Method for constructing a prediction model of pfas adsorption behavior based on interface dynamics data fitting

Abstract

The application discloses a PFAS adsorption behavior prediction model construction method based on interface dynamics data fitting, relates to the PFAS adsorption prediction technical field, and comprises the following steps: collecting interface adsorption capacity continuous change data and bulk phase concentration synchronous change data in the concentration rapid decline stage in the PFAS adsorption process, spreading the interface adsorption capacity continuous change data and the bulk phase concentration synchronous change data under a unified time scale, marking the time node at which the interface adsorption capacity changes from a downward trend to an upward trend, and forming a rebound candidate interval; and synchronously comparing the interface adsorption capacity change rate and the bulk phase concentration change amplitude in the rebound candidate interval. The application identifies a real dynamic section, constructs an interface occupation strengthening index, retains interface adsorption capacity rebound information, gives an independent weight in the parameter calculation stage, improves adsorption capacity and saturation process judgment accuracy, corrects the setting of the dosage, reduces the risk of sudden increase of effluent concentration, and enhances operation stability.

Description

Technical Field

[0001] This invention relates to the field of PFAS adsorption prediction technology, and specifically to a method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting. Background Technology

[0002] The construction of a predictive model for PFAS adsorption behavior based on interfacial kinetic data fitting refers to the mass transfer and adsorption process of perfluorinated and polyfluoroalkyl substances at solid-liquid or gas-liquid interfaces. First, interfacial kinetic parameters such as interfacial tension variation curves, adsorption amount versus time curves, concentration gradient distribution data, and surface potential variation data are collected experimentally to clarify the diffusion rate, interfacial migration rate, and adsorption site occupancy evolution of PFAS molecules on specific adsorbent material surfaces. Then, the above interfacial kinetic data are mathematically expressed, dividing the adsorption process into external diffusion, internal diffusion, and surface binding stages. This is achieved by establishing a time-adsorption amount functional relationship and... An expression for the change in interfacial free energy is derived. Nonlinear regression fitting is performed on measured data to extract the apparent adsorption rate constant, diffusion coefficient, and interfacial binding energy parameters. After obtaining a stable parameter expression system, the fitting results under different initial concentrations, temperatures, ionic strengths, and pH conditions are normalized to construct an adsorption behavior prediction framework that reflects changes in multiple operating conditions. This enables the preliminary prediction of PFAS adsorption capacity, adsorption rate, and adsorption equilibrium time under different environmental conditions and different adsorbent material systems. Essentially, this process is based on experimental interfacial kinetic data, establishing an adsorption behavior prediction model with extrapolation capabilities through parameterized fitting and functionalized expression.

[0003] The existing technology has the following shortcomings: During PFAS adsorption, when the system enters a phase of rapid concentration decrease, some short-chain PFAS molecules may experience a short-term reverse enrichment near the interface, manifested as a renewed upward fluctuation in the interfacial adsorption curve within a downward trend. This phenomenon is a transient response characteristic of interfacial kinetics, not a random measurement error, but it is easily misjudged as abnormal noise and discarded during data preprocessing. If this secondary upward fluctuation is not preserved and identified, the actual interfacial occupancy strength and adsorption potential will be underestimated during the fitting process, leading to an underestimation of adsorption capacity and saturation process by the prediction model. In engineering applications, this may result in an underestimation of the adsorbent dosage, potentially leading to a sudden increase in effluent concentration in the later stages of operation, affecting treatment stability.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide a method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting, so as to solve the problems in the background art mentioned above.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting, comprising the following steps: Step 1: During the PFAS adsorption process, collect continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration during the rapid decrease phase. Expand the continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration on a unified time scale, and mark the time node when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form a rebound candidate interval. Step 2: Simultaneously compare the rate of change of interfacial adsorption amount and the magnitude of change of bulk concentration within the rebound candidate range. Under the condition of continuous decrease in bulk concentration, screen out the dynamic segment where the interfacial adsorption amount still increases, determine the dynamic segment as the real dynamic segment and retain its original change trajectory. Step 3: Focusing on the real dynamic segment, analyze the gradient of interface occupancy change at adjacent time points, calculate the proportion of the interface adsorption recovery in the overall downward trend, and construct an interface occupancy enhancement index. Step 4: Embed the interface placeholder enhancement index into the interface dynamics expression process, and assign independent weights to the real dynamic segments during the parameter acquisition stage to distinguish them from non-real dynamic segments participating in the fitting calculation. Step 5: Based on the interfacial kinetics expression results after assigning independent weights, recalculate the upper limit of adsorption capacity and the turning point of saturation process, and revise the adsorbent dosage determination rule according to the recalculation results.

[0007] Preferably, marking the time point when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form a rebound candidate interval includes the following steps: During the adsorption process, the continuous value of the interfacial adsorption amount changes over time is continuously recorded, and the continuous value of the bulk concentration changes over time is recorded at the same time point. When the bulk concentration shows a continuous decrease, the starting time of the rapid concentration drop phase is determined. From this starting time, continuous data is collected at fixed time intervals to form an original time series record table containing time number, interfacial adsorption amount value and bulk concentration value. Based on the original time series record table, the time number is expanded with a unified time scale. The interfacial adsorption value and the bulk concentration value are filled into the continuous time axis sequence. The changes in the interfacial adsorption value of adjacent time nodes are compared under the continuous time axis sequence to determine and mark the time node when the interfacial adsorption value changes from a downward trend to an upward trend. Extend consecutive time nodes forward and backward in the continuous time axis sequence around the time node where the marking is completed, delineate the continuous time segment containing the time node, and keep the interfacial adsorption amount and bulk concentration value recorded as is in the continuous time axis sequence. The defined continuous time intervals are listed in an independent data interval record table, recording the start time number, trend reversal time node number, and end time number, and fully retaining the interfacial adsorption amount value and bulk concentration value corresponding to each time node within the interval.

[0008] Preferably, the interfacial adsorption values ​​within a continuous time interval are kept in their original order of change without deletion, replacement, or smoothing. The direction of change of interfacial adsorption and the direction of change of bulk concentration are recorded synchronously in the independent data segment record table to ensure that the time point when the interfacial adsorption changes from a decreasing trend to an increasing trend corresponds to the decreasing state of the bulk concentration.

[0009] Preferably, determining the dynamic segment as a true dynamic segment and preserving its original change trajectory includes the following steps: For the interfacial adsorption amount values ​​and bulk concentration values ​​arranged continuously by time number within the rebound candidate interval, the difference of interfacial adsorption amount values ​​corresponding to adjacent time numbers is calculated to form the rate of change of interfacial adsorption amount, and the difference of bulk concentration values ​​is calculated according to the same time number order to form the magnitude of change of bulk concentration, and a synchronous comparison record table is generated. Based on the synchronous comparison record table, the segments in which the change in bulk concentration is continuously less than zero are identified in chronological order. Within these segments, time intervals in which the rate of change in interfacial adsorption is greater than zero are selected and marked to form a set of time intervals in which the interfacial adsorption increases. The set of time intervals for the increase in interfacial adsorption amount is integrated according to the continuity of time number to form a continuous and complete segment, and the corresponding interfacial adsorption amount value and bulk concentration value within the complete segment are extracted to form the change trajectory. Complete segments that meet the condition of continuous decrease in bulk concentration and positive rate of change of interfacial adsorption amount are identified. These complete segments that meet the conditions are determined as the true dynamic segments, and all original change trajectories within the true dynamic segments are retained.

[0010] Preferably, the real dynamic segment is identified by continuous time number in the synchronous comparison record table, and the start time number and end time number of the real dynamic segment are listed separately in the overall rebound candidate interval record table. The interfacial adsorption value and the bulk concentration value in the real dynamic segment are kept in their original order without deletion.

[0011] Preferably, constructing interface placeholder enhancement metrics includes the following steps: Based on the start and end time numbers of the real dynamic segment in a unified time scale sequence, the interface adsorption value within the corresponding time range is extracted. The difference between adjacent time nodes is calculated according to the time number order to form the interface occupancy change gradient, and a continuous record table of interface occupancy change gradient is generated. Based on the continuous record table of interface occupancy change gradient, the interface adsorption amount values ​​of adjacent time nodes in the real dynamic segment are read forward and backward to construct the overall time range record table. The interface occupancy change gradient is calculated in the same way, and the increase in interface adsorption amount in the real dynamic segment and the decrease in interface adsorption amount in the overall time range are statistically analyzed. The ratio of the increase to the decrease in interfacial adsorption is calculated to obtain the percentage of the increase in interfacial adsorption in the overall downward trend. This percentage is then combined with the actual dynamic time intervals and the continuous record table of interfacial occupancy change gradient to form an interfacial occupancy enhancement index record table.

[0012] Preferably, the continuous record table of interface occupancy change gradient retains all positive and negative value segments. The increase and decrease of interface adsorption are both derived from the original records in the unified time scale sequence and are associated with the start and end time numbers of the actual dynamic segments.

[0013] Preferably, assigning independent weights to the actual dynamic segments during the parameter calculation stage includes the following steps: Import the interfacial adsorption values ​​from the unified time scale sequence into the data recording table of the interfacial dynamic expression process, set the time number column and the segment identifier column, and mark the time nodes according to the start time number and end time number of the real dynamic segment in the interfacial occupancy enhancement index recording table. At the same time, write the total increase of interfacial adsorption and the proportion of the increase of interfacial adsorption in the overall downward trend into the additional information column. Based on the segment identifier column, a weight value column is set. A baseline weight value is assigned to the time node of the non-real dynamic segment, and the baseline weight value is assigned to the time node of the real dynamic segment multiplied by the percentage of the interface adsorption amount recovery in the overall downward trend to form an independent weight value. The parameters are obtained based on the data record table containing time number, segment identifier and weight value. The participation of the interface adsorption amount is weighted according to the weight value to form a parameter value result record table containing the range of real dynamic segment time number and the proportion of interface occupancy enhancement index value.

[0014] Preferably, the independent weight values ​​corresponding to the real dynamic segments remain consistent throughout the parameter calculation process, and the total increase in interface adsorption amount and the proportion of the increase in interface adsorption amount in the overall downward trend are simultaneously marked in the parameter value result record table to ensure that the parameter results correspond one-to-one with the interface occupancy enhancement index.

[0015] Preferably, the method for revising the adsorbent dosage determination rule based on the recalculated results includes the following steps: Read the predicted values ​​of interfacial adsorption amount arranged by time number in the interfacial kinetic expression result record table after assigning independent weights, construct the adsorption process change sequence, determine the saturation process transition time based on the change of the predicted values ​​of interfacial adsorption amount at consecutive time nodes, determine the upper limit value of adsorption capacity, and write the upper limit value of adsorption capacity and the saturation process transition time into the interfacial kinetic expression result record table. Based on the upper limit of adsorption capacity and the turning point of saturation, combined with the treatment scale and influent concentration conditions, the adsorbent dosage is calculated according to the correspondence between the total amount of pollutants per unit time and the upper limit of adsorption capacity, and the corrected dosage value is generated and recorded in the dosage adjustment record table. The upper limit of adsorption capacity, the turning point of saturation process, and the corrected dosage value are written into the parameter list of the adsorbent dosage determination rule, and the corresponding parameters in the adsorbent dosage determination rule are updated synchronously when the interface kinetic expression results are updated.

[0016] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention identifies the time point when the interfacial adsorption amount changes from a decreasing trend to an increasing trend during the rapid concentration decline phase, and filters out the real dynamic segment under the condition of continuous decrease in bulk concentration. This avoids misjudging the transient response of interfacial dynamics as abnormal noise and eliminating it. Thus, the information on the recovery of interfacial adsorption amount is completely preserved in the process of expressing interfacial dynamics, so that the adsorption behavior prediction model can truly reflect the interfacial occupancy enhancement characteristics, improve the accuracy of adsorption capacity and saturation process judgment, and enhance the consistency between prediction results and actual operating conditions.

[0017] This invention constructs an interface occupancy enhancement index and assigns independent weights to the actual dynamic segments during the parameter calculation stage. This allows the interface dynamics expression results to fully reflect the contribution ratio of the interface adsorption capacity recovery rate to the overall downward trend. Based on this, the upper limit of adsorption capacity and the turning point of the saturation process are recalculated, and the adsorbent dosage determination rules are corrected accordingly. This improves the rationality of adsorbent dosage setting, reduces the risk of sudden increase in effluent concentration in the later stage of operation, and enhances the stability and continuous operation capability of the treatment process. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0019] Figure 1 This is a flowchart illustrating the method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting, as described in this invention. Detailed Implementation

[0020] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0021] This invention provides, for example Figure 1 The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting, as shown, includes the following steps: Step 1: During the PFAS adsorption process, collect continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration during the rapid decrease phase. Expand the continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration on a unified time scale, and mark the time node when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form a rebound candidate interval. The steps for marking the time points when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form rebound candidate intervals are as follows: After the adsorption operation begins, the continuous values ​​of interfacial adsorption amount over time are continuously recorded. At the same time, the continuous values ​​of bulk concentration over time are also recorded at the exact same time points. When the operation enters the rapid concentration decline phase, the start time of this phase is determined as the moment when the bulk concentration is lower than the previous time point for three consecutive time points. From this moment, continuous data collection is carried out at fixed time intervals, and the time interval remains constant. For example, a collection cycle of ten seconds is used. Starting from the start of the rapid concentration decline phase, the interfacial adsorption amount and the corresponding bulk concentration values ​​at no less than thirty time points are continuously recorded. All records are numbered in chronological order to form an original time series record table containing three items: time number, interfacial adsorption amount, and bulk concentration. During the recording process, the time number is kept continuously increasing, without inserting missing time points or deleting any data at any time point, thereby ensuring that the continuous change data of interfacial adsorption amount and the synchronous change data of bulk concentration are completely corresponding in the time dimension.

[0022] After the original time series record table is formed, the time numbers are processed to unify the time scale. That is, all time numbers are rearranged into a continuous time axis sequence according to a fixed time interval. The interfacial adsorption amount value and the bulk concentration value are respectively filled into the corresponding time axis sequence, so that the continuous change data of interfacial adsorption amount and the synchronous change data of bulk concentration form two corresponding numerical sequences on the same time axis. Under this unified time scale, the interfacial adsorption amount values ​​of adjacent time nodes are compared one by one. When the interfacial adsorption amount value of a certain time node is lower than that of the previous time node, and the interfacial adsorption amount value of the subsequent time node is higher than that time node, the time node is determined as the time node when the interfacial adsorption amount changes from a downward trend to an upward trend, and the time node is marked. At the same time, the interfacial adsorption amount value and the bulk concentration value of no less than three consecutive time nodes before and after the time node are recorded to ensure that the identification of the trend change is based on continuous data changes.

[0023] After marking the time point when the interfacial adsorption amount changes from a decreasing trend to an increasing trend, extend forward by no less than five consecutive time points and backward by no less than five consecutive time points from this time point as the center. The entire continuous time segment including this time point is defined as the rebound candidate interval. When defining the rebound candidate interval, the original values ​​of the continuous change data of interfacial adsorption amount and the synchronous change data of bulk concentration are kept without deletion, smoothing, or replacement. All time numbers and corresponding values ​​within the interval are completely preserved, and the interval is continuously marked in a unified time scale sequence, so that the interval forms an independent segment with a clear start time number and end time number in the overall time axis. At the same time, the direction of change of interfacial adsorption amount and the direction of change of bulk concentration are recorded point by point within the segment to ensure that the interfacial adsorption amount has changed from a decreasing trend to an increasing trend within the segment, while the bulk concentration continues to decrease.

[0024] After the rebound candidate intervals are defined, they are listed separately as independent data segment record tables. These tables sequentially list the start time number, trend reversal time node number, and end time number of each rebound candidate interval, along with the interfacial adsorption amount and bulk concentration values ​​corresponding to each time node within the interval. The time numbering method is consistent with the unified time scale sequence, and no data deletion or replacement is performed. This creates a complete data system during the rapid concentration decline phase, including the original time series record table, the unified time scale expansion results, and the time node markers indicating the transition from a downward to an upward trend in interfacial adsorption. The rebound candidate intervals are clearly identified within this data system, providing a continuous, complete, and unprocessed data foundation for subsequent dynamic segment screening and interfacial occupancy change analysis based on these rebound candidate intervals.

[0025] Step 2: Simultaneously compare the rate of change of interfacial adsorption amount and the magnitude of change of bulk concentration within the rebound candidate range. Under the condition of continuous decrease in bulk concentration, screen out the dynamic segment where the interfacial adsorption amount still increases, determine the dynamic segment as the real dynamic segment and retain its original change trajectory. The steps to identify the dynamic segment as a true dynamic segment and preserve its original change trajectory are as follows: For the interfacial adsorption capacity and bulk concentration values ​​that are already arranged consecutively by time number within the rebound candidate interval, the difference between the interfacial adsorption capacity values ​​corresponding to every two adjacent time numbers is read one by one. Using the interfacial adsorption capacity value of the previous time number as a reference, the change in the interfacial adsorption capacity value of the subsequent time number relative to the previous time number is calculated, and this change is marked as the rate of change of interfacial adsorption capacity for the corresponding time interval. After the rate of change of interfacial adsorption capacity is generated point by point, the bulk concentration values ​​are read in the same time number order, and the difference between the bulk concentration values ​​corresponding to every two adjacent time numbers is calculated to form the magnitude of change of bulk concentration for the corresponding time interval. Subsequently, the rate of change of interfacial adsorption capacity for each time interval and the magnitude of change of bulk concentration for the same time interval are arranged in a one-to-one correspondence to form a synchronous comparison record table containing four items: time start number, time end number, interfacial adsorption capacity rate of change value, and bulk concentration magnitude of change value. In this record table, the time numbers are kept in a continuously increasing order, without inserting extra time points or deleting any time intervals, thereby ensuring that all time intervals within the rebound candidate interval have paired data on the rate of change of interfacial adsorption capacity and the magnitude of change of bulk concentration.

[0026] After the synchronous comparison record table is generated, the volumetric concentration change amplitude values ​​are checked sequentially according to the time number. When the volumetric concentration change amplitude values ​​are all less than zero in multiple consecutive time intervals, these consecutive time intervals are marked as a continuous volumetric concentration decrease segment, and the start and end time numbers of the continuous volumetric concentration decrease segment are recorded. After determining the continuous volumetric concentration decrease segment, the interfacial adsorption change rate value is read for each time interval within the segment. When the interfacial adsorption change rate value is greater than zero in a time interval where the volumetric concentration change amplitude value is less than zero, the time interval is marked as the interfacial adsorption increase time interval, and the time interval is specially marked in the synchronous comparison record table. At the same time, the interfacial adsorption change rate value and the volumetric concentration change amplitude value of the two consecutive time intervals before and after this time interval are recorded to ensure that the determination of the interfacial adsorption increase time interval is based on continuous change data, rather than a single isolated time point.

[0027] After marking the time intervals for the increase in interfacial adsorption, adjacent time intervals with a time difference of one are merged according to the continuity of time numbers. Two or more consecutive time intervals with an interfacial adsorption rate of change greater than zero and a bulk concentration change amplitude of less than zero are integrated into a complete segment, and the start and end time numbers of this complete segment are recorded. After integration, the original interfacial adsorption values ​​and bulk concentration values ​​corresponding to all time nodes in the complete segment are read one by one and rearranged according to the original arrangement order of the unified time scale sequence to form the interfacial adsorption change trajectory and the bulk concentration change trajectory of the segment. During the recording process, the values ​​are kept as they are, without any form of smoothing, replacement, or deletion processing, so that the segment fully reflects the actual change state of continuous increase in interfacial adsorption under the condition of continuous decrease in bulk concentration. Thus, a set of dynamic segments that meet the condition of continuous decrease in bulk concentration and continuous increase in interfacial adsorption are formed within the rebound candidate interval.

[0028] After the dynamic segment set is formed, each complete segment is confirmed to satisfy the condition of continuous decrease in bulk concentration and positive change rate of interfacial adsorption at least three time intervals. Complete segments that meet this condition are identified as true dynamic segments, and all original change trajectories within the true dynamic segments are saved separately, including the start time number, end time number, interfacial adsorption value and bulk concentration value corresponding to each time node. At the same time, the true dynamic segments are clearly marked in the overall rebound candidate interval record table so that they can be directly referenced in the subsequent interfacial occupancy change gradient analysis and kinetic expression process. This completes the synchronous comparison of the change rate of interfacial adsorption and the change amplitude of bulk concentration within the rebound candidate interval. Under the condition of continuous decrease in bulk concentration, dynamic segments that still show an increase in interfacial adsorption are screened out, and these dynamic segments are identified as true dynamic segments and their original change trajectories are completely preserved.

[0029] Step 3: Focusing on the real dynamic segment, analyze the gradient of interface occupancy change at adjacent time points, calculate the proportion of the interface adsorption recovery in the overall downward trend, and construct an interface occupancy enhancement index. The steps to build UI placeholder reinforcement metrics are as follows: Based on the start and end time numbers of the real dynamic segment within a unified time scale sequence, the interfacial adsorption value corresponding to each consecutive time node within that time range is extracted one by one. These values ​​are arranged in ascending order of time number to form a continuous numerical sequence. Within this sequence, adjacent time nodes are grouped together, and point-by-point difference calculations are performed. The interfacial adsorption value of the later time node is subtracted from that of the earlier time node to obtain the interfacial occupancy change gradient value within the corresponding time interval. This gradient value is then bound to the corresponding start and end time numbers to form a complete continuous record table of interfacial occupancy change gradients. This table retains all interfacial occupancy change gradient values ​​without smoothing, deleting negative or positive value segments, or compressing the values. This ensures that every rise and fall of interfacial adsorption in the real dynamic segment over time is fully presented as an interfacial occupancy change gradient. The table also indicates the number of time intervals with positive and negative interfacial occupancy change gradients, thus creating a continuous gradient distribution throughout the entire process of interfacial occupancy change within the real dynamic segment.

[0030] After forming a continuous record table of interface occupancy change gradients for the actual dynamic segments, based on the position of the actual dynamic segment in the overall time series, the interface adsorption values ​​of the ten consecutive time nodes before the start time number of the actual dynamic segment are read forward, and the interface adsorption values ​​of the ten consecutive time nodes after the end time number of the actual dynamic segment are read backward. These twenty time nodes are merged with the time nodes within the actual dynamic segment to form an overall time range record table containing the actual dynamic segment and the time segments before and after it. In this overall time range record table, the interface occupancy change gradient is calculated point by point in the same way as described above, forming a continuous record table of interface occupancy change gradients with an overall downward trend. Continue recording the sequence, and then separately calculate the increase in interface adsorption corresponding to the time intervals in the actual dynamic segment where the interface occupancy change gradient is positive. Sum these increases one by one to obtain the total increase in interface adsorption. At the same time, calculate the decrease in interface adsorption corresponding to the time intervals in the overall time range where the interface occupancy change gradient is negative. Sum these decreases one by one to obtain the total decrease in interface adsorption in the overall downward trend. During the accumulation process, the time numbering order is kept consistent to ensure that each increase and decrease comes from the actual record in the same time scale sequence, thus forming two specific values: the total increase in interface adsorption and the total decrease in interface adsorption in the overall downward trend.

[0031] After obtaining the total increase in interfacial adsorption and the total decrease in interfacial adsorption in the overall downward trend, the ratio of the total increase in interfacial adsorption to the total decrease in interfacial adsorption in the overall downward trend is calculated to obtain the proportion of the increase in interfacial adsorption in the overall downward trend. This proportion is then integrated with the start and end time numbers of the actual dynamic segment and the corresponding continuous record table of interfacial occupancy change gradient to form a data set that includes time range identifiers, interfacial occupancy change gradient distribution data, the total increase in interfacial adsorption, the total decrease in interfacial adsorption in the overall downward trend, and the ratio of the increase in interfacial adsorption to the decrease in interfacial adsorption in the overall downward trend. The interface occupancy enhancement index is recorded in a table containing five components: the percentage increase in the overall downward trend. This table maintains perfect consistency between the time numbering and the unified time scale sequence, ensuring that the interface occupancy enhancement index not only reflects the continuous distribution of the interface occupancy change gradient within the actual dynamic segment but also the quantitative proportion of the increase in interface adsorption within the overall downward trend. This transforms the degree of interface occupancy change enhancement within the actual dynamic segment into a quantifiable interface occupancy enhancement index, providing clear data support for assigning independent weights to the actual dynamic segment based on the interface occupancy enhancement index in the subsequent expression of interface dynamics.

[0032] Step 4: Embed the interface placeholder enhancement index into the interface dynamics expression process, and assign independent weights to the real dynamic segments during the parameter acquisition stage to distinguish them from non-real dynamic segments participating in the fitting calculation. The steps for assigning independent weights to the actual dynamic segments during the parameter calculation phase are as follows: With a unified time-scale sequence already constructed and interfacial adsorption values ​​arranged consecutively by time number, the interfacial adsorption values ​​across the entire time range are sequentially imported into a data recording table representing the interfacial kinetics process. This data recording table includes columns for time number, interfacial adsorption value, and segment identifiers. The start and end times of the actual dynamic segments in the interfacial occupancy enhancement index recording table are read one by one, and all time nodes within that time range are marked as actual dynamic segments in the segment identifier column. Simultaneously, time nodes outside this time range in the unified time-scale sequence are marked as non-actual dynamic segments. The interface dynamics expression process is divided into segments, which are real dynamic segments and non-real dynamic segments existing in parallel in the time dimension. After the segment identification is completed, the total increase of interface adsorption, the total decrease of interface adsorption in the overall downward trend, and the proportion of the increase of interface adsorption in the overall downward trend are filled into the additional information column of the interface dynamics expression process data record table, and kept completely consistent with the time number of the real dynamic segment. Thus, an embedded relationship of one-to-one correspondence between the interface occupancy enhancement index and the real dynamic segment is formed in the data structure of the interface dynamics expression process.

[0033] Before the parameter determination phase begins, a weight value column is added to the data recording table of the interface dynamics expression process. Different weight values ​​are assigned to different time nodes according to the markings in the segment identifier column. For time nodes marked as non-real dynamic segments in the segment identifier column, their weight values ​​are uniformly set to a fixed baseline weight value, for example, set to one. For time nodes marked as real dynamic segments in the segment identifier column, their weight values ​​are set to the baseline weight value multiplied by the proportion of the interface adsorption recovery rate in the overall downward trend. This product result is then filled into the weight value column of each time node corresponding to the real dynamic segment. During the filling process, all time nodes between the start time number and the end time number of the real dynamic segment are kept to use the same weight value. The original arrangement order of the interface adsorption value and the time number is not changed. The difference between real dynamic segments and non-real dynamic segments is only reflected through the weight value column, so that the interface occupancy enhancement index is directly embedded into the data participation method of the parameter determination phase in the form of weight values.

[0034] After all the weight values ​​are filled in, following the established parameter retrieval process for the interface dynamics expression, a complete data record table containing four items—time number, interface adsorption value, segment identifier, and weight value—is read. During parameter retrieval, the interface adsorption values ​​are called sequentially according to the time number, and the degree of interface adsorption participation at each time point is weighted according to the weight value of that time point. This ensures that truly dynamic segments, by being assigned independent weights, reflect an influence ratio in the parameter retrieval results that matches the proportion of the interface occupancy enhancement index, while non-truthful dynamic segments participate in the parameter retrieval according to the baseline weight value. In this approach, no time node data within the actual dynamic segment is deleted during the entire parameter calculation phase, nor is the interface adsorption value smoothed or replaced. Instead, the participation is differentiated solely through weight values. This results in a parameter value record table that includes the weighted participation. The record table retains three corresponding pieces of information: the time number range of the actual dynamic segment, the percentage value of the interface occupancy enhancement index, and the final parameter value. This completes the process of embedding the interface occupancy enhancement index into the interface dynamics expression process. Furthermore, the actual dynamic segment is assigned independent weights during the parameter calculation phase to distinguish it from non-actual dynamic segments participating in the fitting calculation.

[0035] Step 5: Based on the interfacial kinetics expression results after assigning independent weights, recalculate the upper limit of adsorption capacity and the turning point of saturation process, and revise the adsorbent dosage determination rule according to the recalculation results; The steps for revising the adsorbent dosage determination rule based on the recalculated results are as follows: In the interfacial dynamics expression result record table after assigning independent weights, the predicted interfacial adsorption amount for each time node is read sequentially from smallest to largest according to the time number, forming a complete adsorption process change sequence. In this adsorption process change sequence, five consecutive time nodes are taken as an observation unit, and the increment of the predicted interfacial adsorption amount between adjacent time nodes is compared one by one. When the increment of the predicted interfacial adsorption amount between five consecutive time nodes is within a pre-set stable range, for example, the increase in each time interval does not exceed one percent of the predicted value of the previous time node, and the duration of this state reaches no less than ten consecutive time nodes, the consecutive time nodes are considered stable. The starting time number of the segment is recorded as the turning point of the saturation process. At the same time, the predicted value of the interface adsorption capacity corresponding to the time number is recorded as the stage saturation reference value. Based on this, time nodes are read one by one. When the fluctuation range of the predicted value of the interface adsorption capacity around the stage saturation reference value does not exceed 2% of the reference value within the subsequent twenty consecutive time nodes, the stage saturation reference value is determined as the upper limit value of adsorption capacity. The upper limit of adsorption capacity and the turning point of saturation process are added to the interface kinetic expression result recording table, so that the upper limit of adsorption capacity and the turning point of saturation process are directly derived from the interface kinetic expression results after being assigned independent weights.

[0036] After recording the upper limit of adsorption capacity and the turning point of saturation, the predicted values ​​of interfacial adsorption amount corresponding to the time numbers in the original interfacial kinetic expression results without independent weights were extracted one by one and arranged in the same time number order as the adsorption process change sequence after independent weights were assigned. The differences between the two sets of adsorption process change sequences at the upper limit of adsorption capacity and the turning point of saturation were recorded side by side. Then, in the adsorbent dosage determination rule, the upper limit of adsorption capacity after independent weights was taken as the maximum adsorption amount that a unit adsorbent can carry under the current operating conditions, and the turning point of saturation was taken as the time boundary parameter of the effective operating cycle of the adsorbent. Under the condition that the treatment scale is a fixed flow rate and the influent concentration is a fixed value, the adsorbent dosage required for a single cycle was recalculated according to the correspondence between the total amount of pollutants entering the system per unit time and the upper limit of adsorption capacity. The calculated dosage values ​​were recorded item by item in the dosage adjustment record table, and the original dosage values ​​and the corrected dosage values ​​were compared and arranged, so that the determination of the adsorbent dosage is directly affected by the interfacial kinetic expression results after independent weights were assigned.

[0037] After calculating and recording the corrected dosage value, the three parameters—the upper limit of adsorption capacity, the turning point of the saturation process, and the corrected dosage value—are simultaneously written into the parameter list of the adsorbent dosage determination rule. During subsequent operation, whenever the interfacial kinetic expression results are updated, the upper limit of adsorption capacity and the turning point of the saturation process are recalculated according to the above steps, and the corresponding parameters in the adsorbent dosage determination rule are updated with the latest calculation results. This ensures that the adsorbent dosage is always determined based on the interfacial kinetic expression results with independent weights. This completes the specific implementation process of recalculating the upper limit of adsorption capacity and the turning point of the saturation process based on the interfacial kinetic expression results with independent weights, and correcting the adsorbent dosage determination rule according to the recalculation results. This ensures that the adsorbent dosage is consistent with the interfacial occupancy enhancement characteristics reflected by the actual dynamic segment, avoiding deviations in the judgment of the upper limit of adsorption capacity and fluctuations in the effluent concentration in the later stages of operation due to ignoring the actual dynamic segment.

[0038] Beneficial effect 1: This invention identifies the time point when the interfacial adsorption amount changes from a decreasing trend to an increasing trend during the rapid concentration decline phase, and filters out the real dynamic segment under the condition of continuous decrease in bulk concentration. This avoids misjudging the transient response of interfacial dynamics as abnormal noise and eliminating it. Thus, the information on the recovery of interfacial adsorption amount is completely preserved in the process of expressing interfacial dynamics, so that the adsorption behavior prediction model can truly reflect the interfacial occupancy enhancement characteristics, improve the accuracy of adsorption capacity and saturation process judgment, and enhance the consistency between prediction results and actual operating conditions.

[0039] Benefit 2: This invention constructs an interface occupancy enhancement index and assigns independent weights to the actual dynamic segments during the parameter calculation stage. This allows the interface dynamics expression results to fully reflect the contribution ratio of the interface adsorption capacity recovery rate to the overall downward trend. Based on this, the upper limit of adsorption capacity and the turning point of the saturation process are recalculated, and the adsorbent dosage determination rules are corrected accordingly. This improves the rationality of adsorbent dosage setting, reduces the risk of sudden increase in effluent concentration in the later stage of operation, and enhances the stability and continuous operation capability of the treatment process.

[0040] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A method for constructing a PFAS adsorption behavior prediction model based on interfacial kinetic data fitting, characterized in that, Includes the following steps: Step 1: During the PFAS adsorption process, collect continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration during the rapid decrease phase. Expand the continuous change data of interfacial adsorption amount and synchronous change data of bulk concentration on a unified time scale, and mark the time node when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form a rebound candidate interval. Step 2: Simultaneously compare the rate of change of interfacial adsorption amount and the magnitude of change of bulk concentration within the rebound candidate range. Under the condition of continuous decrease in bulk concentration, screen out the dynamic segment where the interfacial adsorption amount still increases, determine the dynamic segment as the real dynamic segment and retain its original change trajectory. Step 3: Focusing on the real dynamic segment, analyze the gradient of interface occupancy change at adjacent time points, calculate the proportion of the interface adsorption recovery in the overall downward trend, and construct an interface occupancy enhancement index. Step 4: Embed the interface placeholder enhancement index into the interface dynamics expression process, and assign independent weights to the real dynamic segments during the parameter acquisition stage, so as to distinguish them from non-real dynamic segments participating in the fitting calculation. Step 5: Based on the interfacial kinetics expression results after assigning independent weights, recalculate the upper limit of adsorption capacity and the turning point of saturation process, and revise the adsorbent dosage determination rule according to the recalculation results.

2. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 1, characterized in that, Marking the time points when the interfacial adsorption amount changes from a decreasing trend to an increasing trend to form rebound candidate intervals includes the following steps: During the adsorption process, the continuous value of the interfacial adsorption amount changes with time is continuously recorded, and the continuous value of the bulk concentration changes with time is recorded at the same time point. When the bulk concentration continuously decreases, the starting time of the rapid concentration drop phase is determined. From this starting time, continuous data is collected at fixed time intervals to form an original time series record table containing time number, interfacial adsorption amount value and bulk concentration value. Based on the original time series record table, the time number is expanded with a unified time scale. The interfacial adsorption value and the bulk concentration value are filled into the continuous time axis sequence. The changes in the interfacial adsorption value of adjacent time nodes are compared under the continuous time axis sequence to determine and mark the time node when the interfacial adsorption value changes from a downward trend to an upward trend. Extend consecutive time nodes forward and backward in the continuous time axis sequence around the time node where the marking is completed, delineate the continuous time segment containing the time node, and keep the interfacial adsorption amount and bulk concentration value recorded as is in the continuous time axis sequence. The defined continuous time intervals are listed in an independent data interval record table, recording the start time number, trend reversal time node number, and end time number, and fully retaining the interfacial adsorption amount value and bulk concentration value corresponding to each time node within the interval.

3. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 2, characterized in that, Within a continuous time interval, the values ​​of interfacial adsorption amount are kept in their original order of change without deletion, replacement, or smoothing. The direction of change of interfacial adsorption amount and the direction of change of bulk concentration are recorded synchronously in the independent data segment record table.

4. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 2, characterized in that, Determining the dynamic segment as a true dynamic segment and preserving its original change trajectory involves the following steps: For the interfacial adsorption amount values ​​and bulk concentration values ​​arranged continuously by time number within the rebound candidate interval, the difference of interfacial adsorption amount values ​​corresponding to adjacent time numbers is calculated to form the rate of change of interfacial adsorption amount, and the difference of bulk concentration values ​​is calculated according to the same time number order to form the magnitude of change of bulk concentration, and a synchronous comparison record table is generated. Based on the synchronous comparison record table, the segments in which the change in bulk concentration is continuously less than zero are identified in chronological order. Within these segments, time intervals in which the rate of change in interfacial adsorption is greater than zero are selected and marked to form a set of time intervals in which the interfacial adsorption increases. The set of time intervals for the increase in interfacial adsorption amount is integrated according to the continuity of time number to form a continuous and complete segment, and the corresponding interfacial adsorption amount value and bulk concentration value within the complete segment are extracted to form the change trajectory. Complete segments that meet the condition of continuous decrease in bulk concentration and positive rate of change of interfacial adsorption amount are identified. These complete segments that meet the conditions are determined as the true dynamic segments, and all original change trajectories within the true dynamic segments are retained.

5. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 4, characterized in that, The actual dynamic segment is identified by continuous time number in the synchronous comparison record table, and the start time number and end time number of the actual dynamic segment are listed separately in the overall rebound candidate interval record table. The interfacial adsorption value and the bulk concentration value within the actual dynamic segment are kept in their original order without deletion.

6. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 4, characterized in that, Building UI placeholder enhancement metrics involves the following steps: Based on the start and end time numbers of the real dynamic segment in a unified time scale sequence, the interface adsorption value within the corresponding time range is extracted. The difference between adjacent time nodes is calculated according to the time number order to form the interface occupancy change gradient, and a continuous record table of interface occupancy change gradient is generated. Based on the continuous record table of interface occupancy change gradient, the interface adsorption amount values ​​of adjacent time nodes in the real dynamic segment are read forward and backward to construct the overall time range record table. The interface occupancy change gradient is calculated in the same way, and the increase in interface adsorption amount in the real dynamic segment and the decrease in interface adsorption amount in the overall time range are statistically analyzed. The ratio of the increase to the decrease in interfacial adsorption is calculated to obtain the percentage of the increase in interfacial adsorption in the overall downward trend. This percentage is then combined with the actual dynamic time intervals and the continuous record table of interfacial occupancy change gradient to form an interfacial occupancy enhancement index record table.

7. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 6, characterized in that, The continuous record table of interface occupancy change gradient retains all positive and negative value segments. The increase and decrease of interface adsorption are both derived from the original records in the unified time scale sequence and are associated with the start and end time numbers of the actual dynamic segments.

8. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 6, characterized in that, Assigning independent weights to the actual dynamic segments during the parameter calculation phase includes the following steps: Import the interfacial adsorption values ​​from the unified time scale sequence into the data recording table of the interfacial dynamic expression process, set the time number column and the segment identifier column, and mark the time nodes according to the start time number and end time number of the real dynamic segment in the interfacial occupancy enhancement index recording table. At the same time, write the total increase of interfacial adsorption and the proportion of the increase of interfacial adsorption in the overall downward trend into the additional information column. Based on the segment identifier column, a weight value column is set. A baseline weight value is assigned to the time node of the non-real dynamic segment, and the baseline weight value is assigned to the time node of the real dynamic segment multiplied by the percentage of the interface adsorption amount recovery in the overall downward trend to form an independent weight value. The parameters are obtained based on the data record table containing time number, segment identifier and weight value. The participation of the interface adsorption amount is weighted according to the weight value to form a parameter value result record table containing the range of real dynamic segment time number and the proportion of interface occupancy enhancement index value.

9. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 8, characterized in that, The independent weight values ​​corresponding to the actual dynamic segments remain consistent throughout the parameter calculation process, and the total increase in interface adsorption and the percentage of the increase in interface adsorption in the overall downward trend are simultaneously marked in the parameter value result record table.

10. The method for constructing a PFAS adsorption behavior prediction model based on interface kinetic data fitting according to claim 8, characterized in that, The steps for revising the adsorbent dosage determination rule based on the recalculated results are as follows: Read the predicted values ​​of interfacial adsorption amount arranged by time number in the interfacial kinetic expression result record table after assigning independent weights, construct the adsorption process change sequence, determine the saturation process transition time based on the change of the predicted values ​​of interfacial adsorption amount at consecutive time nodes, determine the upper limit value of adsorption capacity, and write the upper limit value of adsorption capacity and the saturation process transition time into the interfacial kinetic expression result record table. Based on the upper limit of adsorption capacity and the turning point of saturation, combined with the treatment scale and influent concentration conditions, the adsorbent dosage is calculated according to the correspondence between the total amount of pollutants per unit time and the upper limit of adsorption capacity, and the corrected dosage value is generated and recorded in the dosage adjustment record table. The upper limit of adsorption capacity, the turning point of saturation process, and the corrected dosage value are written into the parameter list of the adsorbent dosage determination rule, and the corresponding parameters in the adsorbent dosage determination rule are updated synchronously when the interface kinetic expression results are updated.

Images