Real-time data based cyclic cooling water fouling trend prediction method and system

Abstract

The present application relates to the field of data processing, in particular to a circulating cooling water fouling trend prediction method and system based on real-time data, the method comprising: using near-wall flow velocity fluctuation adaptive adjustment window to extract temperature field anisotropy, shear dissipation estimation and ion activity lag angle, and combining with calcium concentration to form multi-field coupling characteristics; based on characteristic calculation, logically exclusive activation strength is activated, when the threshold value is exceeded, the temperature gradient weight is smoothly transferred to the shear weight to generate nucleation tendency coefficient; using the square of the nucleation tendency coefficient, the anisotropy and the shear complementary factor to calculate the entropy increase, constructing the exponential decay irreversible degree factor to constrain the prediction value, and combining with the basic fouling rate to obtain the fouling trend prediction value. The present application extracts high spatial resolution multi-field characteristics through adaptive window, dynamically weights and identifies the logic mutual exclusion of high-risk working conditions, reduces false alarms and misses, and realizes accurate and thermodynamic law-conforming quantitative prediction of fouling trend.

Description

Technical Field

[0001] This invention relates to the field of data processing. More specifically, this invention relates to a method and system for predicting scaling trends in circulating cooling water based on real-time data. Background Technology

[0002] In circulating cooling water systems, accurate prediction of scaling trends on the inner walls of heat exchanger tubes is crucial for ensuring heat transfer efficiency and guiding operational control. Existing technologies typically assess scaling risk based on parameters such as temperature, flow rate, pH value, and ion concentration at discrete measurement points on the wall, using empirical models or statistical methods.

[0003] However, under actual operating conditions, the thermal, flow, and chemical parameters near the heat exchanger tube wall are highly unevenly distributed in space, and there are complex coupling relationships between the various physical fields. Traditional prediction methods have limited adaptability to changes in operating conditions, and their prediction accuracy and reliability under complex or extreme conditions are insufficient to meet engineering requirements. Therefore, there is an urgent need for a method that can accurately and stably predict the scaling trend of circulating cooling water under conditions of multi-field coupling and variable operating conditions. Summary of the Invention

[0004] To address the technical problem that traditional prediction methods have limited adaptability to changes in operating conditions and that their prediction accuracy and reliability under complex or extreme conditions cannot meet engineering requirements, this invention provides solutions in the following aspects.

[0005] In the first aspect, the method and system for predicting scaling trends in circulating cooling water based on real-time data include: Obtain the standard deviation of near-wall flow velocity pulsation at each node on the inner wall of the heat exchange tube, and adaptively adjust the axial span of the sliding window based on the pulsation standard deviation to obtain the adjusted window; Within the adjusted window, the anisotropy of thermal boundary layer perturbation is obtained by local gradient directional divergence quantification of the temperature field. The estimated values ​​of shear strain rate turbulent kinetic energy dissipation and hysteresis tangent of ion activity field are extracted and combined with the calcium ion concentration of the mainstream water body to form a set of intrinsic property parameters of the wall. Based on the set of attribute parameters, the temperature deviation degree, shear constraint factor and acid-base drive degree are calculated respectively, and the product of the complementary factor of the temperature deviation degree and shear constraint factor and the acid-base drive degree is used as the logical mutual exclusion activation strength. When the logic mutual exclusion activation intensity exceeds the preset logic mutual exclusion activation intensity threshold, a portion of the temperature gradient weight is transferred to the shear strain rate weight using a smooth transition factor to generate dynamic weights. The nucleation tendency coefficient is determined by the dynamic weights and the set of attribute parameters; The degree of entropy increase in the local system is determined by the product of the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor, and the irreversibility factor of the exponential decay fouling process is constructed accordingly. The scale trend prediction value is obtained by multiplying the irreversibility factor, the nucleation tendency coefficient, and the basic scaling rate determined by the main fluid temperature.

[0006] Optionally, the step of adaptively adjusting the axial span of the sliding window based on the pulsation standard deviation includes: Based on the near-wall flow velocity pulsation standard deviation, combined with the basic physical span and the preset window expansion and contraction sensitivity coefficient, the axial span of the sliding window is generated. The basic physical span is determined based on the characteristic dimensions of the heat exchange pipe and the typical boundary layer thickness, and the window expansion and contraction sensitivity coefficient is given through field calibration.

[0007] Optionally, the step of obtaining the anisotropy of thermal boundary layer perturbation by local gradient direction divergence quantification of the temperature field within the adjusted window includes: Calculate the L2 norm of the gradient vector of the temperature field within the window in the mainstream direction, and the L2 norm of the gradient vector in the wall normal direction; The ratio of the mainstream gradient norm and the wall normal gradient norm to a preset minimum positive number is calculated, and the logarithm of this ratio to the base 10 is taken as the anisotropy of the thermal boundary layer perturbation.

[0008] Optionally, the extraction of the estimated value of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent of the ion activity field includes: Based on the near-wall average velocity gradient and turbulent shear stress within the window, the estimated value of the shear strain rate turbulent kinetic energy dissipation is calculated. The pH time series within the window is cross-correlated with the gradient time series of the temperature field, and the hysteresis tangent of the ion activity field is determined based on the tangent of the phase angle corresponding to the cross-correlation delay.

[0009] Optionally, the set of intrinsic property parameters of the wall surface, which is composed of the calcium ion concentration in the mainstream water body, includes: The intrinsic property parameter set of the wall is a four-dimensional feature vector composed of the anisotropy of the thermal boundary layer perturbation, the estimated value of the shear strain rate turbulent kinetic energy dissipation, the hysteresis tangent of the ion activity field, and the calcium ion concentration of the mainstream water.

[0010] Optionally, the calculation of the temperature deviation, shear constraint factor, and acid-base driving degree includes: The anisotropy of the thermal boundary layer perturbation is normalized and mapped to obtain the temperature deviation with a value between zero and one. The shear constraint factor is obtained by normalizing the estimated value of turbulent kinetic energy dissipation at the shear strain rate; By combining the hysteresis tangent of the ion activity field and the calcium ion concentration in the mainstream water body, the chemical supersaturation state is evaluated, and the acid-base driving degree is obtained.

[0011] Optionally, when the logical mutual exclusion activation intensity exceeds a preset logical mutual exclusion activation intensity threshold, transferring a portion of the temperature gradient weight to the shear strain rate weight using a smooth transition factor includes: Preset logical mutual exclusion activation strength threshold; An S-shaped function with the logical mutual exclusion activation intensity as the independent variable is constructed using a smooth transition sharpness factor. The function value approaches zero when the logical mutual exclusion activation intensity is less than the threshold and approaches one when the logical mutual exclusion activation intensity is greater than the threshold.

[0012] Optionally, generating dynamic weights includes: Obtain the preset basic weights for temperature gradient, shear strain rate, and acid-base correlation; Define the maximum adjustment range. Subtract the product of the smoothing transition factor and the maximum adjustment range from the basic weight of the temperature gradient. The difference is the adjusted temperature gradient weight. Add the product of the smoothing transition factor and the maximum adjustment range to the basic weight of the shear strain rate. The sum is the adjusted shear strain rate weight. The basic weight of the acid-base correlation remains unchanged.

[0013] Optionally, the step of determining the degree of entropy increase in the local system by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor, and constructing an exponentially decaying scale formation irreversibility factor accordingly, includes: The degree of entropy increase of the local system is calculated as the product of the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the difference between the coefficient and the shear constraint factor. The irreversibility factor is the negative of the natural exponential function, the preset constraint strength adjustment coefficient multiplied by the ratio of the local system entropy increase to the reference entropy gain constant.

[0014] Secondly, the circulating cooling water scaling trend prediction system based on real-time data includes: an adaptive window parameter set generation module, used to obtain the near-wall velocity pulsation standard deviation of each node on the inner wall of the heat exchange tube, adaptively adjust the axial span of the sliding window according to the pulsation standard deviation, and obtain the thermal boundary layer perturbation anisotropy by local gradient direction divergence quantification of the temperature field within the adjusted window, extract the estimated value of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent value of ion activity field, and form a wall intrinsic property parameter set with the calcium ion concentration of the mainstream water body; The nucleation tendency dynamic evaluation module is used to calculate the temperature deviation degree, shear constraint factor and acid-base driving degree according to the intrinsic property parameter set of the wall, and use the product of the complementary factor of the temperature deviation degree and the shear constraint factor and the acid-base driving degree as the logical mutual exclusion activation intensity. When the logical mutual exclusion activation intensity exceeds the preset logical mutual exclusion activation intensity threshold, a portion of the temperature gradient weight is transferred to the shear strain rate weight to generate dynamic weight by means of a smooth transition factor, and the nucleation tendency coefficient is determined by the dynamic weight and the property parameter set. The entropy increase constraint prediction module is used to determine the degree of entropy increase of the local system by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor. Based on this, it constructs an exponential decay type scaling process irreversibility factor, and multiplies the irreversibility factor, the nucleation tendency coefficient, and the basic scaling rate determined by the main fluid temperature to obtain a scaling trend prediction value.

[0015] The present invention has the following effects: 1. This invention utilizes the near-wall velocity pulsation standard deviation to dynamically adjust the axial span of the sampling window, enabling the window scale to adaptively match turbulent pulsation changes and effectively avoiding the aliasing of the boundary layer and the mainstream physical field caused by a fixed window. The anisotropy of the thermal boundary layer disturbance extracted within the window accurately characterizes the temperature gradient difference between the normal and mainstream directions. Combined with the estimation of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent of the ion activity field lag angle, the dispersed measurement point data is reconstructed into spatially high-resolution four-dimensional multi-field coupled features, improving the characterization precision and reliability of input information for local fouling nucleation hotspots on the wall.

[0016] 2. This invention generates temperature deviation, shear constraint factor, and acid-base driving force through membership mapping. The product of these three factors defines the logically mutually exclusive activation intensity, accurately identifying the low-frequency, high-risk scaling complex state of high temperature difference gradient, low shear dissipation, and high pH. When the logically mutually exclusive activation intensity exceeds a threshold, the smooth transition factor drives a partial transfer of the temperature gradient weight to the shear strain rate weight, forming a non-orthogonal dynamic weight. This strengthens the evaluation component of flow inhibition factors in nucleation tendency, avoids overestimation of the high temperature difference signal, thereby reducing false alarms and missed alarms, and ensuring that the nucleation tendency coefficient truly reflects the local scaling driving force.

[0017] 3. This invention introduces the degree of local system entropy increase, calculated by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint. Based on this, an exponentially decaying irreversibility factor for the fouling process is constructed as a nonlinear correction term. When local dissipation intensifies, the irreversibility factor decays exponentially, applying a smoothing suppression to excessively high tendency values ​​and avoiding non-physical exponential divergence and spurious peaks in the prediction results. In the real high-risk region, the factor remains high to preserve sensitive response. This constraint helps the predicted values ​​always conform to the thermodynamic irreversibility law, improving quantitative reliability and engineering practicality. Attached Figure Description

[0018] Figure 1 This is a flowchart of steps S1-S3 in the method for predicting scaling trends in circulating cooling water based on real-time data according to an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0020] Reference Figure 1 The method for predicting scaling trends in circulating cooling water based on real-time data includes steps S1-S3, as detailed below: S1: Obtain the near-wall velocity pulsation standard deviation of each node on the inner wall of the heat exchange tube, adaptively adjust the axial span of the sliding window according to the pulsation standard deviation, obtain the anisotropy of thermal boundary layer disturbance by local gradient direction divergence quantification of the temperature field within the adjusted window, extract the estimated value of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent of ion activity field, and form the intrinsic property parameter set of the wall with the calcium ion concentration of the mainstream water body.

[0021] Multiple spatial points are discretized along the axial and circumferential directions of the inner wall of the heat exchange tube at certain intervals and defined as nodes. In this embodiment of the invention, the axial spacing between adjacent nodes is 0.1 meters, and one node is set every 90° in the circumferential direction, that is, there are 4 base points at each axial position. Each node corresponds to a set of wall-embedded sensors, including a high-frequency flow rate sensor, a temperature sensor, and a pH sensor. The node spacing can be adjusted according to the heat exchange tube diameter, sensor array density, and computing resources, with a typical range of 0.05 meters to 0.2 meters.

[0022] For each node, a high-frequency velocity signal is first acquired using a wall array velocity sensor within a preset time interval. This preset time interval is determined based on the dominant frequency scale of the near-wall turbulent pulsation and should contain at least 10 turbulent vortex overturning cycles to ensure the statistical stability of the pulsation standard deviation. The local pulsation standard deviation of the near-wall velocity within this time interval is then calculated.

[0023] Furthermore, the axial span of the sampling window is adaptively adjusted by superimposing the product of the local pulsation standard deviation and the window expansion / contraction sensitivity coefficient on the basic physical span determined by the characteristic scale of the heat exchanger tube and the typical boundary layer thickness, thus obtaining the adjusted window:

[0024] In the formula, The axial span of the adaptively adjusted sampling window. Based on the basic physical span, This is the window scaling sensitivity coefficient, with a value ranging from 0.1 to 0.5. This represents the local fluctuation standard deviation of the near-wall velocity. The sliding window moves along the axial direction in steps of 0.1 meters, with the center of each window aligned with the current node.

[0025] Based on the above calculations Within the defined window, extract the following features respectively: (a) Anisotropy of thermal boundary layer perturbation in the temperature field: Calculate the gradient vectors of the temperature field in the mainstream direction and the wall normal direction within the window, and obtain their L2 norms. Then, calculate the anisotropy of the thermal boundary layer perturbation based on these two L2 norms.

[0026] In the formula, The anisotropy of the aforementioned thermal boundary layer perturbation is given. The L2 norm of the gradient vector in the mainstream direction. Let L2 norm be the gradient vector of the wall normal. To prevent zero from being the smallest positive number. When A value significantly greater than 0 indicates that the temperature drop gradient in the normal direction of the wall is much stronger than the temperature homogenization trend in the mainstream direction, and the thermodynamic driving force of scaling is prominent.

[0027] (ii) Estimates of shear strain rate and turbulent kinetic energy dissipation in the near-wall flow field: First, determine the wall friction velocity, which can be obtained by fitting the average flow velocity distribution within the window to the logarithmic law.

[0028] Then calculate:

[0029] In the formula, This is an estimate of the turbulent kinetic energy dissipation at the shear strain rate. For fluid density, Kármán's constant, This represents the normal distance from the node to the wall. The velocity is the wall friction velocity. A higher value indicates strong fluid shearing and rapid energy dissipation within that window, thus inhibiting crystal nucleus adhesion; conversely, a lower value weakens this inhibition. In practical applications, It can also be determined by wall shear stress. Obtained indirectly.

[0030] (iii) The hysteresis tangent of the ion activity field: Obtain the pH time series and temperature gradient norm series within the current sliding window, perform cross-correlation calculation on these two series, determine the time shift that maximizes the cross-correlation coefficient, and after normalization with the sampling interval, calculate:

[0031] In the formula, It is the hysteresis tangent value. This is the ratio of the aforementioned time shift to the sampling interval. The hysteresis tangent is used to measure the degree of non-equilibrium response of the chemical potential field. The larger the positive value, the more significant the hysteresis of the ion activity field relative to the thermal boundary layer perturbation, and the easier it is to maintain the supersaturated state.

[0032] (iv) Calcium ion concentration in mainstream water bodies: Read the mainstream water body calcium ion concentration output by the online calcium ion analyzer. This value is a global variable and does not depend on the window.

[0033] Finally, the four features calculated above are combined into a local thermal, fluid, and chemical multi-field coupled feature vector for the node. Then, the local thermal, fluid, and chemical multi-field coupled feature vectors of all nodes are used to construct the intrinsic property parameter set of the wall.

[0034] S2: Evaluate the logically exclusive activation strength of multi-field coupling based on the attribute parameter set. When the logically exclusive activation strength exceeds the limit, adjust the contribution ratio of each physical quantity, and then determine the nucleation tendency coefficient.

[0035] After obtaining the eigenvectors characterizing the coupled states of local thermal, fluidic, and chemical fields at a node, it is difficult to identify the complex conditions unique to high fouling risk by directly applying linear weighting based on a single component.

[0036] Because the most dangerous scaling conditions at the actual heat exchanger tube wall are a combination of significant anisotropy of temperature gradient, weakened near-wall shear, and driven by chemical potential field strength, the probability of all three being true at the same time is much lower than that of any single extreme case. The fixed weights under the conventional independent assumption will inevitably lead to the dilution of the contribution of such logically mutually exclusive states.

[0037] Therefore, it is necessary to construct a processing procedure that can characterize the strength of the composite state and dynamically rearrange the weights based on the physical meaning contained in the feature vector.

[0038] For each component in the feature vector, three dimensionless membership degrees are obtained through mapping:

[0039] In the formula, For the degree of temperature deviation, The anisotropy of the aforementioned thermal boundary layer perturbation is given. For calibration reference value, an example value of 0.5 is used. When Approaching 0 Approaching 0, When significantly increased A value close to 1 indicates that the thermodynamic driving force for scaling is extremely strong.

[0040]

[0041] In the formula, The shear constraint factor. This is an estimate of the turbulent kinetic energy dissipation at the shear strain rate. For reference dissipation constants, calibrated according to field conditions, the typical range is 10~500 W / m³. When Far greater than hour A value close to 1 indicates strong inhibition of crystal nucleus attachment by a high-shear environment; conversely, the inhibition weakens.

[0042]

[0043] In the formula, For acid-base drive degree, This is the mixed weighting coefficient, with a value of 0.6. It is the hysteresis tangent value. The concentration of calcium ions in the mainstream water body The reference value for calcium ion saturation is 120 mg / L. The closer the value is to 1, the more favorable the local chemical environment is for the supersaturation precipitation of calcium carbonate.

[0044] Furthermore, based on the aforementioned temperature deviation, shear constraint factor, and acid-base driving degree, the logic mutual exclusion activation strength is calculated together:

[0045] In the formula, For logically mutually exclusive activation strength, For the degree of temperature deviation, The shear constraint factor. This represents the acid-base driving degree. It is a complementary operation to the shear constraint factor, and its physical meaning is the degree of loss of shear inhibition, which is related to... The high-temperature difference characterized by the combination of driving forces corresponds precisely to a state where strong driving forces and low inhibitory forces coexist. This further restricts the activation of this state to conditions of high chemical potential. Therefore, Significance is only achieved when all three conditions are met simultaneously: high anisotropy of temperature gradient, weak shear dissipation, and strong chemical driving force.

[0046] Furthermore, a threshold for the activation intensity of logical mutual exclusion is set, for example, a value of 0.7. When the activation intensity of logical mutual exclusion is greater than the threshold, the contribution ratio of each physical quantity is adjusted. Specifically, a smooth transition factor is used to transfer a portion of the temperature gradient weight to the shear strain rate weight.

[0047] Specifically, under normal operating conditions, the preset basic weights are: temperature gradient basic weight 0.45, shear strain rate basic weight 0.25, and acid-base correlation basic weight 0.30. When entering a high-risk condition of logical mutual exclusion, a smooth transition factor is introduced, the value of which is close to 0 when the activation intensity of logical mutual exclusion just exceeds the threshold, and gradually approaches 1 as the activation intensity of logical mutual exclusion increases.

[0048] The above smooth transition factor is calculated as follows:

[0049] In the formula, As a smooth transition factor, This represents the activation strength of logical mutual exclusion.

[0050] The maximum adjustment range is defined as 0.15. The dynamic weights are then determined according to the following rules: The adjusted temperature gradient weight is obtained by subtracting the product of the smooth transition factor and the maximum adjustment range from the basic weight of the temperature gradient. The adjusted shear strain rate weight is obtained by adding the product of the smooth transition factor and the maximum adjustment range to the basic weight of the shear strain rate. The basic weight of the acid-base correlation remains unchanged. Thus, as the activation intensity of the logical mutual exclusion increases, the temperature gradient weight gradually decreases, while the shear strain rate weight increases by the same amount, and the total weight always remains at 1.

[0051] Finally, the components of the eigenvector constructed in S1 are normalized using Min-Max to obtain the normalized eigenvector. The adjusted temperature gradient weight, adjusted shear strain rate weight, and fixed acid-base correlation weight obtained above are weighted and summed with the normalized eigenvector to obtain a comprehensive score. The comprehensive score is then mapped to the interval of 0 to 1 using the Logistic function to obtain the nucleation tendency coefficient. The higher the nucleation tendency coefficient, the greater the probability that a stable nucleus will form at the current moment.

[0052] S3: The degree of entropy increase of the local system is determined by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor, and an irreversibility factor of the exponentially decaying scaling process is constructed accordingly; the irreversibility factor, the nucleation tendency coefficient, and the basic scaling rate determined by the main fluid temperature are multiplied to obtain the scaling trend prediction value.

[0053] Although the nucleation tendency coefficient has incorporated the combined effects of temperature gradient, flow shear and chemical environment through a logically mutually exclusive weight competition mechanism, in an open thermodynamic system, a high nucleation tendency means an increased probability of the formation of local ordered structures. This process is inevitably accompanied by the generation of entropy. If the scaling rate is directly extrapolated linearly from the nucleation tendency coefficient, the predicted value may show a non-physical exponential growth at micro-elements with extremely strong anisotropy of temperature gradient and almost no shear.

[0054] Therefore, it is necessary to impose an irreversible constraint on the nucleation tendency coefficient that is related to the degree of local entropy increase in order to limit the upper bound of the prediction results and make them conform to the thermodynamic trend.

[0055] The main sources of irreversible entropy generation in the nucleation process include irreversible heat dissipation caused by temperature gradient and the momentum dissipation capacity of the viscous sublayer weakened by insufficient shear. These two factors, together with the intensity of the nucleation tendency itself, determine the level of entropy increase in the local system.

[0056] Therefore, the degree of entropy increase of a local system is defined as the product of the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the difference between 1 and the shear constraint factor.

[0057] The introduction of the square of the nucleation tendency coefficient is due to the nonlinear characteristics of nucleation's consumption of free energy; the higher the tendency, the more intense the entropy generation induced by a unit increase in tendency. The anisotropy of the thermal boundary layer perturbation directly reflects the irreversible contribution of the wall-normal temperature gradient; the larger the value, the greater the entropy generation rate on the heat conduction path. The difference between 1 and the shear constraint factor lies in the fact that when the shear constraint factor is high, the turbulent kinetic energy dissipation itself has already consumed a large amount of energy in the process of resisting nucleus attachment, and the local effective free energy is relatively limited. Conversely, when the shear constraint factor is low, part of the entropy generation that should have been shared by viscous dissipation is now borne by the nucleation process, further aggravating the entropy increase of the local system.

[0058] Further construct the irreversibility factor:

[0059] In the formula, It is an irreversible factor. This is the constraint strength adjustment coefficient, with a value of 1. The degree of entropy increase in the local system. As a reference entropy gain constant, it was calibrated offline based on the heat exchanger tube wall material and typical water conditions. It is an exponential function with the natural constant e as its base.

[0060] when much smaller When the exponential term approaches 0, A value close to 1 has almost no impact on the prediction, which is close to 0. When the value is close to 1, it imposes almost no constraint on the prediction, allowing the scaling tendency to be directly reflected in the output; conversely, when... Significantly increased to the level of When comparable to or exceeding the level, the exponential decay effect causes The rapid decrease physically reflects the reduced ability of the system to maintain orderly growth under high entropy conditions, thereby suppressing the non-physical divergence that may be caused by the pulse spike of the nucleation tendency coefficient.

[0061] Furthermore, the scale trend prediction value is obtained by multiplying the irreversibility factor, nucleation tendency coefficient, and basic scaling rate. When the node is under true high-risk conditions, the high nucleation tendency coefficient and the combination of thermal boundary layer perturbation anisotropy and shear constraint factor prevent excessive increase in local system entropy, maintain a high irreversibility factor, and the scale trend prediction value sensitively responds to local scaling driving forces. When the node is under transient noise peaks or physically unsustainable high temperature differences, the local system entropy increases sharply, and the exponential decay of the irreversibility factor strongly restrains the scale trend prediction value, filtering out spurious high values. The scale trend prediction value obtained in this way is a quantitative measure of the thermodynamically consistent scale trend of the analyzed node at the current moment, directly serving subsequent risk visualization and control decisions.

[0062] The basic scaling rate is determined based on the main fluid temperature using an offline calibrated scaling rate curve. This curve is obtained through laboratory static scaling experiments, measuring the scale growth rate of a standard wall material at different temperatures. In this embodiment of the invention, a monotonically increasing rate curve is calibrated within a temperature range of 20°C to 60°C; the specific values ​​are determined by the implementer based on the actual heat exchanger tube material and water quality conditions.

[0063] The present invention also provides a circulating cooling water scaling trend prediction system based on real-time data. The system includes a processor and a memory, the memory storing computer program instructions. When the computer program instructions are executed by the processor, the circulating cooling water scaling trend prediction method based on real-time data according to the first aspect of the present invention is implemented.

[0064] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and will not be described in detail here.

[0065] It should be noted that those skilled in the art can make various modifications and improvements without departing from the inventive concept, and these all fall within the scope of protection of this invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

Claims

1. A method for predicting scaling trends in circulating cooling water based on real-time data, characterized in that, include: Obtain the standard deviation of near-wall flow velocity pulsation at each node on the inner wall of the heat exchange tube, and adaptively adjust the axial span of the sliding window based on the pulsation standard deviation to obtain the adjusted window; Within the adjusted window, the anisotropy of thermal boundary layer perturbation is obtained by local gradient directional divergence quantification of the temperature field. The estimated values ​​of shear strain rate turbulent kinetic energy dissipation and hysteresis tangent of ion activity field are extracted and combined with the calcium ion concentration of the mainstream water body to form a set of intrinsic property parameters of the wall. Based on the set of attribute parameters, the temperature deviation degree, shear constraint factor and acid-base drive degree are calculated respectively, and the product of the complementary factor of the temperature deviation degree and shear constraint factor and the acid-base drive degree is used as the logical mutual exclusion activation strength. When the logic mutual exclusion activation intensity exceeds the preset logic mutual exclusion activation intensity threshold, a portion of the temperature gradient weight is transferred to the shear strain rate weight using a smooth transition factor to generate dynamic weights. The nucleation tendency coefficient is determined by the dynamic weights and the set of attribute parameters; The degree of entropy increase in the local system is determined by the product of the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor, and the irreversibility factor of the exponential decay fouling process is constructed accordingly. The scale trend prediction value is obtained by multiplying the irreversibility factor, the nucleation tendency coefficient, and the basic scaling rate determined by the main fluid temperature.

2. The method for predicting scaling trends in circulating cooling water based on real-time data according to claim 1, characterized in that, The adaptive adjustment of the axial span of the sliding window based on the pulsation standard deviation includes: Based on the near-wall flow velocity pulsation standard deviation, combined with the basic physical span and the preset window expansion and contraction sensitivity coefficient, the axial span of the sliding window is generated. The basic physical span is determined based on the characteristic dimensions of the heat exchange pipe and the typical boundary layer thickness, and the window expansion and contraction sensitivity coefficient is given through field calibration.

3. The method for predicting the scaling trend of circulating cooling water based on real-time data according to claim 1, characterized in that, Within the adjusted window, the anisotropy of the thermal boundary layer perturbation is obtained by local gradient direction divergence quantification of the temperature field, including: Calculate the L2 norm of the gradient vector of the temperature field within the window in the mainstream direction, and the L2 norm of the gradient vector in the wall normal direction; The ratio of the mainstream gradient norm and the wall normal gradient norm to a preset minimum positive number is calculated, and the logarithm of this ratio to the base 10 is taken as the anisotropy of the thermal boundary layer perturbation.

4. The method for predicting scaling trends in circulating cooling water based on real-time data according to claim 1, characterized in that, The extraction of the estimated values ​​of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent of the ion activity field includes: Based on the near-wall average velocity gradient and turbulent shear stress within the window, the estimated value of the shear strain rate turbulent kinetic energy dissipation is calculated. The pH time series within the window is cross-correlated with the gradient time series of the temperature field, and the hysteresis tangent of the ion activity field is determined based on the tangent of the phase angle corresponding to the cross-correlation delay.

5. The method for predicting the scaling trend of circulating cooling water based on real-time data according to claim 1, characterized in that, The intrinsic property parameter set of the wall surface, which is composed of the calcium ion concentration of the mainstream water body, includes: The intrinsic property parameter set of the wall is a four-dimensional feature vector composed of the anisotropy of the thermal boundary layer perturbation, the estimated value of the shear strain rate turbulent kinetic energy dissipation, the hysteresis tangent of the ion activity field, and the calcium ion concentration of the mainstream water.

6. The method for predicting scaling trends in circulating cooling water based on real-time data according to claim 1, characterized in that, The calculation of temperature deviation, shear constraint factor, and acid-base drive degree includes: The anisotropy of the thermal boundary layer perturbation is normalized and mapped to obtain the temperature deviation with a value between zero and one. The shear constraint factor is obtained by normalizing the estimated value of turbulent kinetic energy dissipation at the shear strain rate; By combining the hysteresis tangent of the ion activity field and the calcium ion concentration in the mainstream water body, the chemical supersaturation state is evaluated, and the acid-base driving degree is obtained.

7. The method for predicting the scaling trend of circulating cooling water based on real-time data according to claim 1, characterized in that, When the logical mutual exclusion activation intensity exceeds a preset logical mutual exclusion activation intensity threshold, a portion of the temperature gradient weight is transferred to the shear strain rate weight using a smooth transition factor, including: Preset logical mutual exclusion activation strength threshold; An S-shaped function with the logical mutual exclusion activation intensity as the independent variable is constructed using a smooth transition sharpness factor. The function value approaches zero when the logical mutual exclusion activation intensity is less than the threshold and approaches one when the logical mutual exclusion activation intensity is greater than the threshold.

8. The method for predicting scaling trends in circulating cooling water based on real-time data according to claim 1, characterized in that, The generation of dynamic weights includes: Obtain the preset basic weights for temperature gradient, shear strain rate, and acid-base correlation; Define the maximum adjustment range. Subtract the product of the smoothing transition factor and the maximum adjustment range from the basic weight of the temperature gradient. The difference is the adjusted temperature gradient weight. Add the product of the smoothing transition factor and the maximum adjustment range to the basic weight of the shear strain rate. The sum is the adjusted shear strain rate weight. The basic weight of the acid-base correlation remains unchanged.

9. The method for predicting scaling trends in circulating cooling water based on real-time data according to claim 1, characterized in that, The method of determining the degree of entropy increase in the local system by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor, and constructing an exponentially decaying scale formation irreversibility factor accordingly, includes: The degree of entropy increase of the local system is calculated as the product of the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the difference between the coefficient and the shear constraint factor. The irreversibility factor is the negative of the natural exponential function, the preset constraint strength adjustment coefficient multiplied by the ratio of the local system entropy increase to the reference entropy gain constant.

10. A circulating cooling water scaling trend prediction system based on real-time data, characterized in that, include: An adaptive window parameter set generation module is used to obtain the near-wall velocity pulsation standard deviation of each node on the inner wall of the heat exchange tube, adaptively adjust the axial span of the sliding window according to the pulsation standard deviation, and obtain the thermal boundary layer perturbation anisotropy by local gradient direction divergence quantification of the temperature field within the adjusted window. It also extracts the estimated value of shear strain rate turbulent kinetic energy dissipation and the hysteresis tangent value of ion activity field, and forms a wall intrinsic property parameter set with the calcium ion concentration of the mainstream water body. The nucleation tendency dynamic evaluation module is used to calculate the temperature deviation degree, shear constraint factor and acid-base driving degree according to the intrinsic property parameter set of the wall, and use the product of the complementary factor of the temperature deviation degree and the shear constraint factor and the acid-base driving degree as the logical mutual exclusion activation intensity. When the logical mutual exclusion activation intensity exceeds the preset logical mutual exclusion activation intensity threshold, a portion of the temperature gradient weight is transferred to the shear strain rate weight to generate dynamic weight by means of a smooth transition factor, and the nucleation tendency coefficient is determined by the dynamic weight and the property parameter set. The entropy increase constraint prediction module is used to determine the degree of entropy increase of the local system by multiplying the square of the nucleation tendency coefficient, the anisotropy of the thermal boundary layer perturbation, and the complementary factor of the shear constraint factor. Based on this, it constructs an exponential decay type scaling process irreversibility factor, and multiplies the irreversibility factor, the nucleation tendency coefficient, and the basic scaling rate determined by the main fluid temperature to obtain a scaling trend prediction value.

Images