An experimental method for separating and purifying dye wastewater based on freeze purification technology.

Abstract

This invention relates to the field of wastewater testing and detection technology, specifically a test method for separating and purifying dye wastewater based on freeze purification technology. The method includes: constructing a micro-contamination layer, a clean layer, a bubble layer, and a concentration layer as the freeze front advances; analyzing the current freeze front using the average freezing rate to determine the displacement response coefficient of each layer; determining the time-cumulative equivalent factor using the interval time and layer thickness when configuring the displacement response coefficients, and judging the displacement response results of each layer as the freeze front advances; recording the trend difference between pigment concentration and the freeze front advance, and splitting the equivalent factor into two categories related to pigment concentration and the freeze front; combining the equivalent factors of pigment concentration and the freeze front to obtain a parameter fitting combination; and deriving the optimal freezing parameters based on the parameter fitting combination; thus achieving both efficiency and stability in freeze analysis.

Description

Technical Field

[0001] This invention relates to the field of wastewater testing and detection technology, specifically a test method for separating and purifying dye wastewater based on freeze purification technology. Background Technology

[0002] In artificial freezing purification technology, progressive continuous freezing is considered a more promising approach. This method involves gradually forming an integral structure of ice crystals along a direction perpendicular to the cooling surface, removing the solute into the liquid phase, and forming an ice layer of moderate thickness. This ice layer is easy to separate, has high purity, and carries very little solute. Nevertheless, this technology still faces several key challenges: (1) how to improve the purity of ice crystals to achieve better purification results; (2) how to reduce the impact of supercooling in the initial cooling stage to avoid the formation of a micro-contamination layer; and (3) how to improve processing efficiency and reduce energy consumption through technological means.

[0003] For example, Chinese Patent Publication No. CN114062417A discloses a wastewater underground infiltration device simulating freeze-thaw cycles and its usage method. This wastewater underground infiltration device, capable of simulating freeze-thaw cycles, includes a SWIS soil column, a refrigeration cycle system, a water injection system, and an online soil parameter detection system. The refrigeration cycle system includes several layers of refrigeration copper pipes arranged sequentially from bottom to top on the upper part of the SWIS soil column for controlling the freeze-thaw temperature of the surface soil. The water injection system includes a cross-shaped water distribution pipe located in the middle of the SWIS soil column, which injects water into the SWIS soil column to simulate domestic sewage. The online soil parameter detection system includes several soil parameter rapid measuring instruments evenly arranged from bottom to top inside the SWIS soil column for real-time collection of soil parameters. This wastewater underground infiltration device simulating freeze-thaw cycles and its usage method can achieve control of the freeze-thaw temperature of the SWIS layer, thereby accurately revealing the impact of the freeze-thaw cycle of the surface soil on the wastewater treatment performance of the SWIS.

[0004] Existing technologies describe methods for separating wastewater through freezing and for purifying wastewater through repeated freezing and thawing. However, when purifying wastewater, it is easy to overlook the changes in the clean layer during freezing, leading to unclear layer monitoring, missing parameter correlations, and incorrect parameter configuration during freezing purification, resulting in reduced wastewater purification effect and unstable purification process. Summary of the Invention

[0005] To solve the above-mentioned technical problems, the technical solution adopted by the present invention is: an experimental method for separating and purifying dye wastewater based on freeze purification technology, comprising: S1, using image analysis to construct a micro-contamination layer, a clean layer, a bubble layer and a concentration layer during the advancement of the freeze front, and recording the thickness and pigment concentration of each layer.

[0006] S2, based on the freezing rate under freezing analysis, analyze the current freezing front with the average freezing rate to determine the displacement response coefficient of each layer.

[0007] S3. Using the interval time and layer thickness when configuring the displacement response coefficient, determine the equivalent factor in time accumulation; based on the equivalent factor, the value range of the upper and lower clean layers, determine the displacement response results of each layer as the freezing front advances.

[0008] S4 takes the displacement response result after judgment as input, records the trend difference between pigment concentration and the advance of the frozen front, and splits the equivalent factor into two categories related to pigment concentration and frozen front, respectively labeling the functional relationship between pigment concentration and frozen front change.

[0009] S5. By combining the equivalent factors of pigment concentration and freezing front, a parameter fitting combination under trend fitting is obtained. Based on the parameter fitting combination, the optimal freezing parameters during freezing are derived.

[0010] The beneficial effects of this invention are as follows: First, this invention constructs a micro-contamination layer, a clean layer, a bubble layer, and a concentration layer during the advancement of the freezing front, recording the thickness and pigment concentration of each layer; further, it transforms the layer data at each moment into a layer change sequence, and establishes a fitting relationship between the average freezing rate and the initial concentration and cold end temperature, using the average freezing rate and cold end temperature as constraints, and synchronizes it to the layer labeling index to achieve monitoring and processing of freezing purification; it can not only record the evolution of the thickness and pigment concentration of each layer over time in real time, but also bind the layer changes with freezing parameters through the fitting relationship, ensuring the reliability of the basic data for subsequent analysis.

[0011] Second, this invention identifies the changes in the micro-contamination layer and the clean layer using a single variable rule, retrieves the cold end temperature and initial concentration, associates them with the average freezing rate to form a parameter group, and records the freezing time and ice formation rate corresponding to the average freezing rate to obtain the displacement response coefficient under the current scenario, thereby achieving quantitative analysis of the changes in the clean layer.

[0012] Third, this invention divides the stratified time series into subsequences and sets equal-effect factors according to scenarios with consistent and inconsistent timestamps to quantify the response of each layer thickness as the freezing front advances. This explains the growth of the clean layer as the freezing front advances, ensuring that the output displacement response results fit the clean layer optimization requirements. It provides reliable input data for subsequent trend analysis of pigment concentration and freezing front, avoiding invalid data from interfering with subsequent modeling.

[0013] Fourth, this invention records the changing trends of pigment concentration and freezing front, separating the equivalent effect factors dominated by pigment concentration and those dominated by freezing front; it fits the functional relationship between the two factors using the least squares method; it derives the freezing period by combining the timestamps of the displacement response results, and identifies the reasons for the differences in their trends; then, it uses frequent itemset mapping of cleanroom indicators, and extracts the optimal freezing parameters by using a descending order of scores and a descending order of the total frequency of cold end temperature and initial concentration, ensuring that the optimal parameters simultaneously satisfy high removal rates of COD, TOC, and color, avoiding poor overall purification effect caused by optimization of a single indicator; thus improving the efficiency and stability of freezing analysis. Attached Figure Description

[0014] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0015] Figure 1 This is a schematic diagram of an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0016] Figure 2 This is a schematic diagram of temperature changes in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0017] Figure 3 This is a schematic diagram illustrating the changes in the micro-pollution layer in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0018] Figure 4 This is a schematic diagram illustrating the changes in the cleanroom layer in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0019] Figure 5 This is a schematic diagram showing the changes in cold end temperature and ice formation rate of an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0020] Figure 6 This is a schematic diagram showing the changes in initial solution concentration and ice formation rate of an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0021] Figure 7 This is a schematic diagram showing the variation of the thickness of each layer in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0022] Figure 8 This is a schematic diagram of the freezing effect of the upper layer of the cleanroom in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0023] Figure 9 This is a schematic diagram of the freezing effect of the lower layer of the cleanroom layer in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0024] Figure 10This is a schematic diagram of step S1 of an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0025] Figure 11 This is a schematic diagram of step S3 in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0026] Figure 12 This is a schematic diagram of step S4 in an experimental method for separating and purifying dye wastewater based on freeze purification technology.

[0027] Figure 13 This is a schematic diagram of step S5 of an experimental method for separating and purifying dye wastewater based on freeze purification technology. Detailed Implementation

[0028] The embodiments of the present invention are described in detail below. The embodiments described below are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention. Where specific techniques or conditions are not specified in the embodiments, they shall be performed in accordance with the techniques or conditions described in the literature in the art or in accordance with the product manual.

[0029] See Figure 1 An experimental method for separating and purifying dye wastewater based on freeze purification technology includes: S1, using image analysis to construct a micro-contamination layer, a clean layer, a bubble layer and a concentration layer during the advance of the freeze front, and recording the thickness and pigment concentration of each layer.

[0030] S2, based on the freezing rate under freezing analysis, analyze the current freezing front with the average freezing rate to determine the displacement response coefficient of each layer.

[0031] S3. Using the interval time and layer thickness when configuring the displacement response coefficient, determine the equivalent factor in time accumulation; based on the equivalent factor, the value range of the upper and lower clean layers, determine the displacement response results of each layer as the freezing front advances.

[0032] S4 takes the displacement response result after judgment as input, records the trend difference between pigment concentration and the advance of the frozen front, and splits the equivalent factor into two categories related to pigment concentration and frozen front, respectively labeling the functional relationship between pigment concentration and frozen front change.

[0033] S5. By combining the equivalent factors of pigment concentration and freezing front, a parameter fitting combination under trend fitting is obtained. Based on the parameter fitting combination, the optimal freezing parameters during freezing are derived.

[0034] Using aqueous solutions of bromoamino acid at different concentrations as an example, this paper explains the treatment method of dye wastewater by freezing purification.

[0035] For example, the current freezing point temperature can be determined by following these steps, and then the temperature field change curve can be further calculated based on the obtained freezing point temperature.

[0036] (1) Prepare aqueous solutions of bromoamino acid of different concentrations and place them in plastic bottles of the same volume. Label each bottle to ensure that the liquid volume in each bottle is consistent. Insert the temperature sensor into the solution and tighten the cap to fix the sensor position.

[0037] (2) Place all plastic bottles into the cold bath, close the bath cover, start the cold bath equipment, set the temperature to -10℃, and start data acquisition.

[0038] (3) Monitor the temperature changes in real time, and end the experiment after the temperature of each solution tends to stabilize.

[0039] Based on the calculated freezing point temperature, the subsequent temperature field change curves are recorded, and the temperature values ​​are converted into a data combination mapping cold end temperature, initial concentration, and average freezing rate.

[0040] When interpreting the four layers, they will be separated as follows: the light red ice layer containing solute at the surface is the micro-contamination layer; the colorless and transparent ice layer in the middle, free of bubbles, is the clean layer; below it is a colorless and transparent ice layer with widely distributed bubbles, i.e., the bubble layer; and at the bottom is the residual high-concentration bromoamino acid aqueous solution, i.e., the concentrated layer. These four layers will be the main part of the current analysis to interpret the freezing parameters configured during the current freeze purification process.

[0041] In step S1, the thickness and pigment concentration of each layer are calculated using a multi-time-step approach based on the images extracted from each layer. The micro-contamination layer, clean layer, bubble layer, and concentration layer of the current freeze analysis are divided into grids. Within each grid, the micro-contamination layer and concentration layer in the freeze process are determined using the corresponding image recognition method. Areas with obvious ice crystals are considered as clean layers, and areas with bubbles below the clean layers are classified as bubble layers.

[0042] Layer thickness quantification will employ a high-definition image scanner and image analysis software such as ImageJ to scan the longitudinal cross-sectional image of the frozen container and establish a pixel-to-actual length conversion relationship. The upper and lower boundaries of each layer will be identified manually or automatically. The thickness of a layer is calculated as (lower boundary pixels - upper boundary pixels) × conversion factor (mm). This conversion factor is set based on known physical dimensions; for example, approximately 12 pixels per millimeter at 1200 dpi, indicating the factor used to convert pixels in the image to physical length.

[0043] The initial sampling point can be selected as the 1-hour freezing time point, and the data can be recorded every 30 minutes to obtain the time series corresponding to each layer thickness.

[0044] Pigment concentration is quantified by using a visible spectrophotometer to detect the characteristic wavelength of the dye. A small amount of sample is taken from each layer, and the absorbance is measured. The pigment concentration is then calculated using a standard curve. The clean layer needs to be distinguished between the upper and lower layers, and the concentration of the corresponding pigment is recorded.

[0045] like Figure 10 As shown, the implementation of step S1 includes: S11, taking dye wastewater of arbitrary initial concentration as input, converting the thickness of the layer and the pigment concentration at each time point into a sequence that progresses over time, to obtain the layer change sequence.

[0046] S12, based on the value range of the stratified change sequence, the average freezing rate and cold end temperature are used as constraints of the stratified change sequence to determine the fitting relationship between the average freezing rate and the initial concentration and cold end temperature, and the obtained fitting relationship is synchronized to the index of each stratified label.

[0047] During the experiment, the average freezing rate, cold end temperature, freezing time, and the concentration of the input dye wastewater will be controlled to record changes in other parameters.

[0048] like Figure 2 As shown, a total of 11 temperature sensors were arranged in the test, starting 1 cm below the cold end and arranged vertically at 1 cm intervals until approximately 11 cm from the lower temperature control plate. During the experiment, different initial concentrations of dye wastewater resulted in variations in supercooling temperature and duration. Generally, higher initial concentrations corresponded to lower freezing points, and the temperature drop would occur in different ways. To compare the progress of the freeze-purification process for dye wastewater with different initial concentrations, the freezing rate and the change in the freezing front per unit time were measured.

[0049] Using 0℃ as the dividing line for determining the location of the freezing front, the progress of the freezing front over time can be clearly reflected by observing the changes in the position of the 0℃ isotherm at different time points.

[0050] As shown in Table 1, the cold end temperature is affected by the initial concentration of the current dye wastewater, and the average freezing rate will also show different forms of difference.

[0051] Table 1. Results of Freezing Rate Test

[0052]

[0053] Table 1 illustrates the freezing rates observed under different cold-end temperatures and initial concentrations. These freezing rates represent the speed of the current freezing process and can provide a preliminary assessment of the influence of various factors on the average freezing rate. The results show that cold-end temperature is the primary factor affecting the average freezing rate; the lower the cold-end temperature, the higher the average freezing rate. The decrease in cold transfer efficiency during the mid-freezing phase is due to the increased ice thickness after the microcontamination layer stabilizes (cold energy needs to penetrate a thicker ice layer), and is unrelated to the cold-end temperature itself. At this point, the average freezing rate slows down with increasing ice thickness. In contrast, the effect of the initial solution concentration is relatively small; as the initial concentration increases, the average freezing rate decreases to some extent.

[0054] The appropriate fitting relationship can be set based on the current freezing rate, cold end temperature, and initial concentration, as shown below.

[0055] ;in, Represents the average freezing rate. The value represents the cold end temperature. The currently fitted average freezing rate is used to represent the fitted value in the current scenario. In the formula, 0.022 and 0.102 are exemplary results obtained by fitting data from 30 batches of experiments. The calculation results will also change when different cold end temperatures are input. However, the currently fitted calculation can reflect the relationship between the average freezing rate and related influencing factors more accurately, providing a reliable reference for the optimization of subsequent freezing and purification test parameters.

[0056] The fitting method using the initial concentration can be expressed as follows: ;in, This represents the initial concentration. The fitting then interprets the result of the current calculation based on the input initial concentration, illustrating the change in freezing efficiency under different conditions.

[0057] It should be noted that during the freezing analysis, the lower layer solution gradually concentrates, and the concentrate can be discharged through the drainage pipe below to ensure that the freezing process continues.

[0058] As freezing progresses, the ice formation rate and freezing time will be recorded upon completion. This will allow for further verification of the currently used freezing parameters. The freezing process will be displayed as a trend, determining the cold end temperature, freezing rate, and freezing time for different dye wastewater conditions. The combined value of these multiple data points will be considered the primary target for current adjustments. Ultimately, by optimizing the freezing parameters, the goal is to reduce energy consumption and improve purification efficiency.

[0059] In one embodiment of the present invention, the initial concentration is detected by a UV-Vis spectrophotometer and the corresponding components and concentrations in the water are recorded, which is used as the basis for subsequent verification of the cleanroom purification rate.

[0060] The initial COD / TOC ratio is used as a removal rate benchmark by potassium dichromate method or TOC analyzer to facilitate verification of the purification effect of subsequent cleanroom layers.

[0061] The cold end temperature range is typically set from -5 to -20°C, and this range will be set according to the characteristics of the strata being monitored; the average freezing rate is generally expressed as... The unit is 0.2 to 0.9. The average freezing rate is mainly affected by the cold end temperature and the current amount of wastewater. The freezing rate refers to the target volume of frozen wastewater, such as 50%-80%, to prevent the concentrated layer from being unable to separate when the upper and lower parts of the clean layer are completely frozen during subsequent testing.

[0062] The implementation process of step S2 includes: for the micro-contamination layer during the advance of the freezing front, using a single variable rule, retrieving the cold end temperature and initial concentration under freezing analysis.

[0063] The cold end temperature, initial concentration, and average freezing rate are used as a set of parameters associated with the microcontamination layer. When there is a correlation between the current parameter set and the thickness of the microcontamination layer, the ratio of the change in the thickness of the microcontamination layer to the distance the freezing front advances is regarded as the displacement response coefficient of the microcontamination layer.

[0064] The displacement response coefficient set here is suitable for explaining the change in the thickness of the current microcontamination layer as the freezing front advances over time, and for explaining the changes in the current microcontamination layer during the freezing process.

[0065] During the freeze-cleaning process, the formation of the surface micro-contamination layer is closely related to the freeze-crystallization mechanism. When the upper temperature control plate begins to transfer cold energy, the temperature of the surface solution in contact with the cold end drops rapidly. Due to supercooling, even if the solution temperature drops to the freezing point, spontaneous crystallization does not immediately occur. However, the rapid temperature drop leads to a decrease in the solubility of bromoacetic acid in water, causing some bromoacetic acid to precipitate from the solution.

[0066] As the temperature decreases further, water molecules solidify and crystallize on the surface of the precipitated bromoacetic acid particles, resulting in heterogeneous nucleation. During this process, due to the supercooled state, the ice crystals easily form dendritic crystal structures. These structures have a strong encapsulation ability and are more likely to trap or adsorb solute molecules, thus forming a micro-contamination layer in the early stages of freezing. The thickness of this micro-contamination layer directly affects the thickness of the subsequent cleanroom layer that can grow.

[0067] like Figure 3As shown, the lower the cold-end temperature and the higher the initial solution concentration, the more pronounced the thickness of the microcontamination layer. When the initial solution concentration remains constant, the average freezing rate of the solution increases with decreasing cold-end temperature, while the thickness of the microcontamination layer gradually increases. However, under a constant cold-end temperature, the thickness of the microcontamination layer increases with increasing initial solution concentration, but the average freezing rate of the solution decreases.

[0068] Based on relevant theories, the reasons for the above correlation are analyzed as follows: When the cold end temperature is low, the freezing rate of the solution accelerates, and the solute concentration at the freezing front increases rapidly. This increase rate is much higher than the rate at which the solute diffuses into the liquid phase, causing bromoacetic acid molecules to remain trapped in the solid phase due to insufficient time to diffuse away, thus forming a relatively thick microcontamination layer in the early stages of freezing. Furthermore, solutions with higher initial concentrations are more likely to reach saturation and precipitate bromoacetic acid during cooling, promoting the formation of low-purity ice crystals on the heterogeneous nucleus, thereby exacerbating the formation of the microcontamination layer.

[0069] Once the microcontamination layer reaches a certain thickness, the ice nucleus formation rate becomes less than the ice crystal growth rate, and a relatively pure ice layer gradually begins to form in the system. Therefore, under conditions of lower cold-end temperatures and higher initial solution concentrations, the system often needs to form a thicker microcontamination layer to allow the cold energy transfer process to reach a new equilibrium state, and then enter the growth stage of pure ice crystals.

[0070] At this point, the micro-contamination layer needs to reach a certain thickness, and its changing trend should allow for the normal generation of a clean layer, thus completing the freeze-purification process.

[0071] The implementation of step S2 also includes: for the clean layer during the advance of the freezing front, obtaining the freezing time and ice formation rate corresponding to the current average freezing rate; and recording the change curves of ice formation rate and freezing time using the initial concentration and cold end temperature as constraints.

[0072] When the curves showing a trend correlation between the ice formation rate and freezing time, the ratio of the change in the current cleanroom thickness to the distance the freezing front advances is considered the displacement response coefficient of the cleanroom. The displacement response coefficient of the cleanroom reflects its sensitivity to the advance of the freezing front.

[0073] As for the subsequent bubble layer, since this part is close to the lower layer of the clean layer and is the part where bubbles migrate during the growth of the clean layer, the thickness of this layer represents the migration behavior of microscopic gases due to the evolution of ice crystal structure. Its thickness does not affect the removal rates of COD, TOC and color, and is not the main part to be optimized at present.

[0074] The aforementioned trend correlations represent the patterns of change in the ice-forming rate and freezing time curves. By comprehensively interpreting these trends in relation to different initial solution concentrations and cold end temperatures, and describing the corresponding trends, the displacement response coefficients under the corresponding trends are obtained.

[0075] like Figure 4 As shown, when the initial solution concentration remains constant, the average freezing rate of the solution gradually increases as the cold end temperature decreases, while the clean layer thickness decreases. Conversely, under the condition of constant cold end temperature, the clean layer thickness decreases as the initial solution concentration increases, while the average freezing rate of the solution gradually slows down. Therefore, the trend of clean layer thickness variation is opposite to that of the micro-contamination layer.

[0076] At this point, it is necessary to pay attention to the changing trend of the cleanroom layer so that the change in the thickness of the layer presents a relatively optimal state, providing a data basis for selecting relevant parameters.

[0077] like Figures 5-6 As shown, when identifying the thickness of the cleanroom layer, it is also necessary to further consider the freezing time and ice formation rate, and use data such as freezing time and ice formation rate as constraints on the thickness of the cleanroom layer to further identify its changes.

[0078] As shown in Table 2, the results of ice formation rate related to the cleanroom thickness under different conditions are given.

[0079] Table 2. Final ice formation rate under different conditions

[0080]

[0081] From Table 2 and Figures 5-6 It is evident that the final ice-forming rate is significantly affected by the cold-end temperature. As the cold-end temperature decreases, the final ice-forming rate increases significantly, while the freezing time shows the opposite trend. This indicates that appropriately lowering the cold-end temperature not only helps shorten the freezing time but also increases the final ice-forming rate, making it an effective way to improve freezing purification efficiency. On the other hand, the effect of the initial solution concentration on the final ice-forming rate has a certain stage-specific characteristic. When the initial concentration is less than 500 mg / L, the effect of concentration change on the final ice-forming rate is relatively small, and the ice-forming rate increases slightly with increasing concentration; however, when the concentration further increases to 1000 mg / L, the final ice-forming rate decreases significantly. Simultaneously, the freezing time consistently increases with increasing initial concentration, indicating that the higher the solute content in the wastewater, the longer the freezing purification time, and the lower the overall efficiency.

[0082] In step S2, the freezing rate is used to describe the changes in each layer, such as the micro-contamination layer, the clean layer, the bubble layer, and the concentration layer. At this time, the progress of the freezing process is described according to the average freezing rate and the corresponding constraints. The initial concentration and the cold end temperature can be regarded as independent variables for decision analysis. The freezing rate, freezing time, and thickness will be used as dependent variables. The dependent and independent variables are randomly combined, and the freezing analysis process is interpreted through multiple combinations.

[0083] In one embodiment of the present invention, during step S3, the response of the clean layer and the micro-contamination layer to the advance of the freezing front is used to accumulate over time, and the change in the time trend is synchronized to the growth of the layer thickness under the time accumulation.

[0084] like Figure 7 As shown, when the freezing front advances, if the initial concentration, cold end temperature and other parameters remain constant, the thickness change will also show a monotonic trend. For example, the concentrated layer shows a monotonic decrease, the clean layer shows a monotonic increase, and the micro-contaminated layer shows a monotonic increase. The difference is that the slope of the three layers changes in different ranges.

[0085] Step S3 will combine the changes in thickness and use the displacement response coefficient related to the thickness change of the frozen front to accumulate over time, and quantify the accumulated value to determine the stability of the stratification as the front advances.

[0086] like Figure 11 As shown, step S3 is implemented as follows: S31, by selecting multiple different segmentation points for each layer of the time series that changes over time, the time series is divided into multiple different subsequences. When setting the segmentation points, Figure 7 The intersection of the intermediate-micro-contamination layer, the clean layer, and the concentration layer is used as the dividing point. The time point with the largest change in slope value in the remaining part of the curve is also selected as the dividing point, thus dividing the time series into multiple time periods.

[0087] S32. According to the time point of each subsequence, compare the timestamp of each segmentation point. When the timestamps of the segmentation points are consistent, the current segmentation point is regarded as the target time point. Calculate the sum of the product of the displacement response coefficient and the layer thickness within the corresponding time period as the output equivalent effect factor.

[0088] When the timestamps match, it indicates that any two layers among the clean layer, concentrated layer, and micro-contaminated layer exhibit the maximum slope change at the same time point. This signifies a synchronous abrupt change in the rate of change of the two layers at that moment, reflecting a key turning point in the freezing process. The time period corresponding to this turning point represents the period from linear change to relative stability. During the period of gradual stabilization, the thickness of each layer changes slowly, gradually progressing to complete freezing. The content of the linear change time period will illustrate the process of gradual and continuous accumulation of thickness. Both of these time periods represent the cumulative process of the current major changes. The time period before and after the dividing point will be the main part of the current time accumulation analysis.

[0089] S33. When the timestamps of the split points are inconsistent, treat each split point as an independent group, perform a one-way ANOVA on each split point, and obtain the time weight of each split point.

[0090] S34. Calculate the clustering results of each segmentation point based on its time weight. Adjust the position of each segmentation point using the average weight after clustering, and output the equivalent effect factor calculated for the time period corresponding to each segmentation point.

[0091] Data from multiple batches is introduced, and timestamps of all split points are recorded. Each split point is treated as an independent group, and each group includes experimental data from at least 100 adjacent batches. In one-way ANOVA, the effect sum of squares and residual sum of squares from the ANOVA are introduced, so that the sum of the effect sum of squares and residual sum of squares equals the total sum of squares. The time weight of each split point is set by the ratio of the effect sum of squares to the total sum of squares for each group.

[0092] Then, a dataset was constructed using time weight as a feature, and K-means clustering was used. After selecting the corresponding K value through the elbow rule, clusters were obtained after the split points were clustered. For each cluster, a weighted average was performed using timestamps and time weights. The timestamp obtained was used as the adjusted position of the split point, and the data in the corresponding time period of the split point was extracted to calculate the equivalent effect factor.

[0093] When outputting the above equivalent factor, it is also necessary to divide it by the length of the corresponding time period to explain the relative accumulation within a unit time period. If there are units in the current calculation process, the data needs to be standardized.

[0094] When verifying the displacement response results, the values ​​of the upper and lower layers of the cleanroom during the evaluation of freeze purification will be introduced, and these values ​​will be used to explain the effect of the current cumulative changes.

[0095] When evaluating the effectiveness of freeze purification, the analysis is mainly based on the changes in three indicators: chemical oxygen demand (COD), total organic carbon (TOC), and color in the melted ice water of the original solution and the clean layer. The calculation methods for these three parameters are disclosed in the existing technology. It is only necessary to calculate the difference between the corresponding parameters of the original solution and the clean layer, divide it by the value in the original solution, and then explain the effectiveness of the currently selected freezing parameters according to the changing trend of the relevant parameters in the final solution of the clean layer.

[0096] When judging the displacement response result in step S3, the implementation method also includes: establishing a mapping between the equivalent effect factor and the index values ​​of the upper and lower clean layers, where the index values ​​of the upper and lower clean layers represent the removal rates of COD, TOC, and color.

[0097] The equivalent effect factor is checked in turn to see if the change over time meets the triggering condition, and the relevant data of the equivalent effect factor that meets the triggering condition is used as the output displacement response result.

[0098] The triggering conditions need to be explained. The coefficient of variation of the current equivalent factor should be less than 10%, and the removal rates of COD, TOC and chroma should all be greater than 85%. When these triggering conditions are met, it means that the current freezing process is stable and the freezing effect is excellent. This is the parameter combination required for the current situation. In this case, the data related to the equivalent factor should be used as the output displacement response result.

[0099] In one embodiment of the present invention, the index values ​​of the upper and lower clean layers can be extracted by collecting ice blocks at corresponding locations and then using ice melt water to measure the removal rates of COD, TOC and color at the corresponding locations.

[0100] like Figures 8-9 As shown, changes in cold-end temperature have a relatively small impact on the quality of the ice melt water in the cleanroom. During the freezing purification process, the main role of cold-end temperature is in controlling the freezing rate: the lower the cold-end temperature, the higher the freezing rate, and the shorter the cleanroom formation time.

[0101] Therefore, in the current multi-batch test scenarios, introducing different dividing points to quantify the cumulative thickness of each layer allows the freezing effect of the upper and lower clean layers to be mapped with the equivalent effect factor, in order to verify whether the change of the equivalent effect factor over time shows a stable trend, thus making the freezing purification effect better.

[0102] In step S4, the changing trends of pigment concentration and freezing front are quantified. The correlation coefficient between the two is calculated in the form of instantaneous change rate. The changing trend of pigment concentration will be based on the pigment change trend at a certain location during freezing to explain the rate at which pigment is transported to the concentration layer by ice crystals during freezing. The selected location can be marked below the cold end. The pigment concentration extracted in multiple sampling processes at 1 cm is combined with the corresponding time points to form the changing trend of pigment. The freezing front advance rate is used to determine its positive correlation with the change of pigment concentration, clarifying the two types of differences: pigment concentration-dominated and freezing front-dominated.

[0103] The two types of differences, namely pigment concentration-dominated and freezing front-dominated, are essentially comparisons of the initial solution concentration and cold end temperature under different conditions. Correlation calculations are performed on the pigment concentration changes and freezing front changes affected by these two conditions to select parameter combinations with significant trend differences. Using the parts with significant trend differences, the two trend-dominated scenarios are separated, and the functional relationships under the corresponding scenarios are clarified, which provides data support for parameter optimization in step S5.

[0104] When calculating the correlation, the Pearson correlation coefficient is used to determine the degree of linear correlation between pigment concentration and freezing front. Data with absolute values ​​of Pearson correlation coefficients less than 0.6 can be selected as the part of the significant difference analysis to obtain data with a sufficient number of dimensions.

[0105] Furthermore, the selected data must also meet the residual test. If the percentage of samples with residual absolute values ​​greater than twice the standard deviation is ≥30%, then it is confirmed that there is a significant trend difference between pigment concentration and freezing front. After excluding the current selected data as random error, the changing trends dominated by pigment concentration and freezing front can be explained.

[0106] like Figure 12 As shown, the implementation of step S4 includes: S41, recording the changing trends of pigment concentration and freezing front respectively, and when the changing trends of the two satisfy a significant trend difference, determining the equivalent effect factor dominated by pigment concentration and the equivalent effect factor dominated by freezing front respectively.

[0107] When decomposing the equivalent effect factor, it is necessary to determine the contribution ratio of the two types of data. This is done by normalizing the data and calculating the ratio of the two data to their sum. Then, multiplying this value by the equivalent effect factor will separate the two dominant forms of the equivalent effect factor.

[0108] S42, based on the pigment concentration-dominated equivalent factor and the freezing front-dominated equivalent factor according to the corresponding timestamps, the output function relationship is obtained by fitting with the least squares method.

[0109] When setting the functional relationship, the pigment concentration decays rapidly in the early stage and then stabilizes in the later stage. In this case, a logarithmic decay form is used to set the functional relationship corresponding to the pigment concentration. The freezing front also advances rapidly in the early stage and then slows down in the later stage. A logarithmic decrease form will be used to explain the functional relationship. The trend of the freezing front is relatively close to that of the pigment concentration. In scenarios where there is a significant inconsistency, it often represents the time period when a micro-contamination layer and a part of the clean layer are formed. This time period represents the period of significant change. Compared with obtaining the steadily increasing thickness data in step S3, step S4 will check the data of the initial reaction and check the dominance of pigment concentration and freezing front in these data to identify the main influence of the initial concentration and freezing front.

[0110] When setting the functional relationship, based on the relevant data input in step S3, the timestamp of the displacement response result is retrieved to obtain the initial value for the current analysis. For example, when setting the functional relationship of pigment concentration, the initial factor value is obtained. This value represents the earliest timestamp value when decomposing the equivalent factor corresponding to pigment concentration. Then, the equivalent factor dominated by pigment concentration can be equal to the initial factor value × exp(-k1×t). Here, exp represents an exponential function with base e, k1 represents the attenuation coefficient obtained by least squares fitting, and t represents the current timestamp, thus obtaining multiple equivalent factor values ​​representing pigment concentration dominance.

[0111] When setting the functional relationship of the freezing front, the maximum equivalent factor value within the corresponding timestamp will be recorded. The maximum equivalent factor will be multiplied by ln(1+k2×t), where k2 represents the propulsion coefficient calculated by the least squares method. Using these two functions to fit two functional relationships is used to quantify the influence of initial concentration and cold end temperature in the initial stage of freezing.

[0112] When labeling the functional relationship between pigment concentration and freezing front changes, the implementation method also includes: retrieving the timestamp of the displacement response result, and using the position of the timestamp, gradually deriving the freezing period corresponding to the equivalent effect factor dominated by pigment concentration and the equivalent effect factor dominated by freezing front; the freezing period represents the growth process during freezing, and can be divided into three periods according to the freezing time, such as the initial freezing period within 8 hours of freezing, during which the freezing front advances rapidly and the resistance to ice crystal growth is small; the freezing time within 8-18 hours is regarded as the middle freezing period, at which time the cold transfer efficiency decreases after the micro-contamination layer stabilizes, the concentration decays faster, and the ice crystals turn into granules, the encapsulation is reduced, and the solute migration is smooth; the part after more than 18 hours is regarded as the late freezing period, at which time the freezing is almost complete and the progress is slow.

[0113] The split equivalent factors are labeled step by step according to the freezing period, and the reasons for the difference between pigment concentration and freezing front trend are recorded. The reasons for the difference will record the main trends of pigment concentration change and freezing front change in the corresponding stage, and explain the reasons that may lead to the current identification difference. These reasons will be mapped to the corresponding equivalent factors with index numbers.

[0114] In one embodiment of the present invention, in step S5, the freezing result obtained in step S3 is used as the objective function, and a parameter fitting combination composed of equivalent effect factors is introduced. The frequent itemset of the parameter fitting combination is used to optimize the objective, and the optimized parameters are used as the output optimal freezing parameters. The obtained optimal fitting parameters will show a more stable process, ensuring that the optimal parameters not only meet the standards in the laboratory scenario, but also adapt to practical applications.

[0115] like Figure 13 As shown, the implementation of step S5 includes: S51, extracting at least one frequent itemset from the parameter fitting combination, indexing each frequent itemset according to its support and confidence, and storing it in the database; the frequent itemsets are connected pairwise by selecting the equivalent factors of pigment concentration and frozen front; when extracting frequent itemsets, if the equivalent factors of pigment concentration and frozen front are frequent, then the related pigment concentration and frozen front values ​​will also be relatively frequent, which will further affect the initial solution concentration and cold end temperature; if the removal rates of COD, TOC, and color can also reach the optimal at this time, then it means that the currently selected frequent itemset can represent a set of optimal freezing parameters, so that the optimal treatment effect can be achieved for dye-related wastewater.

[0116] S52, establish a mapping relationship between each frequent itemset and the index values ​​of the upper and lower cleanroom layers in sequence. When the frequent itemset satisfies the minimum support and minimum confidence, use the index values ​​of the upper and lower cleanroom layers to perform multi-objective optimization.

[0117] The essence of multi-objective optimization is to find the parts where the removal rates of COD, TOC, and color are the largest in the upper and lower layers of the cleanroom. That is, multi-objective optimization is performed through scoring. The removal rates of COD, TOC, and color can be assigned weights of 0.3, 0.3, and 0.4 respectively. After calculating the weighted values ​​of the upper and lower layers of the cleanroom, the weights of the upper and lower layers of the cleanroom are then assigned weights of 0.4 and 0.6 respectively, thus obtaining the score values ​​of COD, TOC, and color in the upper and lower layers of the cleanroom.

[0118] In the scenario of freezing and purifying dye wastewater, removing the color is the top priority. Therefore, the weight of the color removal rate will be slightly higher than that of COD and TOC. At the same time, the upper layer of the clean layer is directly connected to the micro-contamination layer, which is the foundation for the stable growth of the subsequent clean layer. The clean layer connected to this layer has higher values ​​in terms of COD, TOC and color removal rates, so its weight is slightly lower than that of the lower layer of the clean layer to evaluate the stability of pigment, COD and TOC removal during the subsequent ice crystal migration process.

[0119] The support mentioned above is the statistical value of the frequency of a specific frequent itemset in the total frequent itemset. The confidence level is the ratio of the number of samples that simultaneously contain both premise and conclusion to the number of samples that contain premise. In other words, it is the ratio of the equivalent factor of the current pigment concentration and frozen front combination to the number of all frequent itemsets that contain that pigment concentration.

[0120] In the current experiment, to avoid filtering out effective patterns due to insufficient data, the minimum support value can be set to 10%. To balance the associated scenarios and data, 60% can be used as the minimum confidence level to obtain a sufficient number of reliable combinations.

[0121] S53, let the scores of COD, TOC and chromaticity removal rates be used as the subject of multi-objective optimization, and use the scores after multi-objective optimization to filter frequent itemsets; based on the cold end temperature and initial concentration corresponding to the filtered frequent itemsets, the optimal freezing parameters are output.

[0122] In the implementation of the current scheme, the cold end temperature and initial concentration are used as constraints of the cleanroom layer to record relevant parameters under the conditions of cold end temperature and initial concentration. These relevant parameters will include parameters such as freezing time, average freezing rate and ice formation rate. These parameters will be used as an auxiliary part of the optimal freezing parameters for comprehensive output to explain the parameters involved in the optimal freezing scenario.

[0123] Preferably, the filtering method for frequent itemsets includes: based on the score value corresponding to the frequent itemsets, taking the frequent itemset with the largest score value as the root node, arranging the frequent itemsets in descending order of score value, and building child nodes level by level under the root node to form a first-level child node chain. If there are two or more frequent itemsets with the same score value, then the support is selected for descending sorting; if the support is still the same, then the confidence is selected for descending sorting.

[0124] It can be seen that the root node represents the globally optimal freezing parameters in the current scene, and the child nodes are the locally optimal parameters. The first-level child node chain sets scores based on the removal rates of COD, TOC, and chroma, and sorts them accordingly to illustrate a set of locally optimal parameters corresponding to the freezing effect under the corresponding conditions.

[0125] Iterate through the cold end temperature and initial concentration corresponding to each child node. If the current child node and its adjacent node have the same value of cold end temperature or initial concentration, increment the count of the child node by 1 and record the total frequency of occurrence of cold end temperature and initial concentration in each child node. When selecting the current child node and its adjacent node, select the next node of the current node for judgment and record the total frequency of occurrence of cold end temperature and initial concentration in the corresponding node.

[0126] The child nodes are sorted according to the total frequency of occurrence of cold junction temperature and initial concentration in adjacent child nodes to construct a secondary child node chain. The secondary child node chain will further verify the occurrence frequency of cold junction temperature and initial concentration under local optima. This frequency can verify the stability of the corresponding data, making the overall experimental process a knowable and quantifiable scenario.

[0127] When the second-level child node chain and the first-level child node chain are completely identical in sorting position, the current root node and its corresponding child node are used as the optimal frozen parameters for output.

[0128] When the second-level child node chain and the first-level child node chain are not completely consistent in sorting position, the intersection of the second-level child node chain and the first-level child node chain and the root node are selected as the optimal frozen parameters for the output.

[0129] After sorting and verifying the second-level child node chain and the first-level child node chain, the resulting parameters will represent a set of useful parameters with high scores, high stability, and high support. These parameters can be used as the optimal freezing parameters to describe the freezing process in different scenarios.

[0130] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention, which are still covered within the protection scope of the present invention.

Claims

1. A test method for separating and purifying dye wastewater based on freeze-drying purification technology, characterized in that, include: S1 uses image analysis to construct the micro-contamination layer, clean layer, bubble layer and concentration layer during the advance of the frozen front, and records the thickness and pigment concentration of each layer; S2, Based on the freezing rate under freezing analysis, analyze the current freezing front with the average freezing rate to determine the displacement response coefficient of each layer; S3, using the interval time and layer thickness when configuring the displacement response coefficient, determine the equivalent effect factor in time accumulation; Based on the equivalent effect factor and the value range of the upper and lower clean layers, the displacement response of each layer as the freezing front advances is determined. S4. The displacement response result after judgment is used as input to record the difference in the trend between pigment concentration and the advance of the frozen front. The equivalent factor is divided into two categories related to pigment concentration and frozen front, and the functional relationship between pigment concentration and frozen front change is marked respectively. S5. By combining the equivalent factors of pigment concentration and freezing front, a parameter fitting combination under trend fitting is obtained. Based on the parameter fitting combination, the optimal freezing parameters during freezing are derived.

2. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that, The implementation methods for step S1 include: S11, using dye wastewater of arbitrary initial concentration as input, transforms the thickness of the stratification and pigment concentration at each time point into a sequence that progresses over time, thus obtaining the stratification change sequence; S12, based on the value range of the stratified change sequence, the average freezing rate and cold end temperature are used as constraints of the stratified change sequence to determine the fitting relationship between the average freezing rate and the initial concentration and cold end temperature, and the obtained fitting relationship is synchronized to the index of each stratified label.

3. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that, The implementation process of step S2 includes: For the micro-contamination layer during the advance of the freezing front, the cold end temperature and initial concentration under freezing analysis were retrieved using a single variable rule; The cold end temperature, initial concentration, and average freezing rate are used as a set of parameters associated with the microcontamination layer. When there is a correlation between the current parameter set and the thickness of the microcontamination layer, the ratio of the change in the thickness of the microcontamination layer to the distance the freezing front advances is regarded as the displacement response coefficient of the microcontamination layer.

4. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that, The implementation of step S2 also includes: For the clean layer during the advance of the freezing front, the freezing time and ice formation rate corresponding to the current average freezing rate are obtained; the change curves of ice formation rate and freezing time are recorded with the initial concentration and cold end temperature as constraints. When the curves showing a trend correlation between the changes in ice formation rate and freezing time, the ratio of the change in the current cleanroom thickness to the distance the freezing front advances is regarded as the displacement response coefficient of the cleanroom.

5. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that, Step S3 can be implemented in the following ways: S31, by selecting multiple different dividing points for each layer of the time series that changes over time, the time series is divided into multiple different subsequences; S32, according to the time point of each subsequence, compare the timestamp of each segmentation point. When the timestamps of the segmentation points are consistent, the current segmentation point is regarded as the target time point. Calculate the sum of the product of the displacement response coefficient and the layer thickness within the corresponding time period as the output equivalent effect factor. S33, when the timestamps of the split points are inconsistent, treat each split point as an independent group, perform a one-way ANOVA on each split point, and obtain the time weight of each split point; S34. Calculate the clustering results of each segmentation point based on its time weight. Adjust the position of each segmentation point using the average weight after clustering, and output the equivalent effect factor calculated for the time period corresponding to each segmentation point.

6. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 5, characterized in that, When determining the displacement response result in step S3, the implementation method also includes: A mapping is established between the equivalent effect factor and the index values ​​of the upper and lower cleanroom layers. The equivalent effect factor is checked in turn to see if the change over time meets the triggering condition. The relevant data of the equivalent effect factor that meets the triggering condition is used as the output displacement response result.

7. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that, Step S4 can be implemented in the following ways: S41, record the changing trends of pigment concentration and freezing front respectively. When the changing trends of the two satisfy a significant trend difference, determine the equivalent effect factor dominated by pigment concentration and the equivalent effect factor dominated by freezing front respectively. S42, based on the pigment concentration-dominated equivalent factor and the freezing front-dominated equivalent factor according to the corresponding timestamps, the output function relationship is obtained by fitting with the least squares method.

8. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 7, characterized in that, When annotating the functional relationship between pigment concentration and freezing front changes, the implementation methods also include: Retrieve the timestamps of the displacement response results, and use the positions of the timestamps to gradually deduce the freezing periods corresponding to the equivalent effect factors dominated by pigment concentration and the equivalent effect factors dominated by freezing fronts. The split equivalent factors were labeled step by step according to the freezing period, and the reasons for the difference between pigment concentration and freezing front trend were recorded.

9. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 1, characterized in that... Step S5 can be implemented in the following ways: S51, extract at least one frequent itemset from the parameter fitting combination, and index each frequent itemset according to its support and confidence and store it in the database; S52, establish a mapping relationship between each frequent itemset and the index values ​​of the upper and lower cleanroom layers in sequence. When the frequent itemset satisfies the minimum support and minimum confidence, use the index values ​​of the upper and lower cleanroom layers for multi-objective optimization. S53, let the scores of COD, TOC and chromaticity removal rates be used as the subject of multi-objective optimization, and use the scores after multi-objective optimization to filter frequent itemsets; based on the cold end temperature and initial concentration corresponding to the filtered frequent itemsets, the optimal freezing parameters are output.

10. The experimental method for separating and purifying dye wastewater based on freeze-drying purification technology according to claim 9, characterized in that, Methods for filtering frequent itemsets include: Based on the score value corresponding to the frequent itemsets, the frequent itemset with the largest score value is regarded as the root node. The frequent itemsets are arranged in descending order of score value, and child nodes are built under the root node to form a first-level child node chain. Iterate through the cold end temperature and initial concentration corresponding to each child node. If the current child node has the same value of cold end temperature or initial concentration as the adjacent node, increment the count of the child node by 1 and record the total frequency of occurrence of cold end temperature and initial concentration in each child node. The child nodes are sorted according to the total frequency of occurrence of cold end temperature and initial concentration in adjacent byte points, and a two-level child node chain is constructed. When the second-level child node chain and the first-level child node chain are completely identical in sorting position, the current root node and its corresponding child node are used as the optimal frozen parameters for output. When the second-level child node chain and the first-level child node chain are not completely consistent in sorting position, the intersection of the second-level child node chain and the first-level child node chain and the root node are selected as the optimal frozen parameters for the output.

Images