[0037] Step S54: Multiply the initial power adjustment amount by the adaptive gain coefficient to obtain the actual power adjustment amount.

[0038] Furthermore, in step S22, the feedforward pressure compensation amount is positively correlated with the absolute value of the liquid level change rate.

[0039] Compared with the prior art, the beneficial effects of the present invention are that by deeply integrating the liquid level sensor signal into the pressure control loop, the present invention constructs a dual-layer optimization mechanism that combines feedforward compensation based on the liquid level change rate with gain adaptive adjustment based on real-time liquid level. This overcomes the lag of pure feedback control and the mismatch of fixed gain under all operating conditions, thereby significantly improving the pressure control stability and adaptability of the electrolytic copper foil washing pure water supply system when facing upstream flow fluctuations and its own state changes.

[0040] Furthermore, by monitoring liquid level changes in real time and dynamically increasing the sensor sampling frequency accordingly, this invention can obtain more timely status data when the system operating conditions change drastically, providing a foundation for the accurate calculation of subsequent control algorithms and enhancing the system's perception and response speed to sudden disturbances.

[0041] Furthermore, by calculating the feedforward pressure compensation based on the liquid level change rate and dynamically correcting the fixed pressure setpoint, the present invention can pre-adjust the control target before the actual occurrence of water pressure fluctuations, thereby achieving advance suppression of pressure disturbances and effectively reducing the adjustment burden and lag effect of the feedback control loop.

[0042] Furthermore, by introducing upper and lower limits on the dynamic pressure setpoint, this invention ensures that the final control target is always limited within a safe and effective range regardless of changes in the feedforward compensation amount, thus preventing the risk of pressure runaway due to overcompensation or signal abnormalities and ensuring the safety of system operation.

[0043] Furthermore, by querying and applying different adaptive gain coefficients based on the range to which the real-time liquid level value belongs, the present invention enables the controller's adjustment intensity to automatically match the current buffering capacity and system inertia of the pure water storage tank, preventing over-adjustment oscillation at low liquid levels and improving response speed at high liquid levels, thereby achieving stable control across the entire operating range.

[0044] Furthermore, by monitoring the historical fluctuation amplitude of the dynamic pressure setpoint and automatically attenuating the feedforward compensation intensity when it exceeds a threshold, the present invention can intelligently identify and suppress internal setpoint oscillations that may be caused by excessive feedforward action or model mismatch, thereby improving the overall robustness and stability of the composite control system. Attached Figure Description

[0045] Figure 1 This is a flowchart of the pure water pressure adaptive control method based on multi-sensor fusion according to the present invention;

[0046] Figure 2 This is a flowchart of step S2 of the pure water pressure adaptive control method based on multi-sensor fusion of the present invention;

[0047] Figure 3 This is a flowchart of step S3 in the pure water pressure adaptive control method based on multi-sensor fusion of the present invention;

[0048] Figure 4 This is a flowchart of step S4 of the pure water pressure adaptive control method based on multi-sensor fusion of the present invention. Detailed Implementation

[0049] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0050] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0051] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0052] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0053] Please see Figure 1 As shown, it is a flowchart of the pure water pressure adaptive control method based on multi-sensor fusion of the present invention;

[0054] This invention provides a pure water pressure adaptive control method based on multi-sensor fusion, comprising:

[0055] Step S1: Obtain the real-time liquid level value of the pure water storage tank through the liquid level sensor and obtain the real-time pressure value of the water supply network through the pressure sensor.

[0056] Specifically, in step S1, when the absolute value of the difference between two adjacent real-time liquid level values ​​is greater than or equal to a first preset threshold, the frequency of the liquid level sensor acquiring the real-time liquid level value of the pure water storage tank and the frequency of the pressure sensor acquiring the real-time pressure value of the water supply network are increased.

[0057] In a specific embodiment, step S1 acquires the real-time liquid level value Lcurrent and the real-time pressure value Pcurrent of the system using a liquid level sensor deployed in the pure water storage tank and a pressure sensor deployed in the water supply network, respectively. Preferably, the liquid level sensor is a guided wave radar level gauge with a measurement accuracy of ±2mm and an output signal of 4-20mA; the pressure sensor is a pressure transmitter with an accuracy of ±0.25%FS and an output signal of 4-20mA. Both signals are connected to the analog input module of the distributed control system (DCS) and are sampled and recorded periodically by the DCS control cycle. The initial sampling frequency fs is set to 1Hz, that is, one set of (liquid level value, pressure value) data is acquired per second.

[0058] To improve the timeliness of control response when system operating conditions change drastically, this embodiment introduces a dynamic sampling frequency adjustment mechanism. The specific logic is as follows: Let Lprev be the liquid level value acquired at the previous sampling time, and Lcurrent be the liquid level value acquired at the current sampling time. Calculate the absolute value ΔL of the difference between two adjacent liquid level values ​​and compare it with a preset first threshold Lth:

[0059] ΔL = |Lcurrent - Lprev|;

[0060] Wherein, ΔL is the absolute value of the difference between two adjacent liquid level sampling values, in meters (m); Lth is the first preset threshold, in meters (m), and its value ranges from 0.05m to 0.20m. Preferably, Lth is 0.10m. This value is based on 1% to 4% of the total height of the pure water storage tank and is calibrated in conjunction with typical water usage fluctuation rates.

[0061] If ΔL ≥ Lth, it is determined that the liquid level is undergoing a drastic change. In this case, the sampling frequency fs of the liquid level sensor and pressure sensor is synchronously increased to a higher value fsh. This adjustment can be achieved by modifying the scan cycle of the corresponding function block in the DCS or by calling the high-speed sampling interrupt program. The adjusted frequency relationship can be expressed as:

[0062] When ΔL≥Lth, fsr=fsh;

[0063] Otherwise, fs = fsi.

[0064] Where fs is the dynamically adjusted sensor sampling frequency, measured in Hertz (Hz); fsi is the initial sampling frequency, with a value of 1 Hz; and fsh is the high-frequency sampling frequency, ranging from 2 Hz to 10 Hz, preferably 5 Hz. This preferred value ensures data timeliness while avoiding excessive computational and communication load on the DCS due to overly rapid sampling.

[0065] Understandably, the above formula and logical judgment constitute a simple condition-triggered frequency switching rule. Its core parameter Lth acts as a "sensitive valve," used to distinguish between normal small fluctuations in liquid level and drastic changes that may cause pressure disturbances. The value of fsh determines the "time resolution" of the system's perceived state during disturbances.

[0066] Understandably, the rate of change in the liquid level of a pure water storage tank is one of the most direct leading indicators reflecting the supply and demand balance of a system. When the liquid level changes significantly within adjacent sampling intervals (ΔL≥Lth), it indicates a significant change in the consumption or replenishment rate of pure water. This sudden change in flow rate will inevitably be transmitted to the pipeline network and cause pressure fluctuations, often occurring before the pressure sensor actually detects the deviation. Traditional fixed-frequency sampling strategies have inherent drawbacks under such transient conditions: during stable periods, low-frequency sampling is sufficient to meet monitoring needs and saves resources; however, when disturbances occur, low-frequency sampling leads to a delay in system state updates, causing the controller to acquire "outdated" information, thus exacerbating the lag in feedback control. This implementation dynamically increases the sensor sampling frequency by monitoring the sudden changes in the liquid level value itself, essentially proactively enhancing the sensitivity of the system's "sensory organs" in the early stages before or during strong disturbances. Higher frequency liquid level and pressure data provide a data foundation for calculating more accurate liquid level change rates in step S2 and more timely pressure deviations in step S5, making the calculation of feedforward compensation closer to real-time operating conditions and enabling feedback regulation to initiate corrective actions earlier. Therefore, this dynamic sampling mechanism is not simply data collection, but constitutes the front-end intelligent sensing link of the entire adaptive control system. By adaptively adjusting the "time granularity" of data acquisition, it creates the necessary conditions for the accurate execution of subsequent core control algorithms.

[0067] Step S2: Determine the feedforward pressure compensation amount based on the change in the real-time liquid level value;

[0068] Please continue reading. Figure 2 The diagram shows a flowchart of step S2 in the pure water pressure adaptive control method based on multi-sensor fusion of the present invention. Specifically, step S2 includes:

[0069] Step S21: Based on the real-time liquid level value and the previously acquired historical liquid level value, calculate the rate of change of liquid level per unit time;

[0070] Step S22: Query the preset compensation relationship table according to the liquid level change rate to determine the corresponding feedforward pressure compensation amount;

[0071] Specifically, in step S22, the feedforward pressure compensation amount is positively correlated with the absolute value of the liquid level change rate.

[0072] Step S23: Adjust the direction of the feedforward pressure compensation amount based on the sign of the liquid level change rate.

[0073] Specifically, in step S23, when the liquid level change rate is negative, the feedforward pressure compensation amount is positive, and when the liquid level change rate is positive, the feedforward pressure compensation amount is negative.

[0074] In one specific embodiment, step S2 calculates the feedforward pressure compensation amount based on the time-series data of the liquid level sensor, aiming to provide advance compensation for anticipated pressure disturbances. This embodiment is implemented in the DCS through function block programming.

[0075] Step S21: Calculate the liquid level change rate vL per unit time. The DCS reads the real-time liquid level value Lcurrent at the current sampling time and the historical liquid level value Lprev at the previous sampling time, and knows the time interval Δt between the two samplings (this interval is determined by the dynamic sampling frequency fsensor determined in step S1, (Δt=1 / fsensor)). The formula for calculating the liquid level change rate vL is as follows:

[0076] vL=(Lcurrent-Lprev) / Δt;

[0077] Where vL is the rate of change of liquid level, in meters per second (m / s), and its value can be positive (liquid level rises) or negative (liquid level falls); Lcurrent and Lprev are both liquid level values, in meters (m); Δt is the sampling time interval, in seconds (s).

[0078] Understandably, this formula quantifies the instantaneous rate trend of pure water flowing out of (causing a drop in liquid level) or into (causing a rise in liquid level) the storage tank by calculating the change in liquid level per unit time. The absolute value of vL reflects the drasticness of the flow rate change, while its sign indicates the direction of the change.

[0079] Step S22: Based on the liquid level change rate, query the preset compensation relationship table to determine the corresponding basic feedforward compensation amount ΔPffbase. In this embodiment, the compensation relationship table takes the absolute value of the liquid level change rate |vL| as input and the basic feedforward compensation amount ΔPffbase as output. This table is stored in the DCS database or configuration. The compensation relationship table is obtained through system identification or historical data calibration, and its core relationship satisfies that ΔPffbase is positively correlated with |vL|. A simplified linear implementation can be described as follows:

[0080] ΔPffbase=Kff×|vL|

[0081] Wherein, ΔPffbase is the basic feedforward pressure compensation amount, in kilopascals (kPa); Kff is the feedforward compensation coefficient, in kilopascals per second per meter (kPa·s / m), with a value range of 50 kPa·s / m to 200 kPa·s / m. Preferably, Kff is 100 kPa·s / m, and this value is obtained by calibrating the system step response test based on the specific hydraulic resistance characteristics of the pipeline network and the tank area; |vL| is the absolute value of the liquid level change rate, in meters per second (m / s).

[0082] Understandably, the formula ΔPffbase=Kff×|vL| is a specific functional form of the compensation relationship table. The coefficient Kff is essentially a gain that maps the "liquid level change rate" to the "required pressure compensation amount." Its physical meaning can be interpreted as: how much pressure setpoint needs to be pre-adjusted to compensate for the flow rate change predicted by the liquid level changing at a rate of |vL|. The larger the value of Kff, the more sensitive the system is to flow rate changes, and the stronger the required feedforward compensation action.

[0083] Step S23: Based on the sign of the liquid level change rate vL, adjust the direction of the basic feedforward compensation amount ΔPffbase to generate the final feedforward pressure compensation amount ΔPff. The adjustment rules are as follows:

[0084] If vL < 0 (the liquid level drops, indicating an increase in water demand and a downward trend in pipeline pressure), then ΔPff = +ΔPffbase (takes a positive value).

[0085] If vL > 0 (the liquid level rises, indicating a decrease in water replenishment or water usage, and the pipeline pressure tends to rise), then ΔPff = -ΔPffbase (take the negative value).

[0086] Where ΔPff is the feedforward pressure compensation amount of the final output, in kilopascals (kPa). If vL=0, then ΔPff=0.

[0087] Understandably, this step involves the application of a sign function. It combines the magnitude (absolute value) of the compensation calculated in step S22 with a direction (positive or negative sign). The design of the direction rule perfectly aligns with physical logic: to suppress the expected pressure drop, positive compensation (increasing the pressure setpoint) is required; to suppress the expected pressure rise, negative compensation (decreasing the pressure setpoint) is required.

[0088] Understandably, in water supply networks, pressure fluctuations stem from the instantaneous imbalance between water supply and demand. The liquid level in a pure water storage tank directly reflects the accumulated water volume within it, and the rate of change of the liquid level, vL, directly reflects the net flow difference between the inflow and outflow from the tank. When the liquid level drops rapidly (vL is negative and has a large absolute value), it indicates that the current water consumption is significantly greater than the replenishment volume. This net outflow will quickly be transmitted from the storage tank to the network, causing a drop in network pressure. Traditional pressure feedback control only begins to react after the pressure sensor actually detects this drop, resulting in an inherent delay. The core of this implementation method is to utilize the rate of change of the liquid level, vL, as a leading signal that precedes pressure changes. vL is calculated in real time through step S21, and in step S22, based on a preset system characteristic relationship (compensation relationship table or coefficient Kff), the absolute value of vL is converted into an amplitude ΔPffbase of the "pressure setpoint adjustment amount" needed to offset the upcoming pressure disturbance. This preset relationship reflects the dynamic correlation between flow rate changes and pressure changes. Step S23 assigns the correct direction to this adjustment amount based on the sign of vL, thereby generating the final ΔPff. This ΔPff will be fed into subsequent steps to dynamically correct the pressure setpoint. Therefore, the entire process is equivalent to the control system pre-calculating and issuing a compensation command based on the earlier symptom of "liquid level change" before the pressure disturbance (caused by flow rate changes) is fully reflected in the pipeline pressure gauge. This achieves proactive suppression of pressure fluctuations, effectively reduces the lag burden of feedback control, and improves the overall regulation quality and anti-interference capability of the system.

[0089] Step S3: Based on the feedforward pressure compensation amount, the preset fixed pressure setting value is corrected to obtain the dynamic pressure setting value.

[0090] Please continue reading. Figure 3 The diagram shows a flowchart of step S3 in the pure water pressure adaptive control method based on multi-sensor fusion of the present invention. Specifically, step S3 includes:

[0091] Step S31: Obtain the feedforward pressure compensation amount determined in step S2;

[0092] Step S32: The feedforward pressure compensation amount is algebraically superimposed with the preset fixed pressure setting value to generate an initial corrected pressure setting value.

[0093] Step S33: Obtain the preset upper limit and lower limit of the pressure setting value;

[0094] Step S34: When the initial corrected pressure setting value is greater than the upper limit of the pressure setting value, the upper limit of the pressure setting value is determined as the dynamic pressure setting value; when the initial corrected pressure setting value is less than the lower limit of the pressure setting value, the lower limit of the pressure setting value is determined as the dynamic pressure setting value.

[0095] Specifically, in step S34, when the initial corrected pressure setting value is between the upper limit of the pressure setting value and the lower limit of the pressure setting value, the initial corrected pressure setting value is determined as the dynamic pressure setting value.

[0096] Specifically, step S3 further includes:

[0097] Step S35: Record each determined dynamic pressure setpoint and its corresponding acquisition time to obtain a dynamic pressure setpoint sequence;

[0098] Step S36: Based on the dynamic pressure setpoint sequence, calculate the difference between the maximum and minimum values ​​of the dynamic pressure setpoint within a predefined time window to obtain the setpoint fluctuation amplitude;

[0099] Step S37: When the fluctuation amplitude of the set value is greater than or equal to a second preset threshold, the feedforward pressure compensation amount determined in step S2 is multiplied by a first correction coefficient, the first correction coefficient being less than 1, and the value after multiplying by the first correction coefficient is used as a new feedforward pressure compensation amount for the algebraic superposition in step S32.

[0100] In one specific embodiment, step S3 combines the feedforward compensation mechanism with safety boundary and stability monitoring to generate the final dynamic pressure setpoint for control. This embodiment is implemented in the DCS through sequential control logic and calculation function blocks.

[0101] Step S31: Obtain the feedforward pressure compensation amount ΔPff determined in step S2.

[0102] Step S32: Perform algebraic superposition to generate the initial corrected pressure setpoint Psettemp. The calculation formula is:

[0103] Psettemp = Psetstatic + ΔPff;

[0104] Wherein, Psettemp is the initial corrected pressure setting value, in kilopascals (kPa); Psetstatic is the preset fixed pressure setting value, which is determined according to the requirements of the electrolytic copper foil washing process, for example, it can be 400 kPa; ΔPff is the feedforward pressure compensation amount, in kilopascals (kPa).

[0105] Understandably, this formula is the core embodiment of feedforward control. It superimposes the compensation amount ΔPff based on liquid level prediction onto the process base setpoint Psetstatic, enabling the target value of the control system to be dynamically adjusted according to the trend of operating conditions, thereby issuing control commands in advance to counteract anticipated pressure disturbances.

[0106] Steps S33 and S34 involve limiting the initial correction value to obtain the dynamic pressure setpoint Psetdynamic. The system pre-sets an upper limit Psetmax and a lower limit Psetmin for the pressure setpoint. The limiting logic is as follows:

[0107] If Psettemp > Psetmax, then Psetdynamic = Psetmax;

[0108] If Psettemp < Psetmin, then Psetdynamic = Psetmin;

[0109] If Psetmin≤Psettemp≤Psetmax, then Psetdynamic=Psettemp.

[0110] Among them, Psetdynamic is the final output dynamic pressure setpoint, in kilopascals (kPa); Psetmax is the upper limit of the pressure setpoint, which is usually 15% to 25% higher than Psetstatic, for example, 480 kPa, to prevent excessive pressure from impacting the pipeline or equipment; Psetmin is the lower limit of the pressure setpoint, which is usually 15% to 25% lower than Psetstatic, for example, 320 kPa, to ensure the minimum cleaning effect and avoid electrolyte residue.

[0111] Understandably, this limiting logic serves as a safety guarantee and process constraint. It ensures that regardless of how the feedforward compensation is calculated, the final control target is always constrained within the safe and effective range allowed by the process, preventing the setpoint from exceeding the reasonable range due to sensor noise, model errors, or severe interference, thereby guaranteeing the safety and basic process requirements of the electrolytic copper foil washing process.

[0112] Steps S35 to S37 implement dynamic setpoint fluctuation monitoring and feedforward intensity adaptive suppression. This is an advanced stability maintenance mechanism.

[0113] Step S35: The system records the Psetdynamic generated in each control cycle and its timestamp ti in the archive database of the DCS, forming a time series:

[0114] {Psetdynamic(t1),Psetdynamic(t2),...,Psetdynamic(tn)}.

[0115] Step S36: Calculate the fluctuation amplitude Aset of all Psetdynamics recorded within each predefined time window Twindow. The calculation formula is:

[0116] Aset=max(Psetdynamic)-min(Psetdynamic);

[0117] Where Aset is the setpoint fluctuation amplitude, in kilopascals (kPa); max() and min() functions respectively calculate the maximum and minimum values ​​of the dynamic pressure setpoint within the time window Twindow; Twindow is the monitoring time window, in seconds (s), and its value ranges from 30s to 300s. Preferably, Twindow is 120s, which is sufficient to cover the adjustment process of typical disturbances.

[0118] Step S37: Compare the calculated Aset with the second preset threshold Ath. If Aset ≥ Ath, it is determined that the dynamic setpoint itself fluctuates too much, indicating that the feedforward compensation may be too aggressive or inconsistent with the feedback effect, posing a risk of system oscillation. At this time, the system applies a decay correction to the feedforward pressure compensation amount ΔPff obtained from step S2 in the next control cycle, generating a new compensation amount ΔPffnew for step S32:

[0119] ΔPffnew=α×ΔPff

[0120] Wherein, ΔPffnew is the corrected feedforward pressure compensation amount, in kilopascals (kPa); α is the first correction coefficient, dimensionless, with a value range of 0.3 to 0.8, preferably 0.5, which can achieve a balance between suppressing oscillations and retaining necessary feedforward action; Ath is the second preset threshold, i.e. the allowable setpoint fluctuation amplitude limit, in kilopascals (kPa), with a value range of 5% to 15% of the Psetstatic value, for example, 40 kPa, calibrated according to the system's allowable pressure stability requirements.

[0121] Understandably, steps S36 and S37 constitute a closed-loop monitoring and adjustment circuit. Aset is a direct indicator for evaluating the stability of the setpoint within the feedforward-feedback composite control system. Ath is the boundary of the system's tolerance. When Aset exceeds Ath, the strength ΔPff of the feedforward compensation is actively weakened by a coefficient α (less than 1). This is a "negative feedback" adjustment, designed to reduce the setpoint oscillations caused by over-compensation of the feedforward, thereby increasing the damping of the entire control loop and enhancing robustness.

[0122] Understandably, in the electrolytic copper foil washing process, the pure water pressure needs to be highly stable, but the control action itself must also be smooth. Frequent and drastic changes in the pressure target will cause the actuator (booster pump) to operate frequently and significantly, which may induce new pressure fluctuations and is not conducive to the uniform cleaning of the copper foil. Steps S32 to S34 constitute the main path: the feedforward compensation corrects the static setpoint, enabling the control system to "anticipate" a moving target, while the limiting function defines a clear runway boundary for this moving target, ensuring that it is always within the process safety range. Steps S35 to S37 constitute a monitoring path, which continuously observes whether the movement trajectory of the "moving target" (dynamic setpoint) is smooth. If it is detected that the target swings too much up and down within the set time window (Aset≥Ath), this suggests that the feedforward compensation may be too strong, creating an adverse interaction with feedback control or other disturbances, and has a tendency to form oscillations. At this point, the monitoring path does not directly modify the final target value that has already exceeded the limit. Instead, it adopts a "source attenuation" strategy, reducing the feedforward compensation amount ΔPff input to the main path through the coefficient α. This is equivalent to intelligently reducing the "voice" of the feedforward link when it is judged that it may be "overreacting," allowing the control system to rely more on robust feedback adjustment, thereby making the generation process of dynamic setpoints smoother. This design enables the system to actively utilize liquid level information for advance compensation when facing complex electrolytic copper foil production water scenarios with incompletely accurate models, while automatically suppressing internal instability tendencies that may arise due to model errors or interference complexity, fundamentally ensuring the stability and reliability of the water washing pressure setting command.

[0123] Step S4: Determine the adaptive gain coefficient based on the real-time liquid level value, wherein the adaptive gain coefficient is negatively correlated with the real-time liquid level value;

[0124] Please continue reading. Figure 4 The diagram shows a flowchart of step S4 in the pure water pressure adaptive control method based on multi-sensor fusion of the present invention. Specifically, step S4 includes:

[0125] Step S41: Obtain the real-time liquid level value;

[0126] Step S42: Compare the real-time liquid level value with a number of preset liquid level thresholds to determine the liquid level interval to which the real-time liquid level value belongs. The number of liquid level thresholds divide the entire liquid level measurement range into a number of continuous and non-overlapping liquid level intervals.

[0127] Step S43: Based on the liquid level range to which the real-time liquid level value belongs, query a preset gain mapping table to determine the corresponding adaptive gain coefficient. The gain mapping table defines the mapping relationship between different liquid level ranges and the adaptive gain coefficient, and the mapping relationship satisfies that the adaptive gain coefficient is negatively correlated with the real-time liquid level value.

[0128] In one specific embodiment, step S4 adaptively adjusts the controller's gain coefficient based on the real-time liquid level status of the pure water storage tank to match changes in the system's dynamic characteristics. This embodiment implements this through a comparison and lookup table function block in the DCS.

[0129] Step S41: Obtain the real-time liquid level value Lcurrent, which is measured by the liquid level sensor and sampled by the DCS.

[0130] Step S42: Compare Lcurrent with two preset liquid level thresholds to determine its corresponding liquid level range. Define a low liquid level threshold Llow and a high liquid level threshold Lhigh, where Llow < Lhigh. The entire effective liquid level measurement range is divided into three continuous and non-overlapping intervals:

[0131] Interval 1 (low liquid level interval): Lcurrent ≤ Llow;

[0132] Interval 2 (medium liquid level interval): Low < Lcurrent < Lhigh;

[0133] Interval 3 (high liquid level interval): Lcurrent ≥ Lhigh;

[0134] Wherein, Low is the low liquid level threshold, in meters (m), and its value ranges from 15% to 30% of the total height of the storage tank, preferably 20% of the total height; Lhigh is the high liquid level threshold, in meters (m), and its value ranges from 70% to 85% of the total height of the storage tank, preferably 80% of the total height. These two thresholds are calibrated based on the analysis of the storage tank's buffering capacity and the characteristics of system inertial changes.

[0135] Understandably, dividing the liquid level range into three typical operating condition intervals by setting two thresholds is an engineering strategy that simplifies the problem and is easy to implement. It captures the essential trend of the system's dynamic characteristics changing with the liquid level, namely the transition from "low buffering, fast response" at low liquid levels to "high buffering, slow response" at high liquid levels.

[0136] Step S43: Based on the liquid level range determined in step S42, query the preset gain mapping table to obtain the corresponding adaptive gain coefficient Kadapt. This table defines the mapping relationship between the range and the coefficient, satisfying the basic requirement that Kadapt is negatively correlated with Lcurrent. A specific mapping representation is as follows:

[0137] If Lcurrent belongs to interval 1 (low liquid level), then Kadapt = Klow;

[0138] If Lcurrent belongs to interval 2 (medium liquid level), then Kadapt = Kmid;

[0139] If Lcurrent belongs to interval 3 (high liquid level), then Kadapt = Khigh;

[0140] Wherein, Kadapt is the adaptive gain coefficient, dimensionless; Klow, Kmid, and Khigh are the gain values ​​corresponding to the low, medium, and high liquid level ranges, respectively. To ensure a negative correlation, their values ​​should satisfy: Klow < Kmid < Khigh. The value range of Klow is 0.3 to 0.6, preferably 0.4; the value range of Kmid is 0.7 to 1.0, preferably 0.8; and the value range of Khigh is 1.1 to 1.5, preferably 1.2. These coefficient values ​​are tuned through closed-loop debugging of the system at different liquid levels, with the criterion of ensuring both rapid and stable pressure control response.

[0141] Understandably, the gain mapping table achieves adaptive gain adjustment in the form of piecewise constants. The three key parameters, Klow, Kmid, and Khigh, are the knobs that adjust the "force" of the controller. Their values ​​directly reflect the control strategy: gentle control (small gain) is used when the system is fragile (low liquid level), and more aggressive control (large gain) is used when the system has high inertia (high liquid level).

[0142] Understandably, the physical characteristics of the electrolytic copper foil washing water supply system are as follows: the level of the pure water storage tank directly determines the system's hydraulic inertia and buffer capacity. When the level is low, the amount of water in the tank is small, the system's hydraulic inertia is low, and the pressure of the entire pipeline network is very sensitive to changes in the power of the booster pump. If a high control gain is used at this time, even a small pressure deviation will cause the controller to calculate a large power adjustment command, easily leading to frequent start-stop or large power oscillations of the booster pump. This drastic fluctuation in pressure and flow, transmitted to the washing process, will directly cause uneven impact of the water flow on the copper foil surface, affecting the uniformity of cleaning, and may even adversely affect the precision washing nozzles due to pressure transients. Conversely, when the level is high, the water volume in the storage tank is sufficient, the system's hydraulic inertia is large, and the pressure response to changes in pump power becomes sluggish. If a low gain is still used at this time, the controller's response will be sluggish, unable to effectively suppress pressure disturbances, potentially causing the washing pressure to drift slowly, which also fails to meet the stringent requirements of the process for pressure stability. Using a smaller Kadapt (such as Klow) in the low liquid level range effectively reduces the controller's sensitivity, making its action more gentle and avoiding over-adjustment that could cause oscillations, prioritizing system stability. Conversely, using a larger Kadapt (such as Khigh) in the high liquid level range increases the controller's responsiveness, enabling it to exert force more decisively to overcome system inertia and quickly correct pressure deviations. This adaptive mechanism, where gain is negatively correlated with liquid level, ensures that the same control algorithm maintains excellent control quality throughout the entire cycle from production start-up and batch water use to tank replenishment. This provides an inherent adaptive guarantee for the stability of the core water washing pressure, thus laying the control foundation for achieving a uniform and consistent cleaning effect on the electrolytic copper foil.

[0143] Step S5: Based on the dynamic pressure setpoint and the real-time pressure value, calculate the initial power adjustment amount of the booster pump, and correct the initial power adjustment amount according to the adaptive gain coefficient to determine the actual power adjustment amount.

[0144] Specifically, step S5 includes:

[0145] Step S51: Obtain the dynamic pressure setpoint, the real-time pressure value, and the adaptive gain coefficient determined in step S4;

[0146] Step S52: Calculate the difference between the dynamic pressure setpoint and the real-time pressure value to obtain the current pressure deviation;

[0147] Step S53: Calculate the initial power adjustment amount of the booster pump based on the current pressure deviation;

[0148] Step S54: Multiply the initial power adjustment amount by the adaptive gain coefficient to obtain the actual power adjustment amount.

[0149] In one specific embodiment, step S5 is the final calculation and output stage of the control algorithm. Its core is to perform feedback adjustment based on the deviation between the dynamic target and the measured pressure, and apply the adaptive gain coefficient determined in step S4 to this adjustment amount. This embodiment uses a mature PID control algorithm function block (such as the Siemens standard PID control block FB41) as the basis in the DCS, and encapsulates it to implement the specific logic of this invention.

[0150] Step S51: Obtain the three required input variables: the dynamic pressure setpoint Psetdynamic determined in step S3, the real-time pressure value Pcurrent acquired in real time by the pressure sensor and processed by the DCS, and the adaptive gain coefficient Kadapt determined in step S4.

[0151] Step S52: Calculate the current pressure deviation e. The calculation method is the dynamic pressure setpoint minus the real-time pressure value, i.e., e = Psetdynamic - Pcurrent. Here, e is the current pressure deviation in kilopascals (kPa). This value can be positive (actual pressure is lower than the setpoint), negative (actual pressure is higher than the setpoint), or zero.

[0152] Step S53: Based on the current pressure deviation e, the initial power adjustment ΔWinit of the booster pump is calculated using a PID control algorithm. The PID algorithm is a mature and widely used technology in industrial process control; its ideal formula in the continuous time domain is:

[0153] ΔWinit(t)=Kp×[e(t)+1 / Ti×∫e(τ)dτ+Td×de(t) / dt];

[0154] Where ΔWinit(t) is the initial power adjustment at time t, expressed as a percentage (%), representing the adjustment relative to the rated power of the booster pump; Kp is the proportional gain, dimensionless; Ti is the integral time, in seconds (s); and Td is the derivative time, in seconds (s). In the specific implementation of DCS, discrete incremental or positional PID algorithms are typically used. The control parameters Kp, Ti, and Td need to be tuned according to the specific pipeline network characteristics and booster pump characteristics. For example, they can be initially set using the Ziegler-Nichols method or a model-based tuning method, and then fine-tuned based on the actual control effect. Preferably, a set of available parameters is: Kp=2.0, Ti=30s, Td=5s.

[0155] Understandably, the PID controller calculates the required control output ΔWinit based on the proportional, integral, and derivative components of the deviation e. The proportional term provides immediate response, the integral term eliminates steady-state error, and the derivative term predicts the trend of deviation changes and provides damping. The three work together to enable the real-time pressure Pcurrent to track the dynamic setpoint Psetdynamic quickly and smoothly.

[0156] Step S54: Multiply the initial power regulation ΔWinit by the adaptive gain coefficient Kadapt to obtain the final actual power regulation ΔWfinal sent to the booster pump inverter or actuator. The calculation formula is:

[0157] ΔWfinal = Kadapt × ΔWinit;

[0158] Wherein, ΔWfinal is the actual power regulation amount, in percentage (%).

[0159] Understandably, steps S52 and S53 constitute the basic closed-loop feedback control loop. Their function is to calculate the required booster pump power adjustment to eliminate the deviation between the current pressure and the dynamic setpoint using a proportional-integral-derivative algorithm. This calculation process aims to ensure that the system pressure accurately tracks a setpoint that may change at any time. In the electrolytic copper foil washing system, the dynamic characteristics of the pressure control loop are not constant but significantly depend on the real-time liquid level of the pure water storage tank, i.e., the system's buffer capacity. Step S54's role is to perform condition-based intensity modulation on this basic control output. Specifically, the adaptive gain coefficient determined in step S4 is negatively correlated with the real-time liquid level. When the liquid level is low, the system's hydraulic inertia is small, making it sensitive to control actions. At this time, using a smaller adaptive gain coefficient to attenuate the initial power adjustment essentially reduces the overall gain of the control loop, thereby weakening the intensity of the control action and preventing pressure overshoot or continuous oscillation caused by excessively rapid adjustment. Such oscillation directly leads to instability in the washing water flow. When the liquid level is high, the system has high hydraulic inertia and a slow response. In this case, a larger adaptive gain coefficient is used to amplify the initial power adjustment, effectively increasing the total gain of the control loop and thus strengthening the control action. This ensures a rapid response to pressure deviations and setpoint changes, preventing pressure drift due to insufficient adjustment. Therefore, this step, by multiplying the adaptive gain coefficient by the feedback control output, achieves real-time, intelligent scaling of the control force. This allows the same feedback control law to automatically match the appropriate control strength under different tank liquid level conditions, ensuring the stability and speed of the pressure control process across the entire operating range and meeting the precise pressure stability requirements of the electrolytic copper foil washing process.

[0160] Step S6: Adjust the power of the booster pump according to the actual power output to regulate the water supply pressure.

[0161] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. A pure water pressure adaptive control method based on multi-sensor fusion, characterized in that, include: Step S1: Obtain the real-time liquid level value of the pure water storage tank through the liquid level sensor and obtain the real-time pressure value of the water supply network through the pressure sensor. Step S2: Determine the feedforward pressure compensation amount based on the change in the real-time liquid level value; Step S3: Based on the feedforward pressure compensation amount, the preset fixed pressure setting value is corrected to obtain the dynamic pressure setting value. Step S4: Determine the adaptive gain coefficient based on the real-time liquid level value, wherein the adaptive gain coefficient is negatively correlated with the real-time liquid level value; Step S5: Based on the dynamic pressure setpoint and the real-time pressure value, calculate the initial power adjustment amount of the booster pump, and correct the initial power adjustment amount according to the adaptive gain coefficient to determine the actual power adjustment amount. Step S6: Adjust the power of the booster pump according to the actual power output to regulate the water supply pressure.

2. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 1, characterized in that, In step S1, when the absolute value of the difference between two adjacent real-time liquid level values ​​is greater than or equal to a first preset threshold, the frequency of the liquid level sensor acquiring the real-time liquid level value of the pure water storage tank and the frequency of the pressure sensor acquiring the real-time pressure value of the water supply network are increased.

3. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 2, characterized in that, Step S2 includes: Step S21: Based on the real-time liquid level value and the previously acquired historical liquid level value, calculate the rate of change of liquid level per unit time; Step S22: Query the preset compensation relationship table according to the liquid level change rate to determine the corresponding feedforward pressure compensation amount; Step S23: Adjust the direction of the feedforward pressure compensation amount based on the sign of the liquid level change rate.

4. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 3, characterized in that, In step S23, when the liquid level change rate is negative, the feedforward pressure compensation amount is positive; when the liquid level change rate is positive, the feedforward pressure compensation amount is negative.

5. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 4, characterized in that, Step S3 includes: Step S31: Obtain the feedforward pressure compensation amount determined in step S2; Step S32: The feedforward pressure compensation amount is algebraically superimposed with the preset fixed pressure setting value to generate an initial corrected pressure setting value. Step S33: Obtain the preset upper limit and lower limit of the pressure setting value; Step S34: When the initial corrected pressure setting value is greater than the upper limit of the pressure setting value, the upper limit of the pressure setting value is determined as the dynamic pressure setting value; when the initial corrected pressure setting value is less than the lower limit of the pressure setting value, the lower limit of the pressure setting value is determined as the dynamic pressure setting value.

6. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 5, characterized in that, Step S3 further includes: Step S35: Record each determined dynamic pressure setpoint and its corresponding acquisition time to obtain a dynamic pressure setpoint sequence; Step S36: Based on the dynamic pressure setpoint sequence, calculate the difference between the maximum and minimum values ​​of the dynamic pressure setpoint within a predefined time window to obtain the setpoint fluctuation amplitude; Step S37: When the fluctuation amplitude of the set value is greater than or equal to a second preset threshold, the feedforward pressure compensation amount determined in step S2 is multiplied by a first correction coefficient, the first correction coefficient being less than 1, and the value after multiplying by the first correction coefficient is used as a new feedforward pressure compensation amount for the algebraic superposition in step S32.

7. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 6, characterized in that, In step S34, when the initial corrected pressure setting value is between the upper limit of the pressure setting value and the lower limit of the pressure setting value, the initial corrected pressure setting value is determined as the dynamic pressure setting value.

8. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 7, characterized in that, Step S4 includes: Step S41: Obtain the real-time liquid level value; Step S42: Compare the real-time liquid level value with a number of preset liquid level thresholds to determine the liquid level interval to which the real-time liquid level value belongs. The number of liquid level thresholds divide the entire liquid level measurement range into a number of continuous and non-overlapping liquid level intervals. Step S43: Based on the liquid level range to which the real-time liquid level value belongs, query a preset gain mapping table to determine the corresponding adaptive gain coefficient. The gain mapping table defines the mapping relationship between different liquid level ranges and the adaptive gain coefficient, and the mapping relationship satisfies that the adaptive gain coefficient is negatively correlated with the real-time liquid level value.

9. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 8, characterized in that, Step S5 includes: Step S51: Obtain the dynamic pressure setpoint, the real-time pressure value, and the adaptive gain coefficient determined in step S4; Step S52: Calculate the difference between the dynamic pressure setpoint and the real-time pressure value to obtain the current pressure deviation; Step S53: Calculate the initial power adjustment amount of the booster pump based on the current pressure deviation; Step S54: Multiply the initial power adjustment amount by the adaptive gain coefficient to obtain the actual power adjustment amount.

10. The pure water pressure adaptive control method based on multi-sensor fusion according to claim 3, characterized in that, In step S22, the feedforward pressure compensation amount is positively correlated with the absolute value of the liquid level change rate.

Images