Multi-filter-array activated carbon self-adaptive gradient purification method and filtering device

Abstract

The application discloses a multi-filter-element array active carbon self-adaptive gradient purification method and a filtering device, relates to the technical field of water pollution prevention and treatment, and comprises the following steps: based on an array state matrix, a pollutant breakthrough phase tracking algorithm is adopted to convert front breakthrough dynamic characteristic quantities in continuous time into front phase points to perform displacement matching, construct a front migration rate function, and perform difference sequence arrangement to generate a front space distribution curve; a filter element section with concentrated distribution of the front position in the front space distribution curve is identified and marked as a front aggregate section; based on the front aggregate section, gradient position rotation is performed on the front space distribution curve to obtain a gradient cooperative distribution structure; three-stage purification water path reconstruction and flow redistribution of the filter element array are performed according to the gradient cooperative distribution structure, and the array state matrix is updated. The application can inhibit the concentration fluctuation of the purified water pollutants through staggered flow and differential distribution of the hierarchical flow, prolong the effective operation cycle of the filter element, and reduce the consumption and operation and maintenance costs.

Description

Technical Field

[0001] This invention relates to the field of water pollution control technology, and in particular to an adaptive gradient purification method and filtration device using multi-filter array activated carbon. Its IPC main classification number is mainly C02F9, and the keyword is high-efficiency activated carbon. Background Technology

[0002] The field of multi-stage industrial wastewater purification has long relied on porous adsorption media for deep purification. Among these, activated carbon, as a representative adsorption material, has seen continuous development due to its large specific surface area, adjustable pore size distribution, and broad-spectrum adsorption capacity for various pollutants. In recent years, with continuous production and increasingly stringent emission standards, purification devices have gradually evolved from single-tube and fixed-bed types to multi-filter parallel and array-based systems. On the one hand, parallel connection of multiple filter tubes increases treatment capacity and device throughput; on the other hand, multi-stage purification strategies—main filtration, deep purification, and safety protection—are implemented through graded flow and bypass switching to cope with inlet concentration disturbances and operating condition fluctuations. In recent years, online monitoring and automated control technologies have been introduced into adsorption process management. Common practices include collecting operating parameters such as pollutant concentrations, pressure differentials, flow rates, and operating times in the inlet and outlet water of the filter cartridges, and combining these with actuators such as valve on / off and branch switching to adjust the operating status.

[0003] From the perspective of current technological practices, there are still two key shortcomings that are highly related to purification stability and utilization efficiency. On the one hand, multi-filter parallel devices are prone to uneven flow distribution and short-circuit flow phenomena during long-term operation, resulting in some filter cartridges being overloaded and the adsorption front breaking through prematurely, while the remaining filter cartridges remain under low load. This leads to increased fluctuations in the concentration of pollutants in the purified water, more obvious peak responses, and increased replacement frequency. On the other hand, existing control strategies mostly rely on single-point threshold-triggered switching bypass actions, which make it difficult to characterize the speed and spatial distribution of the adsorption front migration along the axial direction, and even more difficult to identify the concentrated advancement of the front at the filter cartridge section level and make targeted adjustments. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides an adaptive gradient purification method for activated carbon in multi-filter arrays to solve the problems of internal load imbalance and uncontrollable gradient structure in filter arrays.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides an adaptive gradient purification method for activated carbon in a multi-filter array, comprising: collecting wastewater filter operation data and generating an array state matrix through feature standardization and integration; based on the array state matrix, employing a pollutant breakthrough phase tracking algorithm to convert the dynamic feature quantity of the breakthrough at the forefront within a continuous time period into the frontline phase point for displacement matching to construct the frontline migration rate function, and performing differential ordering to generate a frontline spatial distribution curve; identifying filter segments with concentrated frontline positions in the frontline spatial distribution curve and marking them as frontline aggregation segments; based on the frontline aggregation segments, performing gradient position rotation on the frontline spatial distribution curve to obtain a gradient cooperative distribution structure; and performing three-stage purification water path reconstruction and flow redistribution of the filter array according to the gradient cooperative distribution structure, and updating the array state matrix.

[0008] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the wastewater filter cartridge operation data includes the influent and effluent pollutant concentrations, pressure difference, flow rate, and operation time data of the filter cartridge;

[0009] The feature standardization integration includes time synchronization calibration, sliding window statistical processing, and multi-parameter normalization integration.

[0010] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the array state matrix includes filter operating status and load information.

[0011] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the step of using a pollutant breakthrough phase tracking algorithm based on the array state matrix to convert the dynamic feature quantity of the breakthrough front in continuous time into the front phase point for displacement matching to construct the front migration rate function is as follows:

[0012] Extract the leading edge breakthrough dynamic features of each filter element in the array state matrix at adjacent time nodes, and use the pollutant breakthrough phase tracking algorithm to perform directional consistency encoding on the leading edge breakthrough dynamic features to generate a filter element leading edge phase point sequence;

[0013] Perform phase point sorting and migration segment compression on the phase point sequence at the leading edge of the filter element to generate an axial phase migration index sequence;

[0014] A front migration rate function is constructed based on the axial migration index sequence execution time-axial coupling calibration.

[0015] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the specific steps for generating the leading edge spatial distribution curve are as follows:

[0016] Based on the front migration rate function, the relative load sequence in the standardized state sequence of the filter element is established by the front differential order reconstruction algorithm, and the differential order is arranged to generate the differential order of the filter element.

[0017] Based on the differential order of the filter elements, the array state matrix is ​​spatially sequenced and mapped to generate the leading edge spatial distribution curve.

[0018] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the specific steps for identifying filter segments with concentrated distribution at the leading edge position in the leading edge spatial distribution curve are as follows:

[0019] The leading edge spatial distribution curve is divided into continuous segments to obtain the continuous segments of the filter element;

[0020] Based on the continuous segment set of the filter element, the local density discrimination algorithm is used to identify the local clustering of the filter element's operating status and load information, and to obtain the leading concentrated distribution segment.

[0021] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the leading edge aggregation section is obtained by physically marking the filter element number based on the set of leading edge concentrated distribution sections.

[0022] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the step of performing gradient position rotation on the leading edge spatial distribution curve based on the leading edge aggregation segment to obtain the gradient cooperative distribution structure is as follows:

[0023] Based on the frontal spatial distribution curve, segmented flow regulation is performed on the frontal aggregation section to generate a filter element misaligned flow operation structure;

[0024] Based on the filter element misaligned flow operation structure, the filter element flow state is updated according to the arrangement order of the leading edge spatial distribution curves to obtain the gradient cooperative distribution structure.

[0025] As a preferred embodiment of the multi-filter array activated carbon adaptive gradient purification method of the present invention, the specific steps of reconstructing the three-stage purification water path and redistributing the flow according to the gradient cooperative distribution structure, and updating the array state matrix are as follows:

[0026] The leading edge peak shaving and staggered peak reconstruction method is used to perform three-stage purification water path reconstruction and branch connection reorganization on the gradient cooperative distribution structure to generate an array flow path topology;

[0027] Based on the array flow path topology, a hierarchical flow redistribution structure is generated by combining a gradient cooperative distribution structure with differentiated proportional allocation.

[0028] The filter array operates based on a graded flow redistribution structure, and wastewater filter operation data is re-collected, standardized, and integrated to generate an updated array state matrix.

[0029] Secondly, this invention provides a multi-filter array activated carbon adaptive gradient purification filtration device, comprising: an adsorption operation monitoring module, which collects wastewater filter operation data and generates an array state matrix through feature standardization integration; an adsorption front analysis module, which, based on the array state matrix, uses a pollutant breakthrough phase tracking algorithm to convert the dynamic feature quantity of the front breakthrough in continuous time into the front phase point for displacement matching to construct the front migration rate function, and performs differential sorting to generate a front spatial distribution curve; a front aggregation identification module, which identifies filter segments with concentrated front positions in the front spatial distribution curve and marks them as front aggregation segments; a gradient adjustment module, which, based on the front aggregation segments, performs gradient position rotation on the front spatial distribution curve to obtain a gradient cooperative distribution structure; and a flow path allocation module, which performs three-stage purification water path reconstruction and flow redistribution of the filter array according to the gradient cooperative distribution structure and updates the array state matrix.

[0030] The beneficial effects of this invention are as follows: By uniformly integrating the concentration of pollutants in the inlet and outlet water of the filter element, the pressure difference, flow rate, and operating time, a comparable array state characterization is formed. Based on this, the migration and spatial distribution of the adsorption front are plotted, enabling early identification of the concentrated advancement of the front and directional allocation of the aggregation section. Furthermore, by combining staggered flow and graded flow differential allocation, the fluctuation of pollutant concentration and peak response in the purified water are suppressed, extending the effective operating cycle of the filter element and reducing consumable and maintenance costs. At the same time, the multi-filter element array, combined with the guided flow method, extends the contact process between the medium and activated carbon, reducing the risk of leakage and backflow and reducing downtime and manual intervention. The combination of the regeneration path and the double-sealed quick-connect structure further reduces maintenance downtime. Attached Figure Description

[0031] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0032] Figure 1 This is a flowchart of a multi-filter array activated carbon adaptive gradient purification method.

[0033] Figure 2 This is a schematic diagram of a multi-filter array activated carbon adaptive gradient purification filtration device.

[0034] Figure 3 A flowchart for identifying frontier clustering zones.

[0035] Figure 4 A flowchart for generating a gradient cooperative distribution structure.

[0036] Figure 5 This is a comparison chart of the stability of total effluent contaminant concentration in the array.

[0037] Figure 6 This is a comparison chart of the dynamic response of influent and effluent concentrations.

[0038] In the diagram: 1. Filter tube; 2. Inlet; 3. Outlet; 4. Sealing ring. Detailed Implementation

[0039] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0040] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0041] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0042] Reference Figures 1-6 This is one embodiment of the present invention, which provides a multi-filter array activated carbon adaptive gradient purification method, including the following steps:

[0043] S1: Collect wastewater filter cartridge operation data and generate an array state matrix through feature standardization integration.

[0044] S1.1: Wastewater filter cartridge operation data includes the concentration of pollutants in the influent and effluent of the filter cartridge, pressure difference, flow rate and operation time data.

[0045] Specifically, the influent and effluent pollutant concentration data are obtained by continuously sampling through online liquid concentration detection devices installed at the influent and effluent ends of each filter element. The online liquid concentration detection devices detect the concentration of the industrial wastewater to be treated before entering the filter element and the concentration of the effluent liquid after flowing through the filter element in real time, and record the corresponding values ​​according to a unified sampling time, thereby forming the influent and effluent pollutant concentrations of the corresponding filter element at the same time point.

[0046] Differential pressure data is obtained by arranging pressure detection devices at the inlet and outlet ends of each filter element. The pressure values ​​at the inlet and outlet ends are collected separately, and the difference between the pressure values ​​at the inlet and outlet ends at the same time point is used as the differential pressure data of the filter element at that time point.

[0047] Flow data is collected by a flow detection device in the inlet branch of each filter element. The flow detection device continuously measures the instantaneous volumetric flow rate through the filter element and records it synchronously at time points that are consistent with the inlet and outlet water pollutant concentration data and pressure difference data, thus forming the flow data of the filter element at the corresponding time point.

[0048] The start time of the filter cartridge is recorded during operation, and the running time is continuously accumulated during operation. At the same time, the current accumulated running time is read according to the same time node as the influent and effluent pollutant concentration data, pressure difference data, and flow rate data, and used as the running time data at the corresponding time node.

[0049] S1.2: Feature standardization integration includes time synchronization calibration, sliding window statistical processing, and multi-parameter normalization integration.

[0050] Specifically, the time synchronization calibration sorts the sampling time corresponding to the wastewater filter cartridge operation data, selects a fixed time interval as a unified time reference sequence, and assigns each filter cartridge a set of influent and effluent pollutant concentration data, pressure difference data, flow rate data and operation time data at each unified time node to obtain time-aligned wastewater filter cartridge operation data.

[0051] A fixed-length time window is constructed for the influent and effluent pollutant concentration data, pressure difference data, and flow rate data of the same filter element at multiple consecutive unified time points. The corresponding filter element operating parameters are summed point by point, and the number of data points within the window is counted. The ratio of the summation result of the filter element operating parameters point by point to the number of data points within the window is used as the average value of the time window. The operating time data remains unchanged as the original cumulative record after time synchronization calibration, thus obtaining the wastewater filter element operating data after sliding window statistical processing.

[0052] The maximum and minimum values ​​of influent and effluent pollutant concentration, pressure difference, flow rate, and operating time for all filter cartridges at the current time node are extracted. The ratio of the difference between the current value and the minimum value to the difference between the maximum value and the minimum value is used as the linearly normalized wastewater filter cartridge operating data, so that parameters of different dimensions are converted into a unified numerical range. The normalized influent and effluent pollutant concentration data, normalized pressure difference data, normalized flow rate data, and normalized operating time data of the same filter cartridge at the same time node are combined and arranged in chronological order to obtain the filter cartridge operating status and load information.

[0053] S1.3: The array status matrix includes filter element operating status and load information.

[0054] Specifically, the operating status of the filter element is characterized by the concentration data of pollutants in the influent and effluent water and the pressure difference data. By extracting and combining the normalized influent liquid concentration data, effluent liquid concentration data and pressure difference data at a unified time point, operating status data reflecting the current adsorption effect and resistance changes of the filter element are formed.

[0055] Load information is characterized by normalized flow data and operating time data. By extracting flow data and cumulative operating time data at the same time point and combining them, load data reflecting the load intensity and usage of the filter element is formed.

[0056] The operating status data and load data of the same filter element at the same time node are arranged in a fixed order and used as a row of data in the array status matrix to form an array status matrix containing filter element operating status and load information.

[0057] It should be noted that the array state matrix is ​​used to uniformly and structurally express the operating status and load information of each filter element corresponding to its number. This allows the data on influent and effluent pollutant concentrations, pressure differences, flow rates, and operating times to form a comparable overall data set within the same time reference and unified dimension range. This provides a basic data basis for establishing the relative load sequence relationship between filter elements and for constructing subsequent frontal spatial distribution curves.

[0058] like Figure 2 The filter tube 1 shown is the core structure, filled with highly efficient activated carbon adsorption material, which is the core carrier for pollutant adsorption. Multiple filter tubes 1 are arranged in an orderly manner along the physical axis to form a filter element array. Each filter tube 1 is an independent filtration unit, corresponding to a single filter element number in the array state matrix, providing a structural basis for subsequent three-stage purification water path switching and staggered flow adjustment. Inlet 2 and outlet 3 are the two ends of the flow interface of filter tube 1. Inlet 2 is the inlet of the industrial wastewater to be treated, and outlet 3 is the outlet of the purified liquid. The inlet 2 and outlet 3 of a single filter tube 1 are respectively connected to the liquid inlet branch and liquid outlet branch of the device. The filter array is connected to the interstage pipelines. The inlet 2 and outlet 3 of the filter array are also connected to online monitoring, pressure and flow detection devices. At the same time, the pipeline layout of inlet 2 and outlet 3 is adapted to the valve on / off and branch reorganization requirements of the flow path adjustment module to ensure the construction of the staged flow path. The sealing ring 4 is a sealing matching structure, which is mainly embedded in the connection sealing surface between filter tube 1 and inlet 2 and outlet 3, and at the joint node between filter tube 1 and pipeline. It provides liquid tightness guarantee for the overall flow structure and is an important component of the sealed flow structure, ensuring that the liquid flows in a directional manner along inlet 2-filter tube 1-outlet 3 without leakage or backflow.

[0059] It should be noted that the internal flow path of filter tube 1 is configured with a guiding structure to guide the industrial wastewater to be treated to form a guided extended flow or extend the flow path within filter tube 1, thereby improving the sufficient contact between the industrial wastewater and the activated carbon adsorption material and reducing the tendency of bypass flow. In some embodiments, filter tube 1 or its corresponding branch is configured with a regeneration passage for performing backwashing or regeneration treatment on the activated carbon adsorption material during maintenance, thereby reducing pressure differential accumulation and extending the effective operating cycle of filter tube 1. In some embodiments, the sealing ring 4 is a double-seal structure and is matched with a quick-connect structure to reduce the risk of leakage during disassembly and assembly and shorten the downtime for replacing filter tube 1.

[0060] S2: Based on the array state matrix, the pollutant breakthrough phase tracking algorithm is used to convert the dynamic feature quantity of the breakthrough of the front edge in continuous time into the front edge phase point for displacement matching to construct the front edge migration rate function, and then perform differential sorting to generate the front edge spatial distribution curve.

[0061] S2.1: Extract the leading edge breakthrough dynamic features of each filter element in the array state matrix at adjacent time nodes, and use the pollutant breakthrough phase tracking algorithm to perform directional consistency encoding on the leading edge breakthrough dynamic features to generate the filter element leading edge phase point sequence.

[0062] Specifically, according to the time node order in the array state matrix, the filter cartridge operation status and load information corresponding to adjacent time nodes are extracted row by row for each filter cartridge number. The attenuation of the influent and effluent pollutant concentration difference, the increase of the pressure difference, and the change of flow load are calculated respectively. The attenuation of the influent and effluent pollutant concentration difference, the increase of the pressure difference, and the change of flow load are combined into the dynamic characteristic quantity of the front breakthrough, while retaining the corresponding time node index and filter cartridge number identifier.

[0063] After obtaining the dynamic characteristics of the breakthrough, the pollutant breakthrough phase tracking algorithm is used to record the breakthrough advancement state when the concentration difference of pollutants in the influent and effluent along the breakthrough dynamic characteristics shows a decreasing trend, the pressure difference shows an increasing trend, and the flow load shows an increasing trend; when the changing direction in the breakthrough dynamic characteristics along the current edge alternates, it is recorded as the breakthrough fluctuation state; when the changing amplitude in the breakthrough dynamic characteristics along the current edge is within a stable range, it is recorded as the breakthrough stable state.

[0064] The breakthrough propulsion state, breakthrough fluctuation state, and breakthrough stable state are combined with the corresponding time node index and filter element number to form a pollutant breakthrough phase coding result with time index and filter element number identifier; the pollutant breakthrough phase coding results formed by the same filter element number in all time node ranges are summarized and arranged in chronological order to generate a filter element time gradient coding sequence.

[0065] It should be noted that the principle of the pollutant breakthrough phase tracking algorithm is based on the changes in influent and effluent pollutant concentration difference, pressure difference, flow rate, and operating time of each filter element in the array state matrix at adjacent time nodes. This transforms the originally dispersed changes in water quality, resistance, and load into a time-axial variation relationship that can be used to determine the frontal migration rate. The significance of the pollutant breakthrough phase tracking algorithm lies in its ability to avoid judging filter element breakthrough solely based on the effluent concentration threshold. It allows the frontal migration rate function to reflect the pollutant propagation trend within the filter array earlier, providing a more stable basis for constructing the frontal spatial distribution curve, identifying frontal aggregation sections, and reconstructing the three-stage purification water path.

[0066] S2.2: Perform phase point sorting and migration segment compression on the phase point sequence at the leading edge of the filter element to generate an axial phase migration index sequence.

[0067] Specifically, based on the physical arrangement order of the filter element array in the flow direction, the axial position index corresponding to each filter element number is recorded to generate an axial position index table. Based on the axial position index table, the filter element time gradient coding sequence is read in a unified time node order, and the pollutant breakthrough phase coding results corresponding to each filter element number at the same time node are rearranged from the inlet side to the outlet side according to the axial position index to form the axial breakthrough phase coding row corresponding to the time node.

[0068] Multiple axial breakthrough phase coding rows arranged in chronological order are superimposed to form an axial breakthrough phase coding matrix; the axial breakthrough phase coding rows of adjacent time nodes are compared column by column in chronological order. When the pollutant breakthrough phase coding results of consecutive axial positions in adjacent time nodes are consistent or change synchronously, the corresponding axial position range is classified into the same migration segment.

[0069] The start time node index, end time node index, and corresponding axial position index range within the same migration segment are compressed and recorded, and all migration segments are summarized in chronological order to generate an axial migration index sequence.

[0070] S2.3: Construct a front migration rate function based on the axial migration index sequence execution time-axial coupling calibration.

[0071] Specifically, the starting time node index, ending time node index, and corresponding axial position index range of each migration segment recorded in the axial migration index sequence are read in chronological order; the time span between the starting time node index and the ending time node index is calculated to form a migration segment time span sequence; based on the axial position index range recorded in the axial migration index sequence, the axial position start index and axial position end index corresponding to each migration segment are extracted, and the midpoint between the axial position start index and the axial position end index is used as the front representative position to form a front representative position sequence arranged in chronological order.

[0072] The difference between the representative positions of the leading edge of adjacent migration segments is statistically analyzed to form a sequence of axial position changes. The time spans of adjacent migration segments are paired and recorded to form a time span pairing sequence. The time span in the time span pairing sequence is used as the independent variable, and the ratio of the corresponding axial position change to the time span is used as the function value to construct the leading edge migration rate function.

[0073] It should be noted that the leading edge migration rate function is a time-axial change mapping relationship constructed based on the start time node index, end time node index, and corresponding axial position index range of each migration segment in the axial migration index sequence. The leading edge migration rate function uses the ratio of the axial position change between adjacent migration segments to the corresponding time span as the function value, which is used to characterize the speed of the adsorption leading edge in the axial direction of the filter array, and to provide a time-dimensional trend reference for establishing the relative load sequence in the filter standardization state sequence.

[0074] S2.4: Based on the leading edge migration rate function, the relative load sequence in the standardized state sequence of the filter element is established by the leading edge differential order reconstruction algorithm, and the differential order is arranged to generate the differential order sequence of the filter element.

[0075] Specifically, the mapping relationship between time and axial position change rate recorded in the front migration rate function is read in the order of time node index. At the same time, the numerical sequence of filter element numbers at the corresponding time nodes in the filter element standardized state sequence is read according to the same time node index, so that the front migration rate function and the filter element standardized state sequence are aligned in the time dimension.

[0076] The frontier difference reconstruction algorithm is used to calculate the number of same-direction changes and the number of opposite-direction changes between the standardized state sequences of filter cartridges corresponding to any two filter cartridge numbers and the frontier migration rate function within the same time node range. The difference between the number of same-direction changes and the number of opposite-direction changes is used as the trend consistency value.

[0077] Sort the trend consistency values ​​corresponding to the filter element numbers from largest to smallest to obtain the load priority relationship; summarize the formed relative load priority relationships into a sortable relationship set and sort them in descending order according to the filter element numbers in the sortable relationship set to generate the filter element differential order.

[0078] It should be noted that the leading edge differential order reconstruction algorithm is a method to establish the relative load order relationship in the filter element standardized state sequence by comparing the consistency between the changing trend of the filter element standardized state sequence corresponding to each filter element number and the changing trend of the leading edge migration rate function, based on the alignment of the leading edge migration rate function with the completion time node of the filter element standardized state sequence. The leading edge differential order reconstruction algorithm uses the trend consistency as the sorting basis, performs pairwise comparisons on all filter element numbers and arranges them in differential order, and reconstructs the filter element differential order corresponding to the leading edge advancement direction.

[0079] S2.5: Based on the differential order of the filter elements, the array state matrix is ​​spatially sequenced and mapped to generate the leading edge spatial distribution curve.

[0080] Specifically, a spatial location index sequence is established according to the order of filter element numbers recorded in the filter element differential sequence, and the spatial location index sequence corresponds one-to-one with the filter element numbers in the filter element differential sequence; the filter element number data rows of each time node in the array state matrix are read according to the time node index order, and the data rows of the corresponding filter element numbers in the array state matrix are rearranged according to the filter element differential sequence.

[0081] After the array state matrix is ​​rearranged, the filter cartridge operating status and load information are extracted from the data row corresponding to each spatial location index, and combined according to a fixed field order to form a spatial representation value sequence that corresponds one-to-one with the spatial location index. Using the spatial location index sequence as the horizontal coordinate and the spatial representation value sequence as the vertical value, the spatial representation values ​​corresponding to each spatial location index are connected in sequence to form a continuous change curve, and finally the leading edge spatial distribution curve is generated.

[0082] S3: Identify filter cartridge segments with concentrated frontal positions in the frontal spatial distribution curve and mark them as frontal aggregation segments.

[0083] S3.1: Divide the leading edge spatial distribution curve into continuous segments to obtain the continuous segments of the filter element.

[0084] Specifically, the filter cartridge number and corresponding numerical characterization data for each position are extracted sequentially according to the horizontal arrangement order in the frontal spatial distribution curve. The difference between the numerical characterization data of two adjacent positions in the frontal spatial distribution curve is calculated, and the adjacent position difference sequence is obtained.

[0085] The ratio of all differences in the sequence of differences between adjacent positions to the total number of differences is used as the average difference, and this average difference is used as the basis for determining continuous segments. Following the arrangement of the leading edge spatial distribution curves, each adjacent position difference is compared with the average difference. When the difference between adjacent positions is less than the average difference, the filter cartridge numbers corresponding to the two positions are grouped into the same continuous segment. When the difference between adjacent positions is greater than the average difference, the next position is used as the starting point of a new segment to restart the segmentation, thus obtaining continuous filter cartridge segments.

[0086] S3.2: Based on the continuous segment set of the filter element, the local density discrimination algorithm is used to locally aggregate and identify the filter element's operating status and load information to obtain the leading edge concentrated distribution segment.

[0087] Specifically, based on the set of continuous filter segments, the operating status and load information of each continuous filter segment are extracted one by one according to the arrangement order in the frontal spatial distribution curve. Then, a local density discrimination algorithm is executed on each continuous filter segment to calculate the corresponding local density value, expressed as:

[0088] ;

[0089] in, This represents the local density value. Indicates the first The number of filter elements contained in a continuous section of a filter element. Indicates the first The first filter element in the continuous segment arranged according to the leading edge spatial distribution curve is... Numerical representation data corresponding to each location This serves as an index for the arrangement order of spatial distribution curves. To indicate the first The average value of all numerical characterization data within a continuous section of a filter element.

[0090] It should be noted that the dimensionless count value of the number of filter cartridges in the expression has the same numerical value representing a linear combination of the differences and deviations between the data, and the dimensions are consistent. The whole is a dimensionless quantity. Therefore, the dimensions of the formula for calculating the local density value are consistent.

[0091] After obtaining the local density values ​​corresponding to all continuous sections of the filter element, the local density values ​​of each continuous section of the filter element are compared, and the continuous section of the filter element with the highest local density value is determined as the front-end concentrated distribution section.

[0092] It should be noted that the local density discrimination algorithm refers to the quantitative evaluation of the degree of numerical concentration within a unit segment based on the similarity of the operating status and load information corresponding to adjacent filter element numbers in the frontal spatial distribution curve within the continuous segment set of filter elements, and the continuous segment with the highest degree of numerical concentration is identified as the frontal concentrated distribution segment.

[0093] S3.3: Based on the set of front-end concentrated distribution segments, perform physical segment marking on the filter element number to obtain the front-end concentrated segments.

[0094] Specifically, the filter element numbers corresponding to the concentrated distribution segments in the frontal zone are extracted one by one according to the arrangement order in the frontal spatial distribution curve, and the physical arrangement position of the corresponding filter element in the filter element array is determined according to the row position of the filter element number in the array state matrix.

[0095] A unified numbering segment label is applied to the physical arrangement position. Continuous filter cartridge numbers belonging to the same set of front-end concentrated distribution segments are assigned the same segment identifier, and the corresponding filter cartridge rows in the array state matrix are labeled as the same segment category to obtain the front-end clustering segments.

[0096] S4: Based on the frontier clustering segment, perform gradient position rotation on the frontier spatial distribution curve to obtain the gradient cooperative distribution structure.

[0097] S4.1: Based on the leading edge spatial distribution curve, segmented flow regulation is performed on the leading edge aggregation section to generate a filter element misaligned flow operation structure.

[0098] Specifically, based on the leading edge spatial distribution curve corresponding to the leading edge aggregation section, the curve position corresponding to each filter element number within the leading edge aggregation section is retrieved one by one according to the arrangement order of the leading edge spatial distribution curves to form a leading edge aggregation section position sequence; according to the order of the leading edge aggregation section position sequence, the filter element number corresponding to the first position of the leading edge aggregation section position sequence is assigned to the first flow segment sequence, the filter element number corresponding to the second position of the leading edge aggregation section position sequence is assigned to the second flow segment sequence, the filter element number corresponding to the third position of the leading edge aggregation section position sequence is assigned to the first flow segment sequence, and so on, alternating until all filter element numbers are classified.

[0099] For each filter element in the first flow segment sequence, the main flow operation is performed sequentially (including opening the filter element inlet branch passage, opening the filter element outlet branch passage, closing the filter element bypass passage, and adjusting the filter element branch regulating valve to the fully open state); for each filter element in the second flow segment sequence, the maintenance flow operation is performed sequentially (including keeping the filter element inlet branch passage open, keeping the filter element outlet branch passage open, keeping the filter element bypass passage closed, and adjusting the filter element branch regulating valve to the small opening state to form a low flow state), forming alternating high flow and low flow within the same leading edge aggregation section and generating a filter element staggered flow operation structure.

[0100] For example, when the front accumulation zone position sequence includes filter element number 1, filter element number 2, filter element number 3 and filter element number 4, the first flow segment sequence includes filter element number 1 and filter element number 3 and performs the main flow operation, and the second flow segment sequence includes filter element number 2 and filter element number 4 and performs the maintenance flow operation.

[0101] It should be noted that the flow segment sequence refers to the flow bearing order table formed by dividing the continuously arranged filter element numbers into several continuous segments according to their spatial positions based on the arrangement order of the filter element numbers in the frontal spatial distribution curve, and using the segment order to represent the sequential flow relationship of each filter element segment in the array flow path topology.

[0102] S4.2: Based on the filter element misaligned flow operation structure, the filter element flow state is updated according to the arrangement order of the leading edge spatial distribution curves to obtain the gradient cooperative distribution structure.

[0103] Specifically, based on the staggered flow operation structure of the filter element, the filter element number at the corresponding position is read one by one according to the arrangement order of the front spatial distribution curve, and the correspondence between the filter element number and the current position index is established according to the order in the front spatial distribution curve.

[0104] According to a fixed rotation cycle, the first flow segment sequence and the second flow segment sequence are subjected to a flow state exchange operation (from main flow operation to maintenance flow operation or from maintenance flow operation to main flow operation); the filter cartridge number currently in the main flow operation is released from the main flow operation and switched to the maintenance flow operation; at the same time, the filter cartridge number currently in the maintenance flow operation state is released from the maintenance flow operation and switched to the main flow operation, so that the filter cartridges complete the sequential update of the flow state in sequence under the constraint of the arrangement order of the frontal spatial distribution curve.

[0105] The next round of flow state exchange is performed according to the arrangement order of the frontal spatial distribution curve. After all the filter cartridge numbers in the current aggregation section have undergone the alternation of main flow operation and maintenance flow operation, a continuous decreasing arrangement relationship of alternating high flow rate and low flow rate is formed under the sequential framework of the frontal spatial distribution curve, and a gradient cooperative distribution structure is obtained.

[0106] It should be noted that the gradient collaborative distribution structure is used to transform the filter element arrangement relationship after position rotation in the frontal spatial distribution curve into a hierarchical structure with a progressive undertaking relationship, so that the main filtration stage, the deep purification stage and the security stage form a continuously decreasing adsorption and propulsion relationship in the flow sequence, thereby providing a clear level division basis for the switching of the three-stage purification water circuit and the differentiated flow redistribution, and ensuring that the filter element array maintains a stable gradient purification state during operation.

[0107] S5: Based on the gradient collaborative distribution structure, perform the three-stage purification water path reconstruction and flow redistribution of the filter array, and update the array state matrix.

[0108] S5.1: The leading edge peak shaving and staggered peak reconstruction method is used to perform three-level purification water path reconstruction and branch connection reorganization on the gradient cooperative distribution structure to generate an array flow path topology.

[0109] Specifically, based on the filter cartridge number, gradient level order, and front-end aggregation section marker recorded in the gradient collaborative distribution structure, the main filtration level filter cartridge number set, the deep purification level filter cartridge number set, and the security level filter cartridge number set are extracted respectively.

[0110] Based on the changes in pollutant concentration, pressure difference, and current flow rate of each filter element in the purified effluent, the peak shaving adjustment priority of each filter element in the leading accumulation zone is determined. The filter element with the highest peak shaving adjustment priority is configured as a low flow rate maintenance branch or a downstream buffer branch, and the filter elements that are not in the leading accumulation zone and have a high load margin are configured as main flow branches.

[0111] Following the flow sequence of the main filtration stage, the deep purification stage, and the safety stage, the inlet branch valves, outlet branch valves, and interstage connecting pipe valves of the corresponding filter elements are controlled to open and close, connecting the outlet of the main filtration stage to the inlet of the deep purification stage and the outlet of the deep purification stage to the inlet of the safety stage. This also allows the leading-edge and non-leading-edge collecting filter elements to share the inlet load in the three-stage purification water circuit, generating an array flow path topology structure in which the main filtration stage, the deep purification stage, and the safety stage are connected in sequence.

[0112] It should be noted that the leading-edge peak shaving and staggered-peak reconstruction method is based on a gradient collaborative distribution structure and leading-edge aggregation sections. It considers the leading-edge propulsion status of each filter element, changes in pollutant concentration in the purified water, pressure difference changes, and flow rate load ratio. Filter elements with concentrated leading-edge propulsion and prone to generating effluent peaks are staggered from high-load continuous flow positions, while filter elements with higher load margins are added to the main flow path. This creates a three-stage purification water path structure with staggered peak loads for the main filtration stage, deep purification stage, and safety stage. The leading-edge peak shaving and staggered-peak reconstruction method can disperse pollutant peaks caused by premature exhaustion of local filter elements into multiple purification paths for buffering, transforming the pollutant concentration in the purified water from sharp fluctuations to gradual changes, while simultaneously improving the load balance and shock load resistance of the filter array.

[0113] S5.2: Based on the array flow path topology, combined with the gradient collaborative distribution structure, a differentiated proportional allocation is performed to generate a hierarchical flow redistribution structure.

[0114] Specifically, according to the arrangement order of the main filtration stage filter element number set, the deep purification stage filter element number set, and the security stage filter element number set in the gradient collaborative distribution structure, the positions of the inlet branch regulating valve and the outlet branch regulating valve corresponding to the filter element number in the array flow path topology are respectively located.

[0115] The current opening position of each branch control valve is read, and the opening of the branch control valve containing the main filter element number set is used as the reference opening value. After the reference opening value is determined, the opening of the branch control valve is adjusted. The target opening position is controlled by the electric control valve actuator according to the proportional calculation. For example, the opening of the branch control valve containing the deep purification filter element number set is adjusted to half of the reference opening value, and the opening of the branch control valve containing the safety filter element number set is adjusted to one-third of the reference opening value.

[0116] According to the different flow distribution ratios in the hierarchical order of the gradient cooperative distribution structure, the total flow of all branches is checked so that the sum of the flow of each branch after adjustment is equal to the total inflow of liquid. In the array flow path topology, a hierarchical flow bearing ratio relationship corresponding to the gradient cooperative distribution structure is formed, and a hierarchical flow redistribution structure is generated.

[0117] S5.3: The filter array is operated based on the graded flow redistribution structure, and the wastewater filter operation data is re-collected, and the features are standardized and integrated to generate an updated array state matrix.

[0118] Specifically, based on the hierarchical flow redistribution structure, the filter array is continuously operated according to the branch regulating valve opening ratios corresponding to the main filter stage filter element number set, the deep purification filter element number set, and the security filter element number set in the hierarchical flow redistribution structure, while maintaining the stability of the array flow path topology.

[0119] During the operation of the filter array, the wastewater filter operation data corresponding to each filter number is collected in real time. The wastewater filter operation data includes the concentration of pollutants in the influent and effluent of the filter, pressure difference, flow rate and operation time data. After the wastewater filter operation data collection is completed, the wastewater filter operation data within the same time section is time-synchronized and calibrated so that the data corresponding to each filter number is under the same time reference.

[0120] The wastewater filter cartridge operation data after time synchronization calibration are arranged in chronological order to form an equal time interval sampling sequence. A fixed number of continuous sampling points are used as a sliding window length. Starting from the first sampling point, continuous sampling points are extracted to form the first window segment. Within the first window segment, the influent and effluent pollutant concentrations, pressure difference, flow rate, and operation time data are arithmetically averaged and recorded as the sliding window statistical results (including influent liquid concentration statistics, effluent liquid concentration statistics, pressure difference statistics, flow rate statistics, and operation time statistics).

[0121] Extract the maximum and minimum values ​​of the influent liquid concentration statistics, effluent liquid concentration statistics, differential pressure statistics, flow rate statistics, and running time statistics. At the same statistical moment, use the statistical values ​​of the same type corresponding to all filter element numbers as the calculation range to obtain the overall maximum and minimum statistical values ​​of the corresponding parameters. Record the difference between the current statistical value and the overall minimum statistical value as the offset of the current statistical value. Record the difference between the overall maximum statistical value and the overall minimum statistical value as the width of the statistical value variation range. Calculate the ratio of the current statistical value offset to the statistical value variation range width, ensuring the result falls between zero and one.

[0122] The normalized parameters are arranged in order of filter element number to form filter element operating status and load information within a unified dimension range, thereby completing the multi-parameter normalization integration; the normalized filter element operating status and load information are written into the array state matrix according to the filter element number order to generate and update the array state matrix.

[0123] like Figure 5The figure shows the overall trend of the total effluent pollutant concentration of the multi-filter array activated carbon device over time under continuous operation. The horizontal axis represents time, and the vertical axis represents the total effluent pollutant concentration. It can be seen that during the inlet disturbance stage, under the traditional fixed rotation or no gradient position rotation operation mode, the total effluent pollutant concentration exhibits significant fluctuations and peaks. However, after adopting the array state matrix construction—frontal differential reconstruction—frontal aggregation segment identification—gradient position rotation operation process of this invention, the overall fluctuation amplitude of the total effluent pollutant concentration is reduced, the peak response is effectively suppressed, and the recovery process is smoother. By spatially reorganizing and misaligning the adsorption front, the risk of local failure caused by premature breakthrough of a single filter element can be avoided, improving the overall adsorption stability of the array and delaying the overall effluent concentration increase trend, thus verifying the effectiveness of this invention in suppressing the concentrated advancement of the front. A magnified view further refines the comparison of the perturbation range, allowing for a clearer observation of the differences in peak height, peak duration, and recovery slope among different operating modes. Notably, the operating curve employing the adaptive gradient control strategy has a lower peak position than the control group and a shorter fluctuation range, indicating a more balanced distribution of adsorption load among the multiple filter elements, thus avoiding the risk of localized breakthroughs caused by concentrated adsorption front advancement. These results demonstrate that spatial reorganization and misalignment control of the adsorption front can improve the overall adsorption stability of the array and enhance its buffering capacity against perturbation shocks, validating the effectiveness of this invention in suppressing concentrated front advancement.

[0124] like Figure 6 The figure shows a dynamic comparison between the inlet pollutant concentration and the total effluent pollutant concentration at the same time scale. The horizontal axis represents time, and the vertical axis represents pollutant concentration. It can be observed that when the inlet pollutant concentration experiences a step increase or pulse disturbance, the total effluent pollutant concentration does not respond synchronously with equal amplitude, but rather exhibits a slow-release change characteristic, with the overall amplitude lower than the inlet disturbance amplitude. This phenomenon indicates that, under the control of the array state matrix and gradient position rotation, the multi-filter array achieves a balanced distribution of adsorption load, forming a gradient unfolding structure between different filter elements at the adsorption front, thereby enhancing the buffering capacity against instantaneous shock loads. This figure further demonstrates that this invention, by structurally expressing the relative adsorption load relationship of multiple filter elements and implementing adaptive gradient adjustment, can effectively reduce effluent concentration fluctuations and improve the overall purification stability and anti-disturbance capability of the array.

[0125] This embodiment also provides a multi-filter array activated carbon adaptive gradient purification filtration device, including: an adsorption operation monitoring module, which collects wastewater filter operation data and generates an array state matrix through feature standardization integration; an adsorption front analysis module, which, based on the array state matrix, uses a pollutant breakthrough phase tracking algorithm to convert the dynamic feature quantity of the front breakthrough in continuous time into the front phase point for displacement matching to construct the front migration rate function, and performs differential sorting to generate a front spatial distribution curve; a front aggregation identification module, which identifies filter segments with concentrated front positions in the front spatial distribution curve and marks them as front aggregation segments; a gradient adjustment module, which, based on the front aggregation segments, performs gradient position rotation on the front spatial distribution curve to obtain a gradient cooperative distribution structure; and a flow path allocation module, which performs three-stage purification water path reconstruction and flow redistribution of the filter array according to the gradient cooperative distribution structure and updates the array state matrix.

[0126] In summary, this invention achieves the following: by uniformly integrating the concentration of pollutants in the influent and effluent of the filter cartridge, pressure difference, flow rate, and operating time, a comparable array state characterization is formed. Based on this, the migration and spatial distribution of the adsorption front are plotted, enabling early identification of the concentrated advance of the front and targeted allocation of aggregation sections. Furthermore, by combining staggered flow and graded flow differential allocation, fluctuations in the concentration of pollutants in the purified effluent and peak response are suppressed, extending the effective operating cycle of the filter cartridge and reducing consumable and maintenance costs. At the same time, the multi-filter cartridge array, combined with the guided flow method, extends the contact process between the medium and activated carbon, reducing the risk of leakage and backflow and minimizing downtime and manual intervention. The combination of the regeneration path and the double-sealed quick-connect structure further reduces maintenance downtime.

[0127] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A multi-filter array activated carbon adaptive gradient purification method, characterized in that, include: Collect wastewater filter cartridge operation data and generate an array state matrix through feature standardization and integration; Based on the array state matrix, the pollutant breakthrough phase tracking algorithm is used to convert the dynamic feature quantity of the breakthrough of the front edge in continuous time into the front edge phase point for displacement matching to construct the front edge migration rate function, and then sort them by difference to generate the front edge spatial distribution curve. Identify filter cartridge segments with concentrated frontal positions in the frontal spatial distribution curve and mark them as frontal aggregation segments; Based on the front clustering segment, the gradient position rotation of the front spatial distribution curve is performed to obtain the gradient cooperative distribution structure; Based on the gradient collaborative distribution structure, the three-stage purification water path reconstruction and flow redistribution of the filter array are performed, and the array state matrix is ​​updated.

2. The multi-filter array activated carbon adaptive gradient purification method as described in claim 1, characterized in that: The wastewater filter cartridge operation data includes the concentration of pollutants in the influent and effluent of the filter cartridge, pressure difference, flow rate, and operation time. The feature standardization integration includes time synchronization calibration, sliding window statistical processing, and multi-parameter normalization integration.

3. The multi-filter array activated carbon adaptive gradient purification method as described in claim 2, characterized in that: The array status matrix includes filter cartridge operating status and load information.

4. The multi-filter array activated carbon adaptive gradient purification method as described in claim 1, characterized in that, The method based on the array state matrix, using the pollutant breakthrough phase tracking algorithm, converts the dynamic features of the breakthrough front over a continuous time period into the front phase points for displacement matching to construct the front migration rate function. The specific steps are as follows: Extract the leading edge breakthrough dynamic features of each filter element in the array state matrix at adjacent time nodes, and use the pollutant breakthrough phase tracking algorithm to perform directional consistency encoding on the leading edge breakthrough dynamic features to generate a filter element leading edge phase point sequence; Perform phase point sorting and migration segment compression on the phase point sequence at the leading edge of the filter element to generate an axial phase migration index sequence; A front migration rate function is constructed based on the axial migration index sequence execution time-axial coupling calibration.

5. The multi-filter array activated carbon adaptive gradient purification method as described in claim 4, characterized in that, The specific steps for generating the frontier spatial distribution curve are as follows: Based on the front migration rate function, the relative load sequence in the standardized state sequence of the filter element is established by the front differential order reconstruction algorithm, and the differential order is arranged to generate the differential order of the filter element. Based on the differential order of the filter elements, the array state matrix is ​​spatially sequenced and mapped to generate the leading edge spatial distribution curve.

6. The multi-filter array activated carbon adaptive gradient purification method as described in claim 1, characterized in that, The specific steps for identifying filter segments with concentrated frontal positions in the frontal spatial distribution curve are as follows: The leading edge spatial distribution curve is divided into continuous segments to obtain the continuous segments of the filter element; Based on the continuous segment set of the filter element, the local density discrimination algorithm is used to identify the local clustering of the filter element's operating status and load information, and to obtain the leading concentrated distribution segment.

7. The multi-filter array activated carbon adaptive gradient purification method as described in claim 6, characterized in that: The aforementioned leading edge aggregation section is obtained by physically marking the filter element number based on the set of leading edge concentrated distribution sections.

8. The multi-filter array activated carbon adaptive gradient purification method as described in claim 7, characterized in that, The step of performing gradient position rotation on the leading edge spatial distribution curve based on the leading edge aggregation segment to obtain the gradient cooperative distribution structure is as follows: Based on the frontal spatial distribution curve, segmented flow regulation is performed on the frontal aggregation section to generate a filter element misaligned flow operation structure; Based on the filter element misaligned flow operation structure, the filter element flow state is updated according to the arrangement order of the leading edge spatial distribution curves to obtain the gradient cooperative distribution structure.

9. The multi-filter array activated carbon adaptive gradient purification method as described in claim 1, characterized in that, The steps for reconstructing the three-stage water purification path and redistributing the flow rate of the filter array based on the gradient cooperative distribution structure, and updating the array state matrix are as follows: The leading edge peak shaving and staggered peak reconstruction method is used to perform three-stage purification water path reconstruction and branch connection reorganization on the gradient cooperative distribution structure to generate an array flow path topology; Based on the array flow path topology, a hierarchical flow redistribution structure is generated by combining a gradient cooperative distribution structure with differentiated proportional allocation. The filter array operates based on a hierarchical flow redistribution structure, and the filter operating parameter data is re-collected, standardized, and integrated to generate an updated array state matrix.

10. A multi-filter array activated carbon adaptive gradient purification filtration device, based on the multi-filter array activated carbon adaptive gradient purification method according to any one of claims 1 to 9, characterized in that, include: The adsorption operation monitoring module collects wastewater filter cartridge operation data and generates an array state matrix through feature standardization and integration. The adsorption front analysis module, based on the array state matrix, uses a pollutant breakthrough phase tracking algorithm to convert the dynamic characteristics of the front breakthrough in continuous time into front phase points for displacement matching to construct the front migration rate function, and then sorts them by difference to generate the front spatial distribution curve. The leading edge aggregation identification module identifies filter cartridge segments with concentrated leading edge positions in the leading edge spatial distribution curve and marks them as leading edge aggregation segments; The gradient adjustment module, based on the leading edge aggregation segment, performs gradient position rotation on the leading edge spatial distribution curve to obtain the gradient cooperative distribution structure; The flow path allocation module performs the three-stage purification water path reconstruction and flow redistribution of the filter array according to the gradient collaborative distribution structure, and updates the array state matrix.

Images