Data fusion processing method and system applied to volcanic rock water purification system

Abstract

The invention provides a data fusion processing method and system applied to a volcanic rock water purification system, and the method comprises the steps: firstly obtaining volcanic rock filter material pore evolution data, pollutant adsorption migration data and filter tank operation feedback data, then carrying out the time sequence correlation capture processing of the three types of data, recognizing a form transition node and a transition amplitude, and carrying out the data fusion processing. The method comprises the following steps: constructing a dynamic association network based on transition nodes and amplitudes, serially connecting an interaction relationship of filter material pores, pollutant migration and operation parameter adjustment, generating a filter material state adaptive adjustment rule according to the dynamic association network, and describing an optimal matching mode of the filter material pores and operation parameters in different pollutant migration states. And according to the filter material state self-adaptive adjustment rule, an operation parameter dynamic correction instruction is generated, and the operation parameter dynamic correction instruction comprises filter speed and backwashing period adjustment parameters, so that the purification efficiency and stability of the volcanic rock water purification system can be improved.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and more specifically, to a data fusion processing method and system applied to a volcanic rock water purification system. Background Technology

[0002] In the field of operation and management of volcanic rock water purification systems, traditional methods mainly focus on the monitoring and analysis of single data points. For example, for volcanic rock filter media, only the initial pore structure parameters are typically tested periodically, while the dynamic changes in the pore morphology of the filter media over time during the water purification process are ignored. Regarding pollutant treatment, attention is often focused only on the initial concentration and final removal rate of pollutants in the water, lacking in-depth research and monitoring of the specific location changes and morphological transformation processes of pollutants within the filter media layer. For filter operation, it is generally carried out according to a preset fixed filtration rate and backwash cycle, rarely considering the actual impact of filtration rate adjustments and backwashing operations on the filter media's condition.

[0003] Existing methods lack comprehensive correlation analysis of data on volcanic rock filter media pore evolution, pollutant adsorption and migration, and filter operation feedback, making it difficult to fully and accurately grasp the actual operating status of volcanic rock water purification systems. This leads to the inability to adjust operating parameters in a timely manner according to the system's dynamic changes during actual operation, easily resulting in problems such as filter media clogging and unstable pollutant removal efficiency. Consequently, the overall purification efficiency and stability of volcanic rock water purification systems are affected, failing to meet increasingly stringent wastewater treatment requirements and water reuse standards. Summary of the Invention

[0004] In view of the aforementioned problems, and in conjunction with the first aspect of the present invention, embodiments of the present invention provide a data fusion processing method applied to a volcanic rock water purification system, the method comprising: Acquire pore evolution data of volcanic rock filter media, pollutant adsorption and migration data, and filter operation feedback data. The pore evolution data of volcanic rock filter media records the temporal changes in the pore morphology between filter media particles as the water purification process progresses. The pollutant adsorption and migration data records the positional changes and morphological transformations of pollutants in the water body within the filter media layer. The filter operation feedback data records the changes in the state of the filter media caused by filtration rate adjustments and backwashing operations. The pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool are subjected to time-series correlation and capture processing to identify the morphological transformation nodes and transformation magnitudes of the pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool in the same time process. A dynamic correlation network is constructed based on the morphological transformation nodes and the transformation amplitude. The dynamic correlation network connects the interaction relationships of filter material pore morphology changes, pollutant migration path changes and operating parameter adjustments with time flow as the axis. The filter media state adaptive adjustment rules are generated based on the dynamic correlation network. The adaptive adjustment rules are used to describe the optimal matching mode between the filter media pore morphology and operating parameters under different pollutant migration states. Based on the adaptive adjustment rules for the filter media state, dynamic correction instructions for the operating parameters of the volcanic rock water purification system are generated. These dynamic correction instructions include real-time filtration rate adjustment parameters and backwashing cycle adjustment parameters.

[0005] Furthermore, embodiments of the present invention also provide a data fusion processing system applied to a volcanic rock water purification system, characterized in that it includes: A processor; a machine-readable storage medium for storing machine-executable instructions of the processor; wherein the processor is configured to perform the above-described data fusion processing method for a volcanic rock water purification system by executing the machine-executable instructions.

[0006] In another aspect, embodiments of the present invention also provide a computer program product, the computer program product including machine-executable instructions, the machine-executable instructions being stored in a computer-readable storage medium, the processor of a computer device reading the machine-executable instructions from the computer-readable storage medium, the processor executing the machine-executable instructions, causing the computer device to execute the above-described data fusion processing method applied to a volcanic rock water purification system.

[0007] Based on the above, by acquiring data on the pore evolution of volcanic rock filter media, pollutant adsorption and migration, and filter operation feedback, key information in the operation of the volcanic rock water purification system is comprehensively covered. Time-series correlation capture processing of these three types of data accurately identifies the morphological transformation nodes and magnitudes within the same timeframe, revealing the dynamic relationships between various system elements. A dynamic correlation network constructed based on morphological transformation nodes and magnitudes clearly connects the interactions between changes in filter media pore morphology, pollutant migration paths, and operational parameter adjustments along a time-flow axis. Adaptive adjustment rules for filter media state generated by the dynamic correlation network describe the optimal matching mode between filter media pore morphology and operational parameters under different pollutant migration states. Dynamic correction instructions for operational parameters generated based on the adaptive adjustment rules include real-time filtration rate adjustment parameters and backwash cycle adjustment parameters, enabling precise dynamic adjustment of the volcanic rock water purification system's operational parameters. This effectively improves the system's purification efficiency, stability, and adaptability, reduces operating costs, and meets wastewater treatment needs under different water quality and quantity conditions. Attached Figure Description

[0008] Figure 1 This is a schematic diagram of the execution flow of the data fusion processing method applied to the volcanic rock water purification system provided in the embodiments of the present invention.

[0009] Figure 2 This is a schematic diagram of exemplary hardware and software components of a data fusion processing system applied to a volcanic rock water purification system provided in an embodiment of the present invention. Detailed Implementation

[0010] The present invention will now be described in detail with reference to the accompanying drawings. Figure 1 This is a schematic flowchart of a data fusion processing method for a volcanic rock water purification system provided by an embodiment of the present invention. The following is a detailed description of the data fusion processing method for a volcanic rock water purification system.

[0011] Step S110: Obtain pore evolution data of volcanic rock filter media, pollutant adsorption and migration data, and filter operation feedback data. The pore evolution data of volcanic rock filter media records the temporal changes in the pore morphology between filter media particles as the water purification process progresses. The pollutant adsorption and migration data records the positional changes and morphological transformations of pollutants in the water within the filter media layer. The filter operation feedback data records the changes in the state of the filter media caused by filtration rate adjustments and backwashing operations.

[0012] In a municipal waterworks' volcanic rock filter deep water purification system, pore evolution data of the volcanic rock filter media is collected using microscopic imaging devices deployed at different depths within the filter. These devices scan the filter layer at preset time intervals, acquiring image sequences containing pore structures. The images record dynamic changes in pore geometry, connectivity paths, and other information during the purification process. Pollutant adsorption and migration data are collected by multi-parameter water quality sensors installed at the filter inlet, different filter layer depths, and the outlet. These sensors monitor the spatial distribution, concentration gradient, and chemical transformation of pollutants in the water in real time, including changes in functional groups of organic pollutants and valence state transitions of heavy metal ions. Filter operation feedback data is automatically recorded by the filter's automatic control system, encompassing real-time operating parameters of the filtration rate adjustment device, start and stop signals of the backwashing equipment, changes in pressure loss of the filter layer before and after backwashing, and changes in filter layer height—parameters reflecting changes in the filter media's state.

[0013] Step S120: Perform time-series correlation capture processing on the pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool to identify the morphological transformation nodes and transformation magnitudes of the pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool in the same time process.

[0014] In the aforementioned scenario of a municipal waterworks volcanic rock filter deep water purification system, time-series correlation capture processing requires first unifying the time reference of the three types of data. Through timestamp alignment technology, the time records of data collected by the microscopic imaging device, water quality sensor, and automatic control system are adjusted to the same time reference system, ensuring data comparability across time dimensions. Subsequently, for each type of data sequence, trend analysis methods are used to identify the time points where significant turning points occur in the data change trends, i.e., morphological transformation nodes. The magnitude of the transformation is determined by calculating the degree of difference in data characteristic values ​​before and after the transformation node, such as the change in filter media pore volume, the proportion of change in pollutant concentration, and the rate of change in filtration rate adjustment.

[0015] Step S121: Extract the cross-sectional morphology time sequence and pore connectivity path time sequence of the filter media pores from the pore evolution data of the volcanic rock filter media. The cross-sectional morphology time sequence records the geometric shape data of the pores at different times, and the pore connectivity path time sequence records the connection mode data between the pores at different times.

[0016] In the above scenario, the image sequence of the pore evolution data of volcanic rock filter material is processed. First, the images are preprocessed by grayscale conversion and noise reduction to enhance the contrast between the pore area and the background. Then, an edge detection algorithm is used to identify the cross-sectional contour of the pores, and the geometric parameters of the contour are extracted, including the perimeter, inscribed circle diameter, and shape factor. These parameters at different times are arranged in chronological order to form a time-series sequence of cross-sectional morphology. For the time-series sequence of pore connectivity paths, the image is segmented to label each independent pore region. Then, the region adjacency analysis method is used to identify the connection channels between adjacent pores, and the number, distribution, and connection mode (such as direct connection, indirect connection, and isolated state) of the connection channels are recorded. The connection mode data at different times are arranged in chronological order to form a time-series sequence of pore connectivity paths.

[0017] Step S1211: Analyze the pore evolution data of the volcanic rock filter material, which includes a sequence of pore images at different depths of the filter material layer acquired through microscopic imaging technology.

[0018] In the above scenario, the pore evolution data of volcanic rock filter media is stored as a collection of image files, with each image file corresponding to a specific filter layer depth and acquisition time. The parsing process includes reading the metadata of the image files to obtain information such as the image acquisition time, corresponding filter layer depth, and resolution. Then, the image data is decoded, converting the image from compressed format into a raw pixel matrix to prepare for subsequent image segmentation and feature extraction.

[0019] Step S1212: Perform image segmentation processing on the pore image sequence to separate the pore region and non-pore region in each pore image.

[0020] In the above scenario, image segmentation employs a combination of thresholding and region growing. First, grayscale histogram analysis is performed on the pore image to determine an initial threshold for distinguishing between pore and non-pore regions. This threshold is then used to binarize the image, initially separating the pore regions. Next, morphological filtering is applied to the pore regions in the binarized image to remove noise interference and small pseudo-pore regions. Then, using the initially segmented pore regions as seed points, a region growing algorithm is employed. Based on the similarity and spatial continuity of pixel grayscale values, adjacent pore regions are further grown and merged, ultimately yielding a complete segmentation result separating the pore and non-pore regions.

[0021] Step S1213: For each segmented pore region, extract the geometric shape parameters of its cross-section, including the perimeter of the outline, the diameter of the inscribed circle, and the shape factor, which together constitute the geometric shape data of the pore.

[0022] In the above scenario, for each segmented pore region, the boundary contour of the pore region is extracted using a contour tracking algorithm. The sum of the distances between all consecutive pixels on the contour is calculated to obtain the contour perimeter. The largest inscribed circle that can be completely contained within the pore region is found inside the contour; the diameter of this largest inscribed circle is the inscribed circle diameter. The shape factor is calculated as the ratio of the square of the contour perimeter to the area of ​​the pore region, used to describe the irregularity of the pore shape; the smaller the ratio, the closer the pore shape is to a circle. The contour perimeter, inscribed circle diameter, and shape factor of each pore region are recorded as geometric shape data.

[0023] Step S1214: Arrange the geometric shape data according to the acquisition time sequence of the image sequence to form a time sequence of the cross-sectional morphology of the filter material pores.

[0024] In the above scenario, the geometric shape data extracted from the segmented pore images acquired at different times are arranged sequentially according to the order of image acquisition time. Each time point corresponds to a set of data containing the geometric shape parameters of all pore regions, thus forming a time-series sequence of cross-sectional morphology that changes over time.

[0025] Step S1215: Perform connectivity analysis on the pore image sequence to identify the connection channels between adjacent pores and the width and length of the connection channels.

[0026] In the above scenario, connectivity analysis is performed based on image segmentation. First, a distance transform is applied to the segmented binarized image, and the distance from each non-pore region pixel to the nearest pore region is calculated. Then, based on the distance transform results, a distance threshold is set, and non-pore regions whose distances to two or more pore regions are less than this threshold are identified as potential connectivity channels. Morphological dilation is performed on these potential connectivity channels to fill the small gaps within the channels. A skeleton extraction algorithm is then used to extract the central axis of the connectivity channel region; the length of this central axis is the length of the connectivity channel. At the start and end positions of the connectivity channel, the width of the channel is measured perpendicular to the central axis, and the average of multiple measurement points is taken as the width of the connectivity channel.

[0027] Step S1215-1: Binarize each image in the segmented pore image sequence, labeling the pore region as the first value and the non-pore region as the second value.

[0028] In the above scenario, the segmented pore image is a grayscale image. During binarization, the grayscale value of the pixel corresponding to the pore region is set to the first value, and the grayscale value of the pixel corresponding to the non-pore region is set to the second value, forming a binarized image containing only two values, which facilitates subsequent connectivity analysis and feature extraction.

[0029] Step S1215-2: The binarized image is preprocessed using morphological erosion operation, and the preprocessed image is analyzed using the connected component labeling algorithm to identify each independent pore region and assign a unique identifier.

[0030] In the above scenario, the morphological erosion operation scans the binarized image using a structuring element of a preset size, removing small protrusions and noise points on the boundaries of the pore regions, making the contours of the pore regions smoother. The preprocessed image is then processed using a connected component labeling algorithm. This algorithm traverses all pixels in the image, grouping interconnected pore region pixels into a connected component and assigning a unique identifier to each connected component, thus distinguishing and labeling each independent pore region.

[0031] Step S1215-3: Perform distance transformation processing on the marked aperture regions and calculate the distance from each non-aperture region pixel to the nearest aperture region.

[0032] In the above scenario, the distance transformation process uses the marked pore regions as the foreground and the non-pore regions as the background. For each non-pore region pixel, the Euclidean distance to the nearest pore region pixel is calculated, and this distance value is assigned to the non-pore region pixel to obtain the distance-transformed image. The gray value of each pixel in the image represents the distance of that pixel to the nearest pore region.

[0033] Step S1215-4: Identify potential connection channel regions based on the distance transformation results. Potential connection channel regions are non-porous regions that are close to two or more pore regions.

[0034] In the above scenario, the distance-transformed image is analyzed, a distance threshold is set, and pixels that are not in the aperture region are filtered out by selecting those with distance values ​​less than the threshold. For these pixels, it is further determined whether they are simultaneously close to two or more aperture regions with different identifiers, i.e., the distance from the pixel to two or more aperture regions is less than the distance threshold. The region formed by the set of pixels that meet the conditions is identified as a potential connection channel region.

[0035] Step S1215-5: Perform morphological dilation on the potential connection channel region, use the skeleton extraction algorithm to extract the central axis of the dilated connection channel region as the path of the connection channel, and measure the length of the central axis of the connection channel as the length parameter of the connection channel.

[0036] In the above scenario, the morphological dilation operation expands the potential connection channel region using a structuring element of a preset size, filling in the small breaks and gaps within the channel to create a continuous region. The skeleton of the dilated connection channel region is then extracted by continuously eroding the boundary pixels until the region shrinks to a single-pixel-wide central axis, which represents the path of the connection channel. Pixels are traversed along the central axis from start to finish, and the cumulative distance between pixels is calculated to obtain the length of the central axis of the connection channel, which serves as the length parameter of the connection channel.

[0037] Step S1215-6: At the start and end positions of the connecting channel, measure the channel width perpendicular to the central axis, take the average value as the width parameter of the connecting channel, and associate and store the start hole identifier, end hole identifier, length parameter and width parameter of each connecting channel to complete the identification of the connecting channel between adjacent holes.

[0038] In the above scenario, the starting and ending points of the connecting channel correspond to two different pore regions. At the starting point, the distances to the boundaries of the connecting channel are measured perpendicular to the central axis of the connecting channel; the sum of these two distances is the channel width at the starting point. The channel width at the ending point is measured using the same method. The average of the channel widths at the starting and ending points is taken as the width parameter of the connecting channel. The starting pore identifier, ending pore identifier, length parameter, and width parameter of each connecting channel are associated and stored in a database to achieve the identification and recording of connecting channels between adjacent pores.

[0039] Step S1216: Determine the connection method between the pores based on the distribution of the connection channels, including direct connection, indirect connection and isolated state.

[0040] In the above scenario, the connection method between pores is determined based on the identification results of the connection channels. If there is a direct connection channel between two pore regions, that is, they are directly connected through a single channel, then these two pore regions are in a directly connected state. If two pore regions are indirectly connected through multiple intermediate pore regions and connection channels, then these two pore regions are in an indirectly connected state. If there is no connection channel between a pore region and any other pore region, then that pore region is in an isolated state.

[0041] Step S1217: Record the connection method and corresponding connection channel parameters of all pores at each time point, and arrange them in chronological order to form a time sequence of pore connection paths.

[0042] In the above scenario, for each pore image acquired at a given time, the connection methods (directly connected, indirectly connected, isolated) of all pore regions at that time, as well as the corresponding connection channel parameters (length and width), are recorded. These data from different times are arranged sequentially in chronological order to form a pore connectivity path time series, which reflects the dynamic changes in the connection methods and connection channel parameters between pores over time.

[0043] Step S1218: Add the same time stamp to the cross-sectional morphology time sequence and the pore connectivity path time sequence.

[0044] In the above scenario, both the cross-sectional morphology time series and the pore connectivity path time series originate from the pore image series, with each series corresponding to a specific acquisition time. The acquisition time of the pore image series is used as a time stamp and added to both the cross-sectional morphology time series and the pore connectivity path time series, maintaining a temporal correspondence between the two time series and facilitating subsequent temporal correlation analysis.

[0045] Step S122: Extract the spatial distribution time sequence and chemical transformation time sequence of pollutants in the filter layer from the pollutant adsorption and migration data. The spatial distribution time sequence records the position coordinate data of pollutants at different times, and the chemical transformation time sequence records the molecular structure data of pollutants at different times.

[0046] In the above scenario, the pollutant adsorption and migration data includes pollutant concentration data collected at different locations in the filter media layer at different times. The extraction of the spatial distribution time series begins by determining the spatial coordinate system of the filter media layer, with the bottom of the filter bed as the origin, the vertical direction upwards as the height, and the horizontal direction as the radial direction. Based on the installation position of the water quality sensors in the filter media layer, the spatial coordinates corresponding to each sensor are obtained. The pollutant concentration data collected by each sensor at different times are associated with their corresponding spatial coordinates and arranged in chronological order to form a spatial distribution time series. This spatial distribution time series reflects the change in the spatial position of pollutants in the filter media layer over time. The extraction of the pollutant chemical form transformation time series involves analyzing the pollutant molecular structure detection data collected by the water quality sensors to extract characteristic parameters characterizing the molecular structure, such as functional group types, chemical bond types, and molecular configurations. These characteristic parameters at different times are then arranged in chronological order to form a pollutant chemical form transformation time series.

[0047] Step S123: Extract the filtration rate change time sequence and the backwashing effect time sequence from the filter operation feedback data. The filtration rate change time sequence records the water flow velocity data at different times, and the backwashing effect time sequence records the filter media state recovery data after backwashing.

[0048] In the above scenario, the filtration rate data in the filter operation feedback data is collected by a filtration rate sensor, recording the speed at which water flows through the filter media layer at different times. The extraction of the filtration rate change time series involves arranging the filtration rate data from different times in chronological order to form a sequence showing the filtration rate changing over time. The extraction of the backwashing effect time series involves collecting filter media state data at different times after the backwashing operation, such as the porosity of the filter media layer, pressure loss, and uniformity of filter media particles. These data reflect the recovery status of the filter media after backwashing. The above data are then arranged in chronological order after the end of backwashing to form the backwashing effect time series.

[0049] Step S124: Unify the time base of the cross-sectional morphology time sequence, the pore connectivity path time sequence, the spatial distribution time sequence, the chemical morphology transformation time sequence, the filtration rate change time sequence, and the backwashing effect time sequence to the same time unit. Using the unified time base as the axis, compare the abrupt changes in pore geometry in the cross-sectional morphology time sequence segment by segment and mark them as filter media pore morphology transformation nodes.

[0050] In the above scenario, the time reference is uniformly adopted in hours as the time unit. The time markers in each time series are converted to ensure that the time unit is consistent across all series. Using the unified time axis as the reference, the cross-sectional morphology time series is divided into multiple time periods. Trend analysis is performed on the pore geometry parameters (profile perimeter, inscribed circle diameter, shape factor) within each time period, and the rate of change of the parameters is calculated. When the rate of change of the parameters at a certain time point exceeds a preset abrupt change threshold, that time point is marked as a filter media pore morphology transformation node, indicating that a significant change in pore geometry has occurred at that moment.

[0051] Step S125: Compare the abrupt change positions of the connection mode in the time sequence of the pore connectivity path and mark them as filter media pore connectivity transition nodes; compare the abrupt change positions of the pollutant location coordinates in the time sequence of the spatial distribution and mark them as pollutant migration transition nodes; compare the abrupt change positions of the molecular structure in the time sequence of the chemical morphology transformation and mark them as pollutant morphology transformation nodes.

[0052] In the above scenarios, for the time series of pore connectivity paths, the changes in the connection modes (direct connection, indirect connection, and isolated state) between pores at different times are analyzed. When the connection mode at a certain time point changes significantly compared to the previous time point, such as from direct connection to indirect connection, or from isolated state to connected state, this time point is marked as a pore connectivity transition node in the filter media. For the spatial distribution time series, the changing trend of pollutant location coordinates is analyzed, and the movement speed of the location coordinates in the horizontal and vertical directions is calculated. When the movement speed at a certain time point exceeds a preset migration mutation threshold, this time point is marked as a pollutant migration transition node. For the chemical form transformation time series, the changes in molecular structure characteristic parameters are analyzed. When the characteristic parameters at a certain time point change significantly, such as the disappearance or appearance of functional groups, or a change in chemical bond type, this time point is marked as a pollutant form transformation node.

[0053] Step S126: Compare the abrupt changes in water flow velocity in the filtration rate change time sequence and mark them as filtration rate adjustment transition nodes; compare the starting position of filter media state recovery in the backwashing effect time sequence and mark it as backwashing operation transition nodes.

[0054] In the above scenario, for the time series of filtration rate changes, the variation of water flow velocity over time is analyzed. The difference in water flow velocity between adjacent time points is calculated. When the difference exceeds a preset filtration rate change threshold, this time point is marked as a filtration rate adjustment transition node, indicating that the filtration rate has been significantly adjusted at that moment. For the time series of backwashing effect, the start time of the backwashing operation is the starting position of the filter media state recovery. The time point corresponding to this starting position is marked as the backwashing operation transition node. This backwashing operation transition node marks the start of the backwashing process, and the filter media state enters the recovery phase.

[0055] Step S127: Calculate the relative rate of change or standardized change of the data sequence before and after each transition node in its own dimension, as the transition magnitude of the transition node in its own data dimension.

[0056] In the above scenario, for each transition node, the values ​​of the corresponding data sequence within a certain time window before and after the node are obtained. For example, for the filter media pore morphology transition node, the average value of the pore geometry parameters within the time window before the node and the average value of the pore geometry parameters within the time window after the node are obtained. The relative rate of change or standardized change of the values ​​before and after the node in their own dimensions is calculated, which is the transition amplitude of the filter media pore morphology transition node. Using the same method, the relative rate of change or standardized change of the corresponding data sequence values ​​at the filter media pore connection transition node, contaminant migration transition node, contaminant morphology transition node, filtration rate adjustment transition node, and backwashing operation transition node in their own dimensions is calculated, which is taken as the transition amplitude of each transition node.

[0057] Step S128: Arrange all morphological transformation nodes in chronological order to form a sequence of transformation nodes on a unified time axis, and associate the transformation magnitude of each transformation node.

[0058] In the above scenario, all identified morphological transformation nodes (including filter media pore morphology transformation nodes, filter media pore connectivity transformation nodes, contaminant migration transformation nodes, contaminant morphology transformation nodes, filtration rate adjustment transformation nodes, and backwashing operation transformation nodes) are collected. The timestamps of each node are extracted, and the nodes are sorted according to the order of their timestamps to form a sequence of transformation nodes on a unified timeline. In this sequence, each transformation node is associated with its corresponding transformation magnitude to facilitate subsequent analysis of the correlation and impact between different transformation nodes.

[0059] Step S130: Construct a dynamic correlation network based on the morphological transformation node and the transformation amplitude. The dynamic correlation network connects the interaction relationships of filter material pore morphology change, pollutant migration path change and operating parameter adjustment with time flow as the axis.

[0060] In the aforementioned scenario of a municipal waterworks volcanic rock filter deep water purification system, the construction of a dynamic interconnected network uses time flow as its core thread, organically integrating morphological transformation nodes and transformation magnitudes to intuitively demonstrate the dynamic interaction between filter media, pollutants, and operating parameters. Through the network structure, it is possible to clearly observe how changes in filter media pore morphology affect pollutant migration paths at different points in time, how adjustments to operating parameters affect filter media pore morphology and pollutant migration, and how changes in pollutant migration, in turn, prompt adjustments to operating parameters.

[0061] Step S131: Construct a network framework with time flow as the vertical axis. The vertical dimension of the network framework is the time process, and the horizontal dimensions are the filter media feature dimension, the pollutant feature dimension, and the operation feature dimension.

[0062] In the above scenario, the network framework adopts a two-dimensional coordinate system. The vertical axis represents the time process, with the order of time from bottom to top corresponding to the sequence of events. The time interval is determined based on the distribution of the transition nodes to ensure that all transition nodes can be clearly displayed. The horizontal dimension is divided into three parallel feature dimensions: filter media feature dimension, contaminant feature dimension, and operational feature dimension. Each feature dimension is used to display the corresponding type of transition node and related data.

[0063] Step S132: Embed the morphological transformation node into the corresponding horizontal dimension position of the network framework according to its feature dimension. Different feature dimension transformation nodes at the same time point are connected by horizontal connection lines.

[0064] In the above scenario, based on the type of morphological transformation node, it is embedded into the corresponding horizontal feature dimension. For example, filter media pore morphological transformation nodes and filter media pore connectivity transformation nodes are embedded into the filter media feature dimension; contaminant migration transformation nodes and contaminant morphological transformation nodes are embedded into the contaminant feature dimension; and filtration rate adjustment transformation nodes and backwashing operation transformation nodes are embedded into the operational feature dimension. For transformation nodes with different feature dimensions that occur at the same time point, they are connected by horizontal connecting lines in the network framework, indicating the temporal synchronicity of these transformation nodes.

[0065] Step S133: For nodes with different feature dimensions that change at the same time point, determine their change direction (increase / decrease) and record their standardized change magnitude level. Adjust the width of the horizontal connecting line according to the correlation of the change direction and the matching degree of the magnitude level. The higher the correlation and the more matched the magnitude, the wider the line width.

[0066] In the above scenario, for different feature dimension transformation nodes connected by horizontal connecting lines at the same time point, the ratio of their transformation amplitudes is calculated. For example, the ratio of the transformation amplitude of the filter media feature dimension transformation node to that of the pollutant feature dimension transformation node, or the ratio of the transformation amplitude of the running feature dimension transformation node to that of the filter media feature dimension transformation node, etc. Based on the magnitude of the ratio, the line width of the horizontal connecting line is set; the larger the ratio, the wider the line width, to visually reflect the relative influence of different feature dimension transformation nodes at the same time point.

[0067] Step S134: In the network framework, establish a longitudinal connection line between the filter material feature dimension change nodes at adjacent time points, and mark the direction of pore morphology change corresponding to the connection line.

[0068] In the above scenario, for two adjacent time points representing changes in filter media characteristics, a vertical connecting line is used to connect them. The direction of change in pore morphology parameters is marked based on the change from the previous node to the next. If the pore morphology parameters increase (e.g., increased pore volume, enhanced connectivity), it is marked as a positive change direction; if the pore morphology parameters decrease (e.g., decreased pore volume, weakened connectivity), it is marked as a negative change direction.

[0069] Step S135: Establish vertical connection lines between pollutant feature dimension change nodes at adjacent time points, and mark the direction of migration path change corresponding to the connection lines.

[0070] In the above scenario, for two pollutant characteristic dimension transition nodes at adjacent time points, they are connected by a vertical connecting line. Based on the change in the pollutant migration path from the previous node to the next node, the direction of the migration path change is marked. If the pollutant migrates towards the depth of the filter layer, it is marked as a downward migration direction; if the pollutant migrates towards the surface of the filter layer, it is marked as an upward migration direction; if the pollutant diffuses horizontally, it is marked as a horizontal migration direction.

[0071] Step S136: Establish vertical connection lines between nodes of change in the running feature dimension at adjacent time points, and mark the parameter adjustment direction corresponding to the connection lines.

[0072] In the above scenario, for two adjacent time points in the operational feature dimension, the transition nodes of the operational feature dimension are connected by a vertical connecting line. The direction of parameter adjustment is marked according to the change in operational parameters from the previous node to the next node. For the filtration rate adjustment parameter, if the filtration rate increases, it is marked as the direction of increase; if the filtration rate decreases, it is marked as the direction of decrease. For the backwash cycle parameter, if the backwash cycle is extended, it is marked as the direction of cycle extension; if the backwash cycle is shortened, it is marked as the direction of cycle shortening.

[0073] Step S137: For transition nodes of the same feature dimension at adjacent time points, calculate the standardized difference in the level of transition amplitude, and adjust the color depth of the vertical connecting line according to the magnitude of the difference in the level of transition amplitude. The greater the difference in level, the darker the color.

[0074] In the above scenario, for transition nodes at adjacent time points within the same feature dimension, the difference in their transition magnitude is calculated. For example, the difference between the transition magnitude level of the preceding and following transition nodes in the filter media feature dimension. Based on the magnitude of the difference, the color depth of the vertical connecting line is adjusted; the larger the difference, the darker the color, to visually reflect the change in the transition magnitude of the same feature dimension transition nodes at adjacent time points.

[0075] Step S138: Identify the time interval in the network framework where both filter media feature dimension and pollutant feature dimension transition nodes appear simultaneously, and add bidirectional arrows within this time interval to indicate the interaction between the two.

[0076] In the above scenario, the timeline of the network framework is traversed to find time intervals where both filter media feature dimension transition nodes and pollutant feature dimension transition nodes exist simultaneously. Within these time intervals, bidirectional arrows are added between the filter media feature dimensions and the pollutant feature dimensions, with the arrows pointing to the two feature dimensions respectively. This indicates that there is an interaction between changes in filter media pore morphology and changes in pollutant migration paths; that is, changes in filter media pore morphology affect pollutant migration paths, and conversely, pollutant migration also affects filter media pore morphology.

[0077] Step S1381: Traverse the time axis of the network framework and check segment by segment whether there are cases where filter material feature dimension transformation nodes and pollutant feature dimension transformation nodes appear simultaneously in the same time interval.

[0078] In the above scenario, the timeline is divided into multiple consecutive time intervals according to the time axis sequence. The length of each time interval is determined based on the distribution density of the transition nodes. For each time interval, it is checked whether it simultaneously contains transition nodes for both filter media characteristic dimensions and contaminant characteristic dimensions. If so, the start and end times of that time interval are recorded.

[0079] Step S1382: For each time interval in which two types of transition nodes occur simultaneously, determine the start and end times of the time interval and mark it as the interaction interval.

[0080] In the above scenario, for the time interval that simultaneously contains both filter material feature dimension transition nodes and pollutant feature dimension transition nodes, the start and end times of this interval are determined as the boundary of the interaction interval, and this interval is marked so that interaction relationship identifiers can be added later.

[0081] Step S1383: Extract the change magnitude of the filter media feature dimension change nodes and the change magnitude of the pollutant feature dimension change nodes within the interaction interval.

[0082] In the above scenario, within the marked interaction range, find all filter media feature dimension transformation nodes and pollutant feature dimension transformation nodes, extract the transformation amplitude data of each node, and calculate the sum of the filter media feature dimension transformation amplitude and the sum of the pollutant feature dimension transformation amplitude, or calculate the average transformation amplitude of the two.

[0083] Step S1384: Calculate the ratio of the characteristic change amplitude of the filter media to the characteristic change amplitude of the pollutants. This ratio is used to characterize the intensity ratio of their interaction.

[0084] In the above scenario, the ratio of the change amplitude of filter media feature dimensions to the change amplitude of pollutant feature dimensions is obtained by dividing the sum (or average change amplitude) of the change amplitude of filter media feature dimensions within the interaction interval by the sum (or average change amplitude) of the change amplitude of pollutant feature dimensions. This ratio reflects the intensity ratio of the interaction between filter media features and pollutant features within the interaction interval.

[0085] Step S1385: Determine the size of the bidirectional arrowheads according to the intensity ratio. The closer the ratio is to 1, the more consistent the arrow sizes.

[0086] In the above scenario, the two arrows of the bidirectional arrow correspond to the interaction directions from the filter media feature dimension to the pollutant feature dimension and from the pollutant feature dimension to the filter media feature dimension, respectively. The size of the two arrows is adjusted according to the intensity ratio. If the intensity ratio is close to 1, it indicates that the interaction strength is comparable, and the size of the two arrows is set to be the same; if the intensity ratio is greater than 1, it indicates that the filter media feature has a stronger effect on the pollutant feature, and the arrow in the corresponding direction is set to be larger; if the intensity ratio is less than 1, it indicates that the pollutant feature has a stronger effect on the filter media feature, and the arrow in the corresponding direction is set to be larger.

[0087] Step S1386: Analyze the direction of change of filter media characteristic dimension data and the direction of change of pollutant characteristic dimension data within the interaction interval to determine the relationship between their interaction directions; mark the interaction direction relationship on the bidirectional arrow, including promoting relationship and inhibiting relationship. Promoting relationship means that the change of one side aggravates the change of the other side, and inhibiting relationship means that the change of one side slows down the change of the other side.

[0088] In the above scenario, we analyze the direction of change in filter media characteristic dimension data (e.g., porosity increase or decrease) and pollutant characteristic dimension data (e.g., migration speed increase or decrease) within the interaction interval. If the filter media characteristic dimension data changes in a certain direction, and the pollutant characteristic dimension data also changes in the same direction with an increased degree of change, then the two have a promoting relationship. If the filter media characteristic dimension data changes in a certain direction, and the pollutant characteristic dimension data changes in the opposite direction or with a decreased degree of change, then the two have an inhibitory relationship. The promoting or inhibitory relationship is clearly labeled on the bidirectional arrows to clarify the directional relationship between the two.

[0089] Step S1387: Calculate the synchronization rate of changes in the two types of feature dimensions within the interaction interval. The higher the synchronization rate, the thicker the double-headed arrow.

[0090] In the above scenario, the synchronization rate of changes in the two types of feature dimensions within the interaction interval is determined by calculating the temporal overlap and trend consistency between the changes in filter media feature dimension data and pollutant feature dimension data. The higher the temporal overlap and the more consistent the trend, the higher the synchronization rate. Based on the synchronization rate, the thickness of the bidirectional arrow is adjusted; the higher the synchronization rate, the thicker the line, reflecting the degree of synchronization in their interaction.

[0091] Step S1388: Add the labeled bidirectional arrows to the interaction area of ​​the network framework and connect the corresponding filter media feature dimension transformation node and pollutant feature dimension transformation node.

[0092] In the above scenario, based on the location and range of the interaction zone, bidirectional arrows labeled with the direction of action, arrow size, and line thickness are added to the network framework. The two ends of the arrows are connected to the corresponding filter material feature dimension transformation node and pollutant feature dimension transformation node within the interaction zone, respectively, to intuitively show the interaction relationship between the two within the zone.

[0093] Step S1389: Number all added bidirectional arrows and establish a correspondence table between arrows and their interaction ranges.

[0094] In the above scenario, each bidirectional arrow added to the network framework is assigned a unique number, and then a corresponding table is established. The table records information such as the arrow number, the start and end time of the interaction interval, and the direction of action, which facilitates the management and querying of the interaction relationships represented by the bidirectional arrows.

[0095] Step S139: Identify the time interval in the network framework after the occurrence of a change node in the operating feature dimension, and the occurrence of a change node in the filter media feature dimension or the pollutant feature dimension. Add a one-way arrow within this time interval to indicate the impact of the operating parameter adjustment on the filter media or the pollutant.

[0096] In the above scenario, the timeline of the network framework is traversed to find the time interval following the occurrence of the operational feature dimension transition node. Within this time interval, if a subsequent filter media feature dimension transition node or contaminant feature dimension transition node appears, it is considered that the adjustment of the operating parameters has affected the filter media pore morphology or contaminant migration. A one-way arrow is added between the operational feature dimension transition node and the subsequent filter media feature dimension transition node or contaminant feature dimension transition node, with the arrow pointing from the operational feature dimension to the filter media feature dimension or contaminant feature dimension, indicating the impact of the operating parameter adjustment on the filter media or contaminant.

[0097] Step S1310: For transition nodes of the same feature dimension at adjacent time points, calculate the difference in the level of the transition amplitude after standardization, and adjust the color depth of the vertical connection line according to the magnitude of the difference in the level of the transition amplitude. The greater the difference in the level, the darker the color.

[0098] In the above scenario, this step is the same as step S137, and will not be described again here.

[0099] Step S1311: Count the frequency of occurrence of various types of connection lines in the network framework, and mark the connection pattern with the highest frequency as the core association pattern.

[0100] In the above scenario, statistics are collected on various connection lines in the network framework, including horizontal connections, vertical connections, and bidirectional arrows. The frequency of occurrence of each connection pattern (such as horizontal connections where the ratio of filter media feature dimensions to pollutant feature dimensions is within a certain range, and vertical connections where the filter media feature dimensions are within a specific range of difference in range) is calculated. The connection pattern with the highest frequency is marked as the core association pattern, which reflects the most important feature association relationship in the system.

[0101] Step S140: Generate adaptive adjustment rules for filter media state based on the dynamic correlation network. The adaptive adjustment rules are used to describe the optimal matching mode between filter media pore morphology and operating parameters under different pollutant migration states.

[0102] In the aforementioned scenario of a municipal waterworks volcanic rock filter deep water purification system, the dynamic correlation network clearly demonstrates the relationships between filter media, pollutants, and operating parameters at different time points. Based on these relationships, the analysis examines how the pore morphology of the filter media changes with the adjustment of operating parameters under different pollutant migration states, and what combination of filter media pore morphology and operating parameters can achieve the best water purification effect, thereby generating adaptive adjustment rules for the filter media state.

[0103] Step S141: Extract all sub-network segments containing pollutant migration and transformation nodes from the dynamic association network. Each sub-network segment contains filter media feature dimension data and operational feature dimension data corresponding to the pollutant migration and transformation node.

[0104] In the above scenario, the dynamic correlation network is traversed to find all network segments containing pollutant migration and transformation nodes. For each pollutant migration and transformation node, relevant filter media characteristic data (such as pore morphology parameters and connection methods) and operational characteristic data (such as filtration rate adjustment parameters and backwashing cycle parameters) within a certain time range before and after it are extracted to form a sub-network segment. Each sub-network segment revolves around a pollutant migration and transformation node, reflecting the filter media state and operational status at that node.

[0105] Step S142: Divide the sub-network segment into multiple pollutant migration state categories according to the transformation magnitude of pollutant migration transition nodes, with each pollutant migration state category corresponding to a pollutant migration intensity range.

[0106] In the above scenario, statistical analysis is performed on the transition amplitudes of pollutant migration transition nodes in all extracted sub-network segments to determine the distribution range of the transition amplitudes. Based on the magnitude of the transition amplitude, the distribution range is divided into multiple continuous intervals, each interval corresponding to a pollutant migration state category, representing a range of pollutant migration intensity, such as low-intensity migration, medium-intensity migration, and high-intensity migration. Each sub-network segment is then assigned to the corresponding pollutant migration state category based on the interval to which the transition amplitude of its pollutant migration transition nodes belongs.

[0107] Step S143: For each pollutant migration state category, extract the filter media pore morphology data from all sub-network segments under that pollutant migration state category, including cross-sectional morphology parameters and pore connectivity path parameters.

[0108] In the above scenario, for each pollutant migration state category, filter media pore morphology data are collected from all sub-network segments within that category. Cross-sectional morphology parameters include the pore perimeter, inscribed circle diameter, and shape factor; pore connectivity path parameters include the length, width, and connection method (direct or indirect) of the connecting channels. These parameters are then aggregated to form the filter media pore morphology dataset corresponding to that pollutant migration state category.

[0109] Step S144: Analyze the correlation between filter media pore morphology data and water purification effect under the same pollutant migration state category, and select the filter media pore morphology parameter range that meets the preset standard for water purification effect as the optimal filter media pore morphology range under the pollutant migration state category.

[0110] In the above scenario, historical water purification performance data corresponding to each pollutant migration state category is retrieved. This data includes whether the effluent water quality meets standards (e.g., whether turbidity, COD, heavy metal ion concentration, etc., meet standards) and purification efficiency (e.g., pollutant removal rate). Correlation analysis is performed between the filter media pore morphology data and the water purification performance data to determine which filter media pore morphology parameters significantly affect the water purification performance. Then, based on preset water purification performance standards (e.g., effluent water quality meets standards and purification efficiency reaches a certain value), the range of values ​​for filter media pore morphology parameters that ensure the water purification performance meets these standards is selected, and this range is determined as the optimal filter media pore morphology range for that pollutant migration state category.

[0111] Step S1441: Retrieve historical water purification effect data corresponding to each pollutant migration status category. The historical water purification effect data includes the effluent water quality compliance status and purification efficiency data.

[0112] In the above scenario, data is retrieved from the historical database of the water purification system. For each pollutant migration state category, the water purification effect data within the corresponding time period is found based on the time stamp of the sub-network segment. The compliance of the effluent water quality is determined by comparing the differences between the effluent water quality indicators and national or design standards; the purification efficiency data is obtained by calculating the pollutant removal rate (e.g., (influent pollutant concentration - effluent pollutant concentration) / influent pollutant concentration).

[0113] Step S1442: Divide the historical water purification effect data into a qualified data group and a non-qualified data group according to a preset standard, wherein the preset standard is determined based on the effluent water quality requirements of the water purification system.

[0114] In the above scenario, the preset standards are formulated based on the design requirements of the water purification system for effluent water quality, including the threshold values ​​for various water quality indicators and the minimum requirements for purification efficiency. Data from historical water purification performance that meet all effluent water quality indicators and the minimum purification efficiency requirements are classified as compliant data groups; data that do not meet these conditions are classified as non-compliant data groups.

[0115] Step S1443: Extract the filter media pore morphology data corresponding to the qualified data group, including cross-sectional morphology parameters and pore connectivity path parameters.

[0116] In the above scenario, based on the time stamp of the compliant data group, filter media pore morphology data for the corresponding time period are extracted from the filter media pore morphology dataset, including cross-sectional morphology parameters (profile perimeter, inscribed circle diameter, shape factor) and pore connectivity path parameters (connection channel length, width, connection method).

[0117] Step S1444: Perform statistical analysis on the various cross-sectional morphological parameters of the qualified data group, calculate the numerical distribution characteristics (such as mean, median and standard deviation) of each type of parameter (such as profile perimeter, inscribed circle diameter, shape factor), and determine the qualified numerical range of each type of parameter.

[0118] In the above scenario, for each cross-sectional morphological parameter (such as profile perimeter) in the compliant data set, the numerical distribution characteristics of all data for that parameter are calculated, such as the mean, median, and standard deviation. The mean reflects the average level of the data, the median reflects the median level of the data, and the standard deviation reflects the dispersion of the data. Based on the mean and standard deviation, the concentrated distribution range of the cross-sectional morphological parameter is determined, which is usually the range of compliant values ​​covered by the mean plus or minus a certain multiple of the standard deviation. This range of compliant values ​​includes the cross-sectional morphological parameter values ​​of most of the compliant data set.

[0119] Step S1445: Perform statistical analysis on the various pore connectivity path parameters of the compliant data group, calculate the numerical distribution characteristics of each type of parameter, and determine the compliant numerical range of each type of parameter.

[0120] In the above scenario, the same method as in step S1444 is used to perform statistical analysis on the pore connectivity path parameters (such as the length and width of the connecting channel) of the compliant data group, calculate the mean, median and standard deviation, and determine the compliant value range for each pore connectivity path parameter.

[0121] Step S1446: Combine the acceptable value ranges of various cross-sectional morphological parameters and the acceptable value ranges of various pore connectivity path parameters in parallel to form an initial set of candidate filter media pore morphological parameter ranges.

[0122] In the above scenario, the concentrated distribution ranges of cross-sectional morphology parameters and pore connectivity path parameters are integrated, and the concentrated distribution range of each parameter is taken as the value interval of that parameter in the initial candidate filter media pore morphology range. For example, the concentrated distribution range of the profile perimeter is [L1, L2], the concentrated distribution range of the inscribed circle diameter is [D1, D2], the concentrated distribution range of the connecting channel length is [C1, C2], etc. Combining the above intervals forms the initial candidate filter media pore morphology range.

[0123] Step S1447: Extract the filter media pore morphology data corresponding to the non-compliant data group, analyze the difference between it and the initial candidate filter media pore morphology range, and determine the parameter value range that needs to be excluded.

[0124] In the above scenario, based on the time stamp of the substandard data group, filter media pore morphology data for the corresponding time period is extracted. This data is then compared with the initial candidate filter media pore morphology range to identify the value range of filter media pore morphology parameters in the substandard data group. If certain parameter value ranges appear frequently in the substandard data group but less frequently in the compliant data group, these parameter value ranges are identified as those to be excluded.

[0125] Step S1448: Remove the parameter value range that needs to be excluded from the initial candidate filter media pore morphology range to form the optimized candidate filter media pore morphology range.

[0126] In the above scenario, the parameter value ranges that need to be excluded are removed from the initial candidate filter media pore morphology range. For each filter media pore morphology parameter, its concentrated distribution range is subtracted from the parameter value range that needs to be excluded to obtain the optimized parameter value range. The optimized value ranges of all parameters are combined to form the optimized candidate filter media pore morphology range.

[0127] Step S1449: Verify the optimized candidate filter media pore morphology range and calculate the historical water purification effect compliance rate corresponding to the filter media pore morphology data within the candidate filter media pore morphology range.

[0128] In the above scenario, the historical water purification effect compliance rate is obtained by extracting all time periods from the historical database where the filter media pore morphology data falls within the optimized candidate filter media pore morphology range, and then calculating the ratio of the number of times the water purification effect meets the standard to the total number of times during these time periods.

[0129] Step S14410: If the historical water purification effect compliance rate reaches the preset threshold, the optimized candidate filter media pore morphology range is determined as the optimal filter media pore morphology range under the pollutant migration state category; if the historical water purification effect compliance rate does not reach the preset threshold, the statistical analysis method of the parameters is adjusted, and the concentrated distribution range is re-determined until the historical water purification effect compliance rate meets the requirements.

[0130] In the above scenario, the preset threshold is determined based on the reliability requirements of the water purification system. If the calculated historical water purification performance meets or exceeds the preset threshold, the optimized candidate filter media pore morphology range is considered to be able to stably guarantee the water purification effect, and it is determined as the optimal filter media pore morphology range under the pollutant migration state category. If the preset threshold is not met, the statistical analysis method of the parameters is adjusted, such as changing the multiple of the mean plus or minus the standard deviation when determining the concentrated distribution range, or using other statistical measures (such as the quartile range) to determine the concentrated distribution range, and the operations of steps S1444 to S1449 are repeated until the historical water purification performance meets the preset threshold requirement.

[0131] Step S145: Extract the operating parameter data from all sub-network segments under each pollutant migration state category, including filtration rate adjustment parameters and backwash cycle parameters.

[0132] In the above scenario, for each pollutant migration state category, operational parameter data from all sub-network segments within that category are collected. Filtration rate adjustment parameters include filtration rate values ​​at different times and the rate of filtration rate adjustment; backwash cycle parameters include the backwash time interval and backwash duration. These parameters are then aggregated to form the operational parameter dataset corresponding to that pollutant migration state category.

[0133] Step S146: Analyze the correlation between operating parameter data and filter media pore morphology parameters under the same pollutant migration state category, and determine the combination of operating parameters that can maintain the filter media pore morphology within the optimal range of filter media pore morphology, as the candidate combination of operating parameters under the pollutant migration state category.

[0134] In the above scenario, multiple regression analysis or machine learning methods are used to analyze the correlation between operating parameter data and filter media pore morphology parameters. With the goal of maintaining the filter media pore morphology parameters within the optimal range, combinations of operating parameter values ​​that can achieve this goal are sought. For example, the changing trends of filter media pore morphology parameters under different filtration rates and backwashing cycles are analyzed to screen out combinations of filtration rates and backwashing cycles that can stabilize the filter media pore morphology parameters within the optimal range. These combinations are then identified as candidate operating parameter combinations for this pollutant migration state category.

[0135] Step S147: Substitute the candidate operating parameter combinations for each pollutant migration state category into the dynamic correlation network for simulation verification, and observe whether the filter media pore morphology remains within the optimal filter media pore morphology range during the simulation process.

[0136] In the above scenario, a simulation model is constructed using a dynamic correlation network, and candidate combinations of operating parameters for each pollutant migration state category are input into the simulation model. Based on the feature correlations recorded in the dynamic correlation network, the simulation model simulates the evolution of the filter media pore morphology under the influence of these operating parameter combinations. During the simulation, the pore morphology parameters are continuously monitored to ensure they remain within the optimal range. If the pore morphology parameters remain within the optimal range throughout the entire simulation period, the candidate operating parameter combination is considered effective.

[0137] Step S148: Adjust the numerical range of candidate operating parameter combinations based on the verification results, eliminate parameter values ​​that cause the filter media pore morphology to deviate from the optimal filter media pore morphology range, and form the optimal operating parameter combination for the pollutant migration state category.

[0138] In the above scenario, for candidate operating parameter combinations where the filter media pore morphology deviates from the optimal range during simulation verification, the reasons for the deviation are analyzed to determine which operating parameter values ​​caused the deviation. Then, the numerical ranges of these operating parameters are adjusted, and the parameter values ​​causing the deviation are eliminated. For the adjusted candidate operating parameter combinations, simulation verification is performed again, and this process is repeated until the candidate operating parameter combinations can consistently maintain the filter media pore morphology within the optimal range. The candidate operating parameter combinations at this point are considered the optimal operating parameter combinations for that pollutant migration state category.

[0139] Step S149: Bind the pollutant migration state category, the optimal filter media pore morphology range, and the optimal combination of operating parameters to form basic adjustment rules.

[0140] In the above scenario, each pollutant migration state category is associated with and bound to the corresponding optimal filter media pore morphology range and optimal operating parameter combination to form a basic adjustment rule. The basic adjustment rule can take the form of "when the pollutant migration state is a certain category, the filter media pore morphology should be maintained within a certain optimal range, and the corresponding operating parameter combination is a certain combination".

[0141] Step S1410: Analyze the transition intervals between different pollutant migration state categories, and supplement each transition interval with a corresponding transition adjustment rule to ensure that the adjustment rules of adjacent pollutant migration state categories can be smoothly connected.

[0142] In the above scenario, there are transitional intervals between different pollutant migration states, representing the process of pollutant migration states changing from one category to another. This study analyzes the characteristics of pollutant migration state changes and the evolution of filter media pore morphology within these transitional intervals. Based on the fundamental adjustment rules for adjacent pollutant migration state categories, transitional adjustment rules are formulated. In these rules, the combination of operating parameters should gradually transition from the optimal combination for the current category to the optimal combination for the adjacent category. This avoids abrupt changes in operating parameters that could lead to drastic changes in filter media pore morphology, ensuring a smooth transition between adjustment rules for adjacent categories.

[0143] Step S1411: Integrate all basic adjustment rules and transitional adjustment rules to form filter media state adaptive adjustment rules. Each filter media state adaptive adjustment rule describes the optimal matching mode between filter media pore morphology and operating parameters under a specific pollutant migration state.

[0144] In the above scenario, the basic adjustment rules for all pollutant migration state categories and the transition adjustment rules for the transition range are integrated and arranged according to the order of pollutant migration state changes, forming a complete set of adaptive adjustment rules for filter media state. Each rule in this set of adaptive adjustment rules for filter media state describes the optimal range that the filter media pore morphology should maintain in a specific pollutant migration state (including stable and transitional states) and the matching combination of operating parameters, thus achieving optimal matching between filter media pore morphology and operating parameters under different pollutant migration states.

[0145] Step S150: Generate dynamic correction instructions for the operating parameters of the volcanic rock water purification system according to the adaptive adjustment rules for the filter media state. The dynamic correction instructions for the operating parameters include real-time adjustment parameters for filtration rate and backwashing cycle adjustment parameters.

[0146] In the aforementioned scenario of a municipal waterworks volcanic rock filter deep water purification system, the current pollutant migration status is monitored in real time and matched with the pollutant migration status category in the filter media state adaptive adjustment rules to determine the corresponding optimal combination of operating parameters. Based on the current filter media pore morphology and the deviation of operating parameters from the optimal values, the operating parameters are adjusted, generating dynamic correction instructions that include real-time filtration rate adjustment parameters and backwash cycle adjustment parameters. These instructions are then sent to the water purification system's control system to achieve automatic adjustment of operating parameters.

[0147] Step S151: Collect real-time pollutant migration data of the volcanic rock water purification system and extract the spatial distribution parameters and chemical speciation parameters of the pollutants.

[0148] In the above scenario, water quality sensors collect real-time data on pollutant migration in the volcanic rock water purification system. This includes spatial distribution parameters such as the spatial location and concentration gradient of pollutants within the filter media layer, as well as chemical morphology parameters such as the molecular structure, functional group types, and chemical bond types of the pollutants. The data collected by the sensors is transmitted in real-time to the data processing unit, which analyzes and extracts the data to obtain the current spatial distribution and chemical morphology parameters of the pollutants.

[0149] Step S152: Match the spatial distribution parameters of the pollutants and the chemical morphology parameters with the pollutant migration state category in the filter media state adaptive adjustment rule to determine the current pollutant migration state category.

[0150] In the above scenario, the real-time extracted spatial distribution parameters and chemical speciation parameters of pollutants are compared with the feature parameters of each pollutant migration state category defined in the adaptive adjustment rule of filter media state. A pattern recognition method is used to calculate the similarity between the current parameter and the feature parameters of each category, and the category with the highest similarity is determined as the current pollutant migration state category. If the current parameter is in the transition interval between two adjacent categories, it is determined to be a transition state.

[0151] Step S153: Based on the matching results, retrieve the optimal filter media pore morphology range and optimal operating parameter combination corresponding to the pollutant migration state category.

[0152] In the above scenario, based on the determined current pollutant migration state category, the corresponding rule entry is searched from the set of adaptive adjustment rules for filter media state, and the optimal filter media pore morphology range and optimal operating parameter combination (including filtration rate adjustment parameters and backwashing cycle parameters) recorded in the rule entry are retrieved.

[0153] Step S154: Collect the current filter media pore morphology data of the volcanic rock water purification system in real time, compare the various parameters contained therein with the corresponding standard value ranges of various parameters in the optimal filter media pore morphology parameter range set, and calculate the deviation parameters between the two.

[0154] In the above scenario, a microscopic imaging device is used to acquire real-time images of the filter media pores. After image segmentation and feature extraction, the current filter media pore morphology data (cross-sectional morphology parameters and pore connectivity path parameters) is obtained. The current filter media pore morphology data is compared with the retrieved optimal filter media pore morphology range. For each pore morphology parameter, the difference between the current parameter value and the center or boundary value of the optimal range is calculated. These differences are combined into a deviation parameter, which is used to characterize the degree of deviation between the current filter media pore morphology and the optimal range.

[0155] Step S155: Based on the deviation of various parameters, comprehensively judge the degree of deviation of the overall state of the filter media pores, and adjust the filtration rate adjustment parameter and backwashing cycle parameter in the optimal operating parameter combination according to this comprehensive deviation. The greater the comprehensive deviation, the greater the adjustment range.

[0156] In the above scenario, a mapping relationship is established between the deviation parameter and the adjustment range of the operating parameters. When the deviation parameter is zero, it indicates that the current filter media pore morphology is within the optimal range, and no adjustment of the operating parameters is required. When the deviation parameter is not zero, the adjustment range is determined based on the magnitude of the deviation parameter; the larger the deviation parameter, the larger the adjustment range. For the filtration rate adjustment parameter, if the current filter media pore morphology parameter is less than the lower limit of the optimal range (e.g., the pores are too small), the filtration rate is increased; if it is greater than the upper limit of the optimal range (e.g., the pores are too large), the filtration rate is decreased. For the backwashing cycle parameter, if the current filter media pore morphology parameter is less than the lower limit of the optimal range, the backwashing cycle is shortened; if it is greater than the upper limit of the optimal range, the backwashing cycle is extended.

[0157] For example, step S1551: Determine the numerical range of the deviation parameter, divide the numerical range into multiple deviation levels, and each deviation level corresponds to a different adjustment range coefficient.

[0158] In the above scenario, based on historical data and experimental results, the possible numerical range of the deviation parameter is determined. This range is then divided into multiple consecutive intervals, each corresponding to a deviation level, such as slight deviation, moderate deviation, or severe deviation. An adjustment coefficient is assigned to each deviation level; the higher the deviation level, the larger the adjustment coefficient, indicating that a greater adjustment to the operating parameters is required.

[0159] Step S1552: Assign a corresponding filtration rate adjustment coefficient and backwash cycle adjustment coefficient to each deviation level. The higher the deviation level, the larger the adjustment coefficient.

[0160] In the above scenario, based on the degree of influence of different deviation levels on the pore morphology of the filter media, a filtration rate adjustment coefficient and a backwashing cycle adjustment coefficient are assigned to each deviation level. For example, a minor deviation level corresponds to a smaller adjustment coefficient, while a severe deviation level corresponds to a larger adjustment coefficient.

[0161] Step S1553: Extract the baseline filtration rate parameter and baseline backwashing cycle parameter from the optimal operating parameter combination.

[0162] In the above scenario, the filtration rate adjustment parameter and backwash cycle parameter in the optimal operating parameter combination are used as benchmark values, referred to as the benchmark filtration rate parameter and the benchmark backwash cycle parameter, respectively. These two benchmark parameters are extracted from the optimal operating parameter combination.

[0163] Step S1554: Multiply the reference filtration rate parameter by the filtration rate adjustment range coefficient corresponding to the current deviation level to obtain the adjustment amount of the filtration rate parameter.

[0164] In the above scenario, the deviation level is determined based on the current deviation parameter, and the corresponding filtration speed adjustment coefficient is obtained. The baseline filtration speed parameter is then multiplied by the filtration speed adjustment coefficient to obtain the adjustment amount of the filtration speed parameter.

[0165] Step S1555: Determine the adjustment direction of the filtration rate parameter based on the deviation direction between the filter media pore morphology data and the optimal filter media pore morphology range. If the current pore morphology parameter is less than the lower limit of the optimal filter media pore morphology range, increase the filtration rate; if it is greater than the upper limit of the optimal filter media pore morphology range, decrease the filtration rate.

[0166] In the above scenario, compare the current filter media pore morphology parameters with the lower and upper limits of the optimal filter media pore morphology range. If the current parameters are lower than the lower limit, it indicates that the filter media pores may be at risk of clogging, and the filtration rate needs to be increased to enhance the flushing effect of the water flow on the pores. If the current parameters are higher than the upper limit, it indicates that the filter media pores may be too loose, and the filtration rate needs to be reduced to prolong the residence time of pollutants in the filter media layer and improve the adsorption effect. Determine the direction for adjusting the filtration rate parameters based on the comparison results.

[0167] Step S1556: Multiply the reference backwash cycle parameter by the backwash cycle adjustment range coefficient corresponding to the current deviation level to obtain the adjustment amount of the backwash cycle parameter.

[0168] In the above scenario, a method similar to step S1554 is used to multiply the baseline backwash cycle parameter by the backwash cycle adjustment range coefficient corresponding to the current deviation level to obtain the adjustment amount of the backwash cycle parameter.

[0169] Step S1557: Determine the adjustment direction of the backwashing cycle based on the deviation direction. If the current pore morphology parameter is less than the lower limit of the optimal filter media pore morphology range, shorten the backwashing cycle. If it is greater than the upper limit of the optimal filter media pore morphology range, extend the backwashing cycle.

[0170] In the above scenario, if the current filter media pore morphology parameters are less than the lower limit of the optimal range, it indicates a high degree of pore blockage, requiring a shorter backwashing cycle and more frequent backwashing to restore the pore morphology. If the current parameters are greater than the upper limit of the optimal range, it indicates that the filter media pores are relatively unobstructed, allowing for a more appropriate extension of the backwashing cycle to reduce disturbance to the filter media structure. The adjustment direction of the backwashing cycle is determined based on the direction of deviation.

[0171] Step S1558: Calculate the adjusted filtration rate parameter value, which is the result of calculating the baseline filtration rate parameter and the filtration rate adjustment amount in the adjustment direction.

[0172] In the above scenario, based on the determined filtration rate adjustment direction, if it is necessary to increase the filtration rate, the adjusted filtration rate parameter value is the base filtration rate parameter plus the filtration rate adjustment amount; if it is necessary to decrease the filtration rate, the adjusted filtration rate parameter value is the base filtration rate parameter minus the filtration rate adjustment amount.

[0173] Step S1559: Calculate the adjusted backwash cycle parameter value. This backwash cycle parameter value is the result of calculating the baseline backwash cycle parameter and the backwash cycle adjustment amount according to the adjustment direction.

[0174] In the above scenario, the adjustment direction is based on the determined backwashing cycle. If it is necessary to shorten the backwashing cycle, the adjusted backwashing cycle parameter value is the base backwashing cycle parameter minus the backwashing cycle adjustment amount; if it is necessary to extend the backwashing cycle, the adjusted backwashing cycle parameter value is the base backwashing cycle parameter plus the backwashing cycle adjustment amount.

[0175] Step S15510: Check whether the adjusted parameter value exceeds the maximum and minimum allowable range. If it does, limit it to the allowable range.

[0176] In the above scenario, the filtration rate and backwash cycle parameters of the water purification system have designed maximum and minimum allowable ranges to ensure the safe and stable operation of the system. The adjusted filtration rate parameter value and backwash cycle parameter value are compared with the corresponding maximum and minimum ranges, respectively. If they exceed the range, the parameter value is limited to the maximum or minimum value within the allowable range.

[0177] Step S15511: Record the parameter values ​​before and after adjustment, as well as the corresponding deviation parameters and adjustment range coefficients, to form a parameter adjustment record.

[0178] In the above scenario, the baseline filtration rate parameters and baseline backwash cycle parameters before adjustment, the adjusted filtration rate parameter values ​​and backwash cycle parameter values, the corresponding deviation parameters and adjustment range coefficients, etc., are recorded to form a parameter adjustment record, which is stored in the system's log database to facilitate subsequent system operation analysis and parameter optimization.

[0179] Step S156: Extract historical correlation segments from the dynamic correlation network that are similar to the current pollutant migration state category and filter media pore morphology deviation parameters, and obtain the operational parameter adjustment effect data in the historical correlation segments.

[0180] In the above scenario, a search is performed in the dynamic correlation network to find historical correlation segments that are the same or similar to the current pollutant migration state category and have similar filter media pore morphology deviation parameters. For the found historical correlation segments, data on the recovery of filter media pore morphology and changes in water purification effect after the adjustment of operating parameters are extracted.

[0181] Step S157: Based on the operating parameter adjustment effect data, further optimize the adjusted filtration rate adjustment parameters and backwash cycle parameters, and convert the optimized filtration rate adjustment parameters into specific filtration rate change rate and change duration to form real-time filtration rate adjustment parameters.

[0182] In the above scenario, the operational parameter adjustment effect data of historical related segments are analyzed. If similar adjustments in the past have resulted in good restoration of filter media pore morphology and stable water purification effect, the current adjustment parameters are maintained. If the historical adjustment effect is poor, the adjusted filtration rate adjustment parameters and backwash cycle parameters are fine-tuned based on the feedback from the effect data. The optimized filtration rate adjustment parameters are converted into the filtration rate change rate per unit time (such as the filtration rate increase or decrease per minute) and the change time required to complete the adjustment. These two parameters together constitute the real-time filtration rate adjustment parameters.

[0183] Step S158: Convert the optimized backwash cycle parameters into specific cycle shortening or extension durations to form backwash cycle adjustment parameters.

[0184] In the above scenario, the required shortening or extension time is calculated based on the difference between the optimized backwash cycle parameter and the current backwash cycle. If the optimized backwash cycle is shorter than the current cycle, the cycle is shortened; if it is longer than the current cycle, the cycle is extended. This shortening or extension time is determined as the backwash cycle adjustment parameter.

[0185] Step S159: Arrange the real-time filtration rate adjustment parameters and the backwashing cycle adjustment parameters in chronological order and add corresponding execution time markers.

[0186] In the above scenario, the real-time filtration rate adjustment parameters and backwash cycle adjustment parameters are prioritized by time according to the system's operational priority and the urgency of the adjustments. Typically, the real-time filtration rate adjustment parameters need to be executed immediately, while the backwash cycle adjustment parameters are executed after the current backwash cycle ends. A corresponding execution time stamp is added to each adjustment parameter to clearly define the time when the parameter takes effect.

[0187] Step S1510: Integrate the arranged parameters and time stamps to generate dynamic correction instructions for the operating parameters of the volcanic rock water purification system.

[0188] In the above scenario, the real-time filtration rate adjustment parameters, backwash cycle adjustment parameters, and corresponding execution time markers, arranged in chronological order, are integrated to form a formatted dynamic correction instruction for operating parameters. The instruction includes information such as parameter type (filtration rate adjustment or backwash cycle adjustment), adjustment value (rate of change, duration of change, shortening or extending duration), and execution time. This instruction is sent to the control system of the volcanic rock water purification system, which then performs the corresponding operating parameter adjustment operation based on the instruction content.

[0189] Based on the same inventive concept, please refer to Figure 2 This paper shows a schematic block diagram of a data fusion processing system 100 for performing the above-described inspection video stream processing method in a volcanic rock water purification system, provided in an embodiment of this application. The data fusion processing system 100 for the volcanic rock water purification system may include a communication unit 110, a machine-readable storage medium 120, and a processor 130.

[0190] In this embodiment, both the machine-readable storage medium 120 and the processor 130 are located in the data fusion processing system 100 applied to the volcanic rock water purification system and are separately configured. However, it should be understood that the machine-readable storage medium 120 may also be independent of the data fusion processing system 100 applied to the volcanic rock water purification system and may be accessed by the processor 130 via a bus interface. Alternatively, the machine-readable storage medium 120 may also be integrated into the processor 130 and may communicate and interact with external systems through the communication unit 110.

[0191] The processor 130 is the control center of the data fusion processing system 100 applied to the volcanic rock water purification system. It connects to various parts of the data fusion processing system 100 via various interfaces and lines. By running or executing software programs and / or modules stored in the machine-readable storage medium 120, and by calling data stored in the machine-readable storage medium 120, it performs various functions and processes data of the data fusion processing system 100, thereby providing overall monitoring of the data fusion processing system 100. Optionally, the processor 130 may include one or more processing cores; for example, the processor 130 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor. The machine-readable storage medium 120 is used to store machine-executable instructions for executing the scheme of this application, and the processor 130 is used to execute the machine-executable instructions stored in the machine-readable storage medium 120 to implement the inspection video stream processing method provided in the aforementioned method embodiments.

[0192] It should be noted that, in order to simplify the description of the present invention and thus help to understand one or more embodiments of the invention, multiple features may sometimes be grouped into one embodiment, drawing or description thereof in the foregoing description of the embodiments of the present invention.

Claims

1. A data fusion processing method applied to a volcanic rock water purification system, characterized in that, The method includes: Acquire pore evolution data of volcanic rock filter media, pollutant adsorption and migration data, and filter operation feedback data. The pore evolution data of volcanic rock filter media records the temporal changes in the pore morphology between filter media particles as the water purification process progresses. The pollutant adsorption and migration data records the positional changes and morphological transformations of pollutants in the water body within the filter media layer. The filter operation feedback data records the changes in the state of the filter media caused by filtration rate adjustments and backwashing operations. The pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool are subjected to time-series correlation and capture processing to identify the morphological transformation nodes and transformation magnitudes of the pore evolution data of the volcanic rock filter media, the adsorption and migration data of pollutants, and the operation feedback data of the filter pool in the same time process. A dynamic correlation network is constructed based on the morphological transformation nodes and the transformation amplitude. The dynamic correlation network connects the interaction relationships of filter material pore morphology changes, pollutant migration path changes and operating parameter adjustments with time flow as the axis. The filter media state adaptive adjustment rules are generated based on the dynamic correlation network. The adaptive adjustment rules are used to describe the optimal matching mode between the filter media pore morphology and operating parameters under different pollutant migration states. Based on the adaptive adjustment rules for the filter media state, dynamic correction instructions for the operating parameters of the volcanic rock water purification system are generated. These dynamic correction instructions include real-time filtration rate adjustment parameters and backwashing cycle adjustment parameters.

2. The data fusion processing method applied to a volcanic rock water purification system according to claim 1, characterized in that, The process of performing time-series correlation capture processing on the pore evolution data of the volcanic rock filter media, the pollutant adsorption and migration data, and the filter operation feedback data to identify the morphological transformation nodes and transformation magnitudes of the pore evolution data of the volcanic rock filter media, the pollutant adsorption and migration data, and the filter operation feedback data within the same time process includes: The cross-sectional morphology time series and pore connectivity path time series of the filter media pores are extracted from the pore evolution data of the volcanic rock filter media. The cross-sectional morphology time series records the geometric shape data of the pores at different times, and the pore connectivity path time series records the connection mode data between the pores at different times. The spatial distribution time series and chemical transformation time series of pollutants in the filter layer are extracted from the pollutant adsorption and migration data. The spatial distribution time series records the location coordinate data of pollutants at different times, and the chemical transformation time series records the molecular structure data of pollutants at different times. Extract the filtration rate change time sequence and the backwashing effect time sequence from the filter operation feedback data. The filtration rate change time sequence records the water flow velocity data at different times, and the backwashing effect time sequence records the filter media state recovery data after backwashing. The time bases of the cross-sectional morphology time sequence, the pore connectivity path time sequence, the spatial distribution time sequence, the chemical morphology transformation time sequence, the filtration rate change time sequence, and the backwashing effect time sequence are unified to the same time unit. Using the unified time base as the axis, the abrupt change positions of the pore geometry in the cross-sectional morphology time sequence are compared segment by segment and marked as filter media pore morphology transformation nodes. By comparing the abrupt changes in the connection mode in the time sequence of the pore connectivity path, the nodes are marked as filter media pore connectivity transition nodes; by comparing the abrupt changes in the pollutant location coordinates in the time sequence of spatial distribution, the nodes are marked as pollutant migration transition nodes; by comparing the abrupt changes in molecular structure in the time sequence of chemical morphology transformation, the nodes are marked as pollutant morphology transformation nodes. The positions of abrupt changes in water flow velocity in the time sequence of filtration rate changes are marked as filtration rate adjustment transition nodes; the starting positions of filter media state recovery in the time sequence of backwashing effect are marked as backwashing operation transition nodes. Calculate the relative rate of change or standardized change of the data sequence before and after each transition node in its own dimension, and use it as the magnitude of the transition in its own data dimension. Arrange all transformation nodes in chronological order to form a sequence of transformation nodes on a unified timeline, and associate the transformation magnitude of each transformation node.

3. The data fusion processing method applied to a volcanic rock water purification system according to claim 2, characterized in that, The extraction of the cross-sectional morphology time series and pore connectivity path time series of the filter media pores from the volcanic rock filter media pore evolution data includes: The pore evolution data of the volcanic rock filter material was analyzed. This pore evolution data includes a sequence of pore images at different depths of the filter material layer acquired by microscopic imaging technology. The pore image sequence is processed by image segmentation to separate the pore region and non-pore region in each pore image; For each segmented pore region, the geometric shape parameters of its cross-section are extracted, including the perimeter of the outline, the diameter of the inscribed circle, and the shape factor, which together constitute the geometric shape data of the pore. The geometric shape data are arranged in the order of the acquisition time of the image sequence to form a time sequence of the cross-sectional morphology of the filter material pores; Connectivity analysis is performed on the pore image sequence to identify the connecting channels between adjacent pores and the width and length of the connecting channels; The connection method between the pores is determined based on the distribution of the connection channels, including direct connection, indirect connection, and isolated state; Record the connection methods and corresponding connection channel parameters of all pores at each time point, and arrange them in chronological order to form a time sequence of pore connectivity paths; Add the same time stamp to the cross-sectional morphology time sequence and the pore connectivity path time sequence.

4. The data fusion processing method applied to a volcanic rock water purification system according to claim 3, characterized in that, The connectivity analysis of the pore image sequence, identifying the connection channels between adjacent pores and the width and length of the connection channels, includes: Each image in the segmented pore image sequence is binarized, with the pore region labeled as the first value and the non-pore region labeled as the second value; Morphological erosion was used to preprocess the binarized image, and the connected component labeling algorithm was used to analyze the preprocessed image to identify each independent pore region and assign a unique identifier. Perform distance transformation on the marked pore regions and calculate the distance from each non-pore region pixel to the nearest pore region. Based on the distance transformation results, potential connection channel regions are identified. Potential connection channel regions are non-porous regions that are close to two or more pore regions. Morphological dilation is performed on the potential connection channel region. The central axis of the dilated connection channel region is extracted using a skeleton extraction algorithm and used as the path of the connection channel. The length of the central axis of the connection channel is measured and used as the length parameter of the connection channel. At the start and end points of the connecting channel, the channel width perpendicular to the central axis is measured, and the average value is taken as the width parameter of the connecting channel. The start point hole identifier, end point hole identifier, length parameter and width parameter of each connecting channel are associated and stored to complete the identification of the connecting channel between adjacent holes.

5. The data fusion processing method applied to a volcanic rock water purification system according to claim 1, characterized in that, The construction of a dynamic correlation network based on the morphological transformation nodes and the transformation amplitude includes: A network framework is constructed with time flow as the vertical axis. The vertical dimension of the network framework is the time process, and the horizontal dimensions are the filter media feature dimension, the pollutant feature dimension, and the operational feature dimension. The morphological transformation nodes are embedded into the corresponding horizontal dimension positions of the network framework according to their respective feature dimensions, and different feature dimension transformation nodes at the same time point are connected by horizontal connecting lines. For nodes with different feature dimensions that change at the same time point, determine their change direction (increase / decrease) and record their standardized change magnitude level. Adjust the width of the horizontal connecting line according to the correlation between the change direction and the matching degree of the magnitude level. The higher the correlation and the more matched the magnitude, the wider the line width. In the network framework, a longitudinal connection line is established between nodes where the filter material feature dimension changes at adjacent time points, and the direction of pore morphology change corresponding to the connection line is marked. Establish vertical connection lines between pollutant feature dimension change nodes at adjacent time points, and mark the direction of migration path change corresponding to the connection lines; Establish vertical connection lines between nodes that change the dimension of operational features at adjacent time points, and mark the parameter adjustment direction corresponding to the connection lines; For transition nodes of the same feature dimension at adjacent time points, calculate the standardized difference in the level of transition amplitude, and adjust the color depth of the vertical connecting line according to the magnitude of the difference in the level of transition amplitude; the greater the difference in level, the darker the color. Identify the time interval in the network framework where nodes simultaneously transition between filter media feature dimensions and pollutant feature dimensions, and add bidirectional arrows within this time interval to indicate the interaction between the two. After identifying the node that changes the operating feature dimension in the network framework, identify the time interval in which the filter media feature dimension or the pollutant feature dimension changes. Within this time interval, add a one-way arrow to indicate the impact of operating parameter adjustments on the filter media or pollutants. The frequency of occurrence of various types of connections in the statistical network framework is counted, and the connection pattern with the highest frequency is marked as the core association pattern; Using the core association pattern as the framework, and supplementing it with other connection patterns, a dynamic association network is formed that includes time flow, feature dimensions, transformation nodes, transformation magnitude, and interaction relationships.

6. The data fusion processing method applied to a volcanic rock water purification system according to claim 5, characterized in that, The identification network framework includes a time interval where both filter media feature dimension and pollutant feature dimension transition nodes occur simultaneously. Within this time interval, a bidirectional arrow is added to indicate the interaction between the two, including: Traverse the timeline of the network framework and check segment by segment whether there are cases where filter material feature dimension transformation nodes and pollutant feature dimension transformation nodes appear simultaneously in the same time interval. For each time interval in which two types of transition nodes occur simultaneously, determine the start and end times of the time interval and mark it as the interaction interval; Extract the change magnitude of filter media feature dimension change nodes and pollutant feature dimension change nodes within the interaction interval; Calculate the ratio of the characteristic change amplitude of the filter media to the characteristic change amplitude of the pollutants. This ratio is used to characterize the intensity ratio of the interaction between the two. The size of the bidirectional arrowheads is determined based on the intensity ratio; the closer the ratio is to 1, the more consistent the arrow sizes. Analyze the changing directions of filter media characteristic dimension data and pollutant characteristic dimension data within the interaction interval to determine the relationship between their interaction directions. The direction of action is marked on the double-headed arrow, including promoting and inhibiting relationships. A promoting relationship means that a change in one party exacerbates the change in the other party, while an inhibiting relationship means that a change in one party slows down the change in the other party. The synchronization rate of changes in the two feature dimensions within the statistical interaction interval is calculated. The higher the synchronization rate, the thicker the double-headed arrow. Add the labeled bidirectional arrows to the interaction area of ​​the network framework to connect the corresponding filter media feature dimension transformation nodes and pollutant feature dimension transformation nodes. Number all added bidirectional arrows and establish a correspondence table between arrows and their interaction ranges.

7. The data fusion processing method applied to a volcanic rock water purification system according to claim 1, characterized in that, The step of generating adaptive adjustment rules for filter media state based on the dynamic correlation network includes: Extract all sub-network segments containing pollutant migration and transformation nodes from the dynamic association network. Each sub-network segment contains filter media feature dimension data and operational feature dimension data corresponding to the pollutant migration and transformation node. The sub-network segments are divided into multiple pollutant migration state categories according to the transformation magnitude of pollutant migration transition nodes, and each pollutant migration state category corresponds to a pollutant migration intensity range; For each pollutant migration state category, extract the filter media pore morphology data from all sub-network segments under that pollutant migration state category, including cross-sectional morphology parameters and pore connectivity path parameters; The correlation between filter media pore morphology data and water purification effect under the same pollutant migration state category was analyzed, and the range of filter media pore morphology parameters that meet the preset standard for water purification effect was selected as the optimal filter media pore morphology range under the pollutant migration state category. Extract operational parameter data from all sub-network segments under each pollutant migration state category, including filtration rate adjustment parameters and backwash cycle parameters; Analyze the correlation between operating parameter data and filter media pore morphology parameters under the same pollutant migration state category, and determine the combination of operating parameters that can maintain the filter media pore morphology within the optimal range of filter media pore morphology, as the candidate combination of operating parameters under the pollutant migration state category. The candidate operating parameter combinations for each pollutant migration state category were substituted into the dynamic correlation network for simulation verification. The simulation process was observed to see whether the filter media pore morphology remained within the optimal filter media pore morphology range. Based on the verification results, adjust the numerical range of the candidate operating parameter combinations, eliminate parameter values ​​that cause the filter media pore morphology to deviate from the optimal filter media pore morphology range, and form the optimal operating parameter combination for this pollutant migration state category. The basic regulation rules are formed by binding the pollutant migration state category, the optimal filter media pore morphology range, and the optimal combination of operating parameters. Analyze the transition intervals between different pollutant migration state categories, and supplement each transition interval with a corresponding transition adjustment rule to ensure that the adjustment rules for adjacent pollutant migration state categories can be smoothly connected. By integrating all basic and transitional adjustment rules, adaptive adjustment rules for filter media state are formed. Each adaptive adjustment rule for filter media state describes the optimal matching mode between filter media pore morphology and operating parameters under a specific pollutant migration state.

8. The data fusion processing method applied to a volcanic rock water purification system according to claim 7, characterized in that, The analysis establishes a correlation between filter media pore morphology data and water purification effect under the same pollutant migration state category. It then selects the range of filter media pore morphology parameters that meet preset standards for water purification effect, defining this as the optimal range for filter media pore morphology under that pollutant migration state category. This includes: Retrieve historical water purification effect data corresponding to each pollutant migration status category, including effluent quality compliance status and purification efficiency data; The historical water purification effect data is divided into a compliant data group and a non-compliant data group according to a preset standard, which is determined based on the effluent water quality requirements of the water purification system. Extract the filter media pore morphology data corresponding to the compliant data set, including cross-sectional morphology parameters and pore connectivity path parameters; Statistical analysis was performed on various cross-sectional morphological parameters of the compliant data groups to calculate the numerical distribution characteristics of each parameter and determine the compliant numerical range for each parameter. Statistical analysis was performed on various pore connectivity path parameters of the compliant data group, the numerical distribution characteristics of each type of parameter were calculated, and the compliant numerical range of each type of parameter was determined. The acceptable value ranges of various cross-sectional morphological parameters and the acceptable value ranges of various pore connectivity path parameters are combined side by side to form an initial set of pore morphological parameter ranges for candidate filter media. Extract the filter media pore morphology data corresponding to the non-compliant data groups, analyze the difference between them and the initial candidate filter media pore morphology range, and determine the parameter value range that needs to be excluded. The parameter value range that needs to be excluded is removed from the initial candidate filter media pore morphology range to form the optimized candidate filter media pore morphology range; The optimized candidate filter media pore morphology range was verified, and the historical water purification effect compliance rate corresponding to the filter media pore morphology data within the candidate filter media pore morphology range was calculated. If the historical water purification effect compliance rate reaches the preset threshold, the optimized candidate filter media pore morphology range will be determined as the optimal filter media pore morphology range under the pollutant migration state category. If the historical water purification effect compliance rate does not reach the preset threshold, the statistical analysis method of the parameters will be adjusted, and the concentrated distribution range will be re-determined until the historical water purification effect compliance rate meets the requirements.

9. The data fusion processing method applied to a volcanic rock water purification system according to claim 1, characterized in that, The step of generating dynamic correction instructions for the operating parameters of the volcanic rock water purification system based on the adaptive adjustment rules of the filter media state includes: Real-time acquisition of pollutant migration data from the volcanic rock water purification system, and extraction of spatial distribution parameters and chemical speciation parameters of pollutants; The spatial distribution parameters and chemical speciation parameters of the pollutants are matched with the pollutant migration state categories in the adaptive adjustment rules of the filter media state to determine the current pollutant migration state category. Based on the matching results, retrieve the optimal filter media pore morphology range and optimal operating parameter combination corresponding to the pollutant migration state category; Real-time acquisition of the current filter media pore morphology data of the volcanic rock water purification system, comparison of the various parameters contained therein with the standard value range of the corresponding parameters in the optimal filter media pore morphology parameter range set, and calculation of the deviation parameters between the two. The overall deviation of the filter media pore state is judged by comprehensively considering the deviation of various parameters, and the filtration rate adjustment parameter and backwashing cycle parameter in the optimal operating parameter combination are adjusted according to this comprehensive deviation. The greater the comprehensive deviation, the greater the adjustment range. Extract historical correlation segments from the dynamic correlation network that are similar to the current pollutant migration state category and filter media pore morphology deviation parameters, and obtain the operational parameter adjustment effect data in the historical correlation segments; Based on the operational parameter adjustment effect data, the adjusted filtration rate adjustment parameters and backwash cycle parameters are further optimized, and the optimized filtration rate adjustment parameters are converted into specific filtration rate change rate and change duration to form real-time filtration rate adjustment parameters. The optimized backwash cycle parameters are converted into specific cycle shortening or lengthening durations to form backwash cycle adjustment parameters. Arrange the real-time filtration rate adjustment parameters and the backwash cycle adjustment parameters in chronological order and add corresponding execution time markers; By integrating and arranging the parameters and time stamps, dynamic correction instructions for the operating parameters of the volcanic rock water purification system are generated.

10. A data fusion processing system applied to a volcanic rock water purification system, characterized in that, include: processor; A machine-readable storage medium for storing machine-executable instructions of the processor; The processor is configured to execute the data fusion processing method for a volcanic rock water purification system according to any one of claims 1 to 9 by executing the machine-executable instructions.

Images