An ion migration analysis method and system based on solid-liquid interface microenvironment modeling

Abstract

The application discloses an ion migration analysis method and system based on solid-liquid interface microenvironment modeling, and particularly relates to the field of water quality analysis and biological deodorization operation monitoring, and comprises the following steps: obtaining multi-period data of a water phase of a biological deodorization solid-liquid interface; an intelligent sensing system performs water analysis to obtain target ion concentration; an electrochemical variable research or analysis material is used to obtain pH value and oxidation-reduction potential; dissolved oxygen and biological activity are synchronously obtained; an edge computing node forms a record sequence according to a sampling time; based on the record sequence, a difference of a concomitant quantity of adjacent records is calculated, and a state code is generated according to a symbol; a record between two adjacent times of repeating the state code forms a turnaround section, and a turnaround point is determined according to an absolute value of a rank difference; through identification of a turnaround state of the concomitant quantity of the solid-liquid interface, a forward path and a return path are matched, and a reversible concentration change in the same period is deducted; a cross-period residual response is offset and distributed, and a target ion net response and a migration stage are obtained.

Description

Technical Field

[0001] This invention relates to the field of water quality analysis and biological deodorization operation monitoring technology. More specifically, this invention relates to an ion migration analysis method and system based on solid-liquid interface microenvironment modeling. Background Technology

[0002] Current biological deodorization operation monitoring mainly focuses on the ion concentration, pH, redox potential, dissolved oxygen, and biological activity in the solid-liquid interface region on the packing surface. The changes in the magnitude and direction of each parameter are compared according to the sampling time sequence, and the migration state of pollutants between the liquid phase and the biofilm is analyzed accordingly. After the malodorous gas enters the liquid phase through gas-liquid mass transfer, the start and stop of the spray, the adjustment of the spray volume, and the process of draining will change the thickness of the liquid film on the surface of the packing. The biological deodorization device cannot be interrupted during normal treatment due to ion migration analysis, and it is also difficult to directly obtain the volume of the interfacial liquid film and the amount of target ions retained. In this case, thickening of the liquid film will cause a decrease in the concentration of the target ion, while thinning of the liquid film will cause a rebound in the concentration of the target ion. The above concentration changes are similar to the response formed by the target ion entering the biofilm or being released from the biofilm in the sampling results. Existing methods, if they directly establish interface microenvironment models based on concentration differences, change directions, or time correlations between adjacent time points, are prone to misrepresenting the concentration decrease caused by liquid film dilution as ion migration and the concentration increase caused by liquid film concentration as ion migration, thus causing the identification results of subsequent migration stages to deviate from the actual process. The technical problem to be solved by this application is: how to distinguish between liquid film dilution, liquid film concentration and actual migration of target ions from multidimensional state data of the solid-liquid interface under the condition that the biological deodorization device is running continuously and the volume of the interfacial liquid film and the amount of ions retained cannot be directly measured. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide an ion migration analysis method and system based on solid-liquid interface microenvironment modeling. By identifying the return state of the accompanying quantity at the solid-liquid interface, matching the outgoing and return paths and subtracting the reversible concentration changes within the same period, and then offsetting and allocating the remaining response across periods, the net response and migration stage of the target ion are obtained, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: an ion migration analysis method based on solid-liquid interface microenvironment modeling, comprising: S1. Acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. S2. Calculate the accompanying difference between adjacent records based on the record sequence and generate a status code according to the symbol. Form a return segment based on the records between two consecutive repetitions of the status code. Determine the turnaround point according to the absolute value of the rank difference to form the outgoing path and the reverse return path. S3. The discrete Fraser matching algorithm establishes a distance table. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order. The pairing sequence is formed by backtracking from the last cell. S4. Read the target ion concentration difference according to the pairing sequence. Subtract the first value of the reverse concentration difference in ascending absolute value. Retain the same concentration difference and form the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. Establish a time unfolding network. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and target ion net response sequence. S5. Using the return segment as the state node, connect adjacent return segments with the same first state code and consistent paired state positions, and write the target ion net response to form a solid-liquid interface microenvironment evolution model; when the net response is in the same direction and the pairing order is not reversed, record the stable migration stage; when the direction is reversed, record the transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, record the abnormal migration stage, and generate ion migration analysis results.

[0005] In a preferred embodiment, S1 includes: S11. The intelligent sensing system performs water analysis on the solid-liquid interface aqueous phase at the same packing location, obtains the target ion concentration and dissolved oxygen according to the sampling time, and uses electrochemical variables to study or analyze the material to obtain acidity, alkalinity and redox potential. The bioactivity and sampling time of each field obtained simultaneously are written into the original record. The process of performing water analysis is as follows: the intelligent sensing system reads the original response values ​​of the target ion detection channel and the dissolved oxygen detection channel at the same packing position at a unified sampling time, performs numerical conversion based on the zero point value, range value and calibration relationship of each detection channel, and writes the conversion results and sampling time into the target ion concentration field and the dissolved oxygen field. Target ion concentration and dissolved oxygen refer to the molar concentration of target ions per unit volume of the aqueous phase at the solid-liquid interface at the same sampling time; dissolved oxygen refers to the mass concentration of oxygen existing in the molecular state per unit volume of the aqueous phase at the solid-liquid interface at the same sampling time. The process of using electrochemical variables to study or analyze materials to obtain acidity / alkalinity and redox potential is as follows: the intelligent sensing system reads the potential difference between the acidity / alkalinity detection electrode and the reference electrode, and converts the acidity / alkalinity according to the two-point calibration relationship; simultaneously, the potential difference between the redox detection electrode and the reference electrode is read, and the potential difference is converted to a unified reference electrode standard to obtain the redox potential; Bioactivity refers to the ratio of the change in dissolved oxygen in the aqueous phase at the solid-liquid interface between adjacent sampling times at the same packing location to the corresponding sampling duration, and is used to characterize the oxygen consumption response of the biofilm during the corresponding sampling period. S12. Edge computing nodes form electrochemical change segments with adjacent acidity and redox potentials. Linear interpolation is performed according to the time position of each field sampling time in the electrochemical change segment to generate the electrochemical state corresponding to each field sampling time. The electrochemical state corresponding to the target ion concentration sampling time is recorded as the baseline state. The linear interpolation process is as follows: For two electrochemical sampling times before and after the target sampling time, the corresponding acidity and redox potential are read respectively. The proportion of the time between the target sampling time and the previous electrochemical sampling time to the interval between the two electrochemical sampling times is used as the weight of the later time, and the remaining proportion is used as the weight of the earlier time. The acidity and redox potential of the two times are weighted and summed respectively to obtain the electrochemical state corresponding to the target sampling time. S13. For dissolved oxygen and biological activity in the current operating cycle and adjacent operating cycles, calculate the sum of absolute values ​​of the field differences between the corresponding electrochemical state and the baseline state. Take the first item in ascending order according to the sum of absolute values ​​of field differences, sampling time difference and operating cycle. Write the dissolved oxygen, biological activity, target ion concentration and baseline state corresponding to the first item into the same interface record. Arrange the interface records according to the sampling time of the target ion concentration to form a record sequence.

[0006] In a preferred embodiment, S2 includes: S21. For adjacent interface records in the record sequence, calculate the difference between the next accompanying quantity and the previous accompanying quantity for each field, write the negative code, hold code or positive code according to the sign of the difference, and concatenate the field codes according to the order of the accompanying quantity fields to form the status code corresponding to the interface record. The accompanying quantities refer to the pH, redox potential, dissolved oxygen, and biological activity recorded on the same interface as the target ion concentration. The four accompanying quantity fields are arranged in a fixed order of pH, redox potential, dissolved oxygen, and biological activity. Field coding refers to the direction mark generated based on the difference sign of the same accompanying value field in adjacent interface records. When the difference is less than zero, a negative code is written; when the difference is equal to zero, a hold code is written; and when the difference is greater than zero, a positive code is written. The four field codes are concatenated in the order of the accompanying value fields to form the status code corresponding to the current interface record. S22. Starting from the first status code, search for the same status code. The first detected same status code and the interface record between them are divided into a return segment. Then, continue searching from the last status code after the last position of the return segment to form a sequence of return segments arranged according to the sampling time.

[0007] In a preferred embodiment, S2 further includes: S23. In each return segment, assign ranks according to the ascending order of the values ​​of each field of the accompanying quantity. When the field values ​​are the same, assign ranks according to the order of sampling time. Calculate the sum of the absolute values ​​of the field rank differences between each interface record and the first interface record of the return segment. Determine the first interface record as the turnaround point according to the descending order of the absolute values ​​of the field rank differences and the ascending order of the sampling time. S24. The outgoing path is formed by recording the first and second interfaces of the return segment to the turning point, and the return path is formed by recording the reverse arrangement of the first and second interfaces of the return segment to the turning point. The corresponding relationship between the status code, the rank of the accompanying quantity, and the target ion concentration in each path is retained.

[0008] In a preferred embodiment, S3 includes: S31. The edge computing node reads the rank and status code of the associated quantity of the aqueous phase at the solid-liquid interface in the outgoing and returning paths. It swaps the negative and positive codes in the status code of the returning path and retains the holding code to form a mirrored returning path. Then, for any outgoing record and any mirrored returning record, it calculates the absolute value of the difference in the status code, the absolute value of the difference in the rank of the associated quantity, and the absolute value of the difference in the cross product of the outgoing and returning order in sequence. The node forms the cell cost word according to the field order and writes it into the paired grid. The target ion concentration does not participate in the calculation of the cell cost word. The outbound record refers to any interface record arranged according to the sampling time between the first interface record of the return segment and the turnaround point. The outbound record carries the outbound sequence, sampling time, status code, rank of the accompanying quantity, and target ion concentration. Mirror retrace record refers to the interface record obtained by reversing the sampling time of the interface record from the turnaround point to the end of the retrace segment, rewriting the negative code in the status code of each retrace record to a positive code, rewriting the positive code to a negative code, and retaining the hold code. The mirror retrace record retains the retrace sequence, sampling time, rank of the accompanying quantity, and target ion concentration. S32. The Discrete Fraser matching algorithm starts recursively from the first cell of the paired grid. The current cell accepts the left neighbor cell, the upper neighbor cell, and the upper left neighbor cell respectively. The peak cost word of the predecessor path and the cost word of the current cell are sorted in descending order in lexicographical order and written as the new peak cost field. The cumulative field of the predecessor is added to the cost word of the current cell field field by field and written as the new cumulative field. For the left neighbor acceptance or the upper neighbor acceptance, a repetition count is added once. For the upper left neighbor acceptance, the repetition count is maintained. The first item is sorted in ascending order in the order of peak field, cumulative field, repetition count, and upper left neighbor cell priority over left neighbor cell and upper neighbor cell to form the forward cost table. The discrete Fraser matching algorithm recursively starts from the first cell of the paired grid as follows: the paired grid corresponding to the first outgoing record and the first mirrored return record is taken as the first cell. The cost word of the first cell is written into the peak field and the cumulative field of the first cell at the same time, and the repetition count of the first cell is written to zero. Each cell in the first row only accepts the left neighbor cell, each cell in the first column only accepts the upper neighbor cell, and the remaining cells accept the left neighbor cell, the upper neighbor cell, and the upper left neighbor cell respectively. The current cell accepts the result according to the peak field, cumulative field, repetition count and predecessor order recorded in S32, until the last cell completes the recursion. The peak cost word of the preceding path refers to the last sorted unit cost word obtained after comparing the cost words of each unit in lexicographical order on the connecting path from the first cell of the paired grid to the predecessor cell of the current cell. It is used to indicate the peak cost of the path that has already appeared on the connecting path. Lexicographical order refers to comparing two cost words field by field in the order of the absolute difference between the status code digits, the absolute difference between the rank of the accompanying quantity and the absolute difference between the cross products of the outgoing and returning order. If the previous field is the same, the next field is compared. The comparison result of the first field with different values ​​is used as the sorting result of the two cost words.

[0009] In a preferred embodiment, S3 further includes: S33. Perform a recursive process from the last cell of the paired grid to the first cell in the opposite direction to S32 to form a backward cost table. Merge the forward cost, cell cost word and backward cost cell by cell, and retain the paired cells whose merged result is consistent with the path cost word of the last cell to form a candidate paired grid that simultaneously satisfies the water phase state order, path monotonically connected and discrete Fraser cost constraints. The pairing cell refers to the grid position in the grid determined by the intersection of a outgoing record and a mirrored return record. The pairing cell is written with outgoing sequence, return sequence, cell cost word, forward cost, backward cost and acceptance direction, which is used to represent a candidate pairing relationship between two records. S34. Read pairing units from the last cell to the first cell and from the first cell to the last cell of the candidate pairing grid to form two index sequences. If the two index sequences are inconsistent, compare them bit by bit according to the index field and delete the first difference pairing unit from the index sequence that is later in the sequence. Then, re-execute S32 and S33 until the two index sequences are consistent. Write the outbound record and return record corresponding to the consistent index sequence as a pairing sequence.

[0010] In a preferred embodiment, S4 includes: S41. Read the target ion concentration of each outgoing record, return record and its previous record according to the pairing sequence. Subtract the concentration of the previous record from the current record concentration to form the outgoing concentration difference and the return concentration difference. When the two concentration differences have opposite signs, take the first value in ascending order of absolute value as the reversible quantity and subtract it from the two concentration differences. When the signs are the same, retain the two concentration differences. Write the subtracted balance and the retained concentration difference as the residual response with the running cycle, packing position, pairing position, response sign and response quantity. S42. Establish supply nodes with negative residual responses and demand nodes with positive residual responses. Only under the same target ion, the same filler position, and the same pairing position, establish offset edges from the early supply node to the late demand node. Use the total number of residual response nodes plus one as the carry base. Multiply the operating cycle interval, the number of status codes, and the difference in record index by the square of the carry base, the carry base, and one, and then add them to form the offset edge cost. Take the first value in ascending order of the absolute values ​​of supply and demand as the offset edge capacity. Add a net response node. Write the total demand minus the total supply as the net response node balance. Use the cube of the carry base plus the index of the residual response node to form the net response edge cost. Establish a time-unfolded network. The early supply node refers to a node whose remaining response is negative and whose combined order of the corresponding cycle number and record index is higher than that of another remaining response. The late demand node refers to a node that corresponds to the same target ion, the same filler position, and the same pairing position as the early supply node, whose remaining response is positive and whose combined order is after that of the early supply node.

[0011] In a preferred embodiment, S4 further includes: S43. The edge computing node adopts the network simplex algorithm. The initial base tree is established by the net response edges of each supply node pointing to the net response node and the net response edges of the net response node pointing to each demand node. The initial flow is written according to the balance of the net response node and the node potential is calculated based on the tree edge cost. The non-tree edge reduction cost is calculated round by round. The negative first edge is selected in ascending order according to the reduction cost, source node index and target node index and replaced into the base tree. The first value of the loop formed by the replaced edge is taken in ascending order according to the remaining capacity to perform augmentation. Then, the first edge of the loop edge index that the flow reaches the capacity is taken in ascending order and replaced out of the base tree until the non-tree edge reduction cost is non-negative. The process of using the network simplex algorithm is as follows: First, an initial base tree is established by expanding all nodes and net response edges in the network over time, satisfying the balance of each node. The reduced cost of each base tree edge is set to zero, and the node potential is calculated accordingly. Then, the reduced cost of each non-tree edge is calculated. Non-tree edges with negative reduced costs are selected and swapped into the base tree. A loop is formed by the unique path between the swapped edge and the two endpoints of the base tree. The flow is increased or decreased along the loop until at least one edge in the loop reaches its capacity or its flow reaches zero. The corresponding edge is then swapped out of the base tree. The calculation of node potential, reduced cost, and swapping is repeated until the reduced cost of all non-tree edges is non-negative. Initial flow refers to the flow written to the net response edge after the initial base tree is established. The flow of the net response edge from the supply node to the net response node is equal to the corresponding supply, and the flow of the net response edge from the net response node to the demand node is equal to the corresponding demand. The outflow of each node minus the inflow is equal to the node balance. Tree edge cost refers to the edge cost carried by each edge in the current base tree; for offset edges, the tree edge cost is the offset edge cost formed by combining the running cycle interval, the out-of-digit status code, and the difference in record index according to the carry-in radix; for net response edges, the tree edge cost is the sum of the cube of the carry-in radix and the index of the corresponding remaining response node. The process of calculating the node potential is as follows: set the node potential of the net response node to zero, and expand from the net response node along the base tree edge by edge; when the direction of the tree edge is from an assigned node to an unassigned node, add the value of the assigned node to the cost of the tree edge and write it as the unassigned node potential; when the direction of the tree edge is from an unassigned node to an assigned node, subtract the cost of the tree edge from the value of the assigned node and write it as the unassigned node potential, until all nodes have obtained the node potential; The process of calculating the reduced cost of non-tree edges is as follows: For any non-tree edge pointing from the source node to the target node, add the cost of the non-tree edge to the potential of the source node, and then subtract the potential of the target node. The difference is taken as the reduced cost of the non-tree edge. The source node index refers to the integer order assigned to the non-tree edge starting node after arranging all remaining response nodes in ascending order of runtime number, record index, and node category. The augmentation process is as follows: After the swapped edge enters the base tree, the source node is connected to the target node along the direction of the swapped edge, and a loop is formed by returning from the target node to the source node in the original base tree; for the edge in the loop that is in the same direction as the swapped edge, the edge capacity is calculated and the current flow is subtracted to obtain the positive margin; for the edge in the opposite direction, the current flow is read as the negative margin; the first value of all margins is taken as the augmentation amount; the augmentation amount is added to the edge in the same direction and subtracted from the edge in the opposite direction. S44. Based on the time expansion network, check the inflow, outflow and net response node balance of each node, and check the flow and edge capacity of each edge. If any check result is inconsistent, recalculate the potential of the corresponding node and continue to execute S43. When all check results are consistent, write the offset edge flow as the cross-cycle offset amount, and arrange the net response edge flow according to the running cycle and record index to form the target ion net response sequence. Inflow refers to the sum of all network edge traffic with the current node as the target node, and outflow refers to the sum of all network edge traffic with the current node as the source node. During node verification, outflow minus inflow should equal the current node's balance.

[0012] In a preferred embodiment, S5 includes: S51. Read the segment head status code, pairing status position, target ion net response and cross-cycle cancellation amount corresponding to each return segment, establish a status node with the return segment, establish a directed edge between adjacent status nodes with the same segment head status code and the same pairing status position, and write the target ion net response direction, response amount and pairing order into the status node to form a solid-liquid interface microenvironment evolution model. S52. Compare the target ion net response direction and pairing order of adjacent state nodes along the directed edge. Merge state nodes with the same target ion net response direction and whose pairing order is not earlier than that of the previous state node into a stable migration stage. Write the first state node interval from the change of target ion net response direction to the new direction as the transition migration stage. S53. For any internal state node, when the current target ion net response direction is different from the previous and subsequent state nodes, the previous and subsequent state nodes have the same direction, and the current cross-cycle cancellation amount is zero, the current state node is separated from the corresponding migration stage and written as the abnormal migration stage. Then, the stable migration stage, transitional migration stage, and abnormal migration stage are arranged in the order of the running cycle to generate the ion migration analysis results of the target ion in the aqueous phase at the solid-liquid interface.

[0013] An ion migration analysis system based on solid-liquid interface microenvironment modeling, the system comprising a data acquisition module, a state segmentation module, a path pairing module, a response allocation module, and a stage discrimination module: The data acquisition module is used to acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. The state segmentation module calculates the accompanying difference between adjacent records based on the record sequence and generates a status code according to the symbol. The records between two consecutive repetitions of the status code form a return segment. The turnaround point is determined according to the absolute value of the rank difference, forming the outgoing path and the reverse return path. The path pairing module is used to establish a distance table using the discrete Fraser matching algorithm. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order, and the pairing sequence is formed by backtracking from the last cell. The response allocation module is used to read the target ion concentration difference according to the pairing sequence. The reverse concentration difference is subtracted by the first value in ascending absolute value order, and the same-direction concentration difference is retained. The remaining amount forms the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. A time unfolding network is established. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and the target ion net response sequence. The stage discrimination module is used to write the target ion net response to form a solid-liquid interface microenvironment evolution model, with the return segment as the state node, connecting adjacent return segments with the same first state code and consistent paired state positions; when the net response is in the same direction and the paired order is not reversed, it is recorded as a stable migration stage; when the direction is reversed, it is recorded as a transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, it is recorded as an abnormal migration stage, generating ion migration analysis results.

[0014] The technical effects and advantages of this invention are as follows: This scheme distinguishes between dilution and concentration response and target ion migration without measuring liquid film volume and ion retention by using associated quantity back identification and discrete Fraser pairing, thereby reducing migration misjudgment. By combining the outbound concentration difference with the return concentration difference, reversible concentration changes within the same cycle can be separated, reducing interference caused by spray start-up and shutdown, spray volume adjustment and drainage process. A time-expanded network is established for the residual response across cycles, and a simplex distribution is performed on the network to ensure a unified acceptance of the retention and release responses, thereby improving the consistency of the net response calculation for the target ion. Establishing offset edges within the same target ion, filler position, and pairing state position range can suppress cross-matching of responses from different sources, thereby improving the physical correspondence and traceability of the results. A microenvironment evolution model of the solid-liquid interface is constructed using status codes, return segments, and pairing sequences. This model can identify stable, transitional, and abnormal migration stages according to the operating cycle, thereby improving the reliability of stage division. Attached Figure Description

[0015] Figure 1 This is a flowchart of the ion migration analysis process of the present invention; Figure 2 This is a diagram of the ion migration analysis system of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figure 1-2 The present invention provides an ion migration analysis method based on solid-liquid interface microenvironment modeling, comprising: S1. Acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. The target ion concentration, pH, redox potential, dissolved oxygen and bioactivity in the aqueous phase at the solid-liquid interface are derived from different detection channels, and each detection channel has an independent sampling period and response time. To ensure a unified time reference for all fields recorded on the same interface, the sampling time of the target ion concentration is used as the interface recording time. Simultaneously, the actual sampling time, cycle number, packing position code, and calibration version of each field are retained. The original response conversion, electrochemical state interpolation, and candidate record matching are then performed sequentially. This implementation process includes the following: The aqueous phase detection results are first uniformly registered according to the detection object and source location; the intelligent sensing system writes the target ion number, the filler position code and the operating cycle number respectively for the target ion, the filler position code and the operating cycle number, and each detection channel shares the edge computing node clock and records the actual sampling time; The original response values ​​of the target ion detection channel and the dissolved oxygen detection channel are converted into the target ion concentration and dissolved oxygen mass concentration in S11 according to the zero-point response value, range response value, conversion relationship, output unit and calibration version in their respective calibration tables. The potential difference between the acidity and alkalinity detection electrode and the reference electrode is converted into acidity and alkalinity according to the two-point calibration relationship. The potential difference between the redox detection electrode and the reference electrode is converted to a unified reference electrode standard to form the redox potential. Biological activity is calculated by dividing the difference between the previous dissolved oxygen value and the next dissolved oxygen value by the time difference between the two sampling times. A positive value indicates oxygen consumption response, and a negative value indicates dissolved oxygen replenishment response. The sampling time for biological activity is the next dissolved oxygen sampling time. The conversion result, actual sampling time and calibration version are written to the original record. The repeated sampling time is retained and written to the last record in the sequence. When the original response exceeds the calibration range, an out-of-bounds flag is written and the corresponding record is stopped from proceeding to subsequent processing. The time offset between different detection channels is corrected by electrochemical state interpolation; for any sampling time to be converted, the edge computing node reads two electrochemical sampling times before and after the sampling time and forms an electrochemical change segment with two electrochemical records. In S12, the electrochemical state conversion generates the weight of the next time step according to the ratio of the time between the sampling time to be converted and the previous electrochemical sampling time to the total time of the electrochemical change segment. The weight of the previous time step is generated by subtracting the weight of the next time step from the weight of the next time step. Then, the weighted summation is performed on the acidity and alkalinity before and after the sampling time and the redox potential before and after the sampling time to be converted, respectively, to obtain the electrochemical state corresponding to the sampling time to be converted. The electrochemical state corresponding to the sampling time of the target ion concentration is written as the baseline state, and the electrochemical states corresponding to the sampling times of dissolved oxygen and biological activity are written into the candidate records. When the sampling time to be converted is the same as the electrochemical sampling time, the corresponding acidity and redox potential are read directly. When the two electrochemical sampling times are the same, the last record in the order of writing is retained. When there is a lack of electrochemical records on one side of the sampling time to be converted, the extrapolation is stopped and the corresponding original record is written to the record area to be supplemented. The interface records candidate matching based on the similarity of electrochemical states and time sequence; the dissolved oxygen corresponding to the same dissolved oxygen sampling time and the biological activity calculated with the sampling time as the last time are combined into candidate records, and the electrochemical state corresponding to the candidate records is read. The relevant matching processing is recorded in S13. The difference in pH is divided by the pH detection resolution to form the difference in pH number of cells, and the difference in redox potential is divided by the redox potential detection resolution to form the difference in redox potential number of cells. The sum of the absolute values ​​of the two cell differences is written as the sum of the absolute values ​​of the field differences. The first item is selected in ascending order according to the absolute value of the field difference, the time difference between the candidate sampling time and the target ion concentration sampling time, the difference between the cycle number of the candidate running cycle and the current running cycle, and the candidate record index. The dissolved oxygen, biological activity, source sampling time, source running cycle, target ion concentration and baseline status in the first item are written into the same interface record, and an occupancy mark is written to the candidate record. When located at the beginning or end of multi-cycle data, only the existing adjacent running cycles are read. When all candidate records are occupied, there is a lack of complete electrochemical states, or there are no candidate records, the target ion concentration and the baseline state are written into the interface record to be supplemented. After the original record is added, it is re-matched. Finally, the interface records are arranged according to the sampling time of the target ion concentration to form a record sequence. Taking the same packing position in the biological deodorization tower as an example, after the target ion detection channel generates the target ion concentration record, the edge computing node reads the pH and redox potential records before and after the sampling time, converts them into the baseline state, and then reads the unoccupied dissolved oxygen and biological activity candidate records from the current operating cycle and the adjacent operating cycles that have been obtained. After converting the candidate electrochemical state and the baseline state into the resolution grid number, the node performs sorting and writes the first item of the sorting and the target ion concentration together into the interface record. The resulting record sequence retains the detection values ​​of each field, the source sampling time, the operating cycle number, the filler position code, and the calibration version, providing definite input for subsequent calculation of accompanying quantity differences, generation of status codes, and identification of return segments.

[0018] S2. Calculate the accompanying difference between adjacent records based on the record sequence and generate a status code according to the symbol. Form a return segment based on the records between two consecutive repetitions of the status code. Determine the turnaround point according to the absolute value of the rank difference to form the outgoing path and the reverse return path. Regarding the departure and return process of the aqueous phase state at the solid-liquid interface in the recorded sequence, this implementation first converts the accompanying quantity changes into a state code with a clear direction, then extracts the return segment based on this, and determines the turnaround point, outbound path, and return path within the return segment, so that the state code, accompanying quantity rank, and target ion concentration read by subsequent discrete Fraser matching have a unified recording source; the specific processing is as follows: First, establish the accompanying quantity change code around the records of adjacent interfaces; the edge computing node divides the value of each field by the corresponding detection resolution or calculation resolution according to the fixed field order of pH, redox potential, dissolved oxygen and biological activity, and takes the integer cell number, and then subtracts the cell number of the previous interface from the cell number of the subsequent interface record. The relevant processing is recorded in S21. When the cell number difference is less than zero, a negative code is written; when the cell number difference is equal to zero, a hold code is written; when the cell number difference is greater than zero, a positive code is written. The four fields are then concatenated in a fixed order to form a status code that points from the previous screen record to the current screen record. The status code is then attached to the current screen record. If the first interface record of the record sequence is missing a previous interface record, no status code is generated. If the detection field is missing or has an out-of-bounds marker, the current interface record is written with a missing code marker and the search for the return segment is stopped. After the status code is written, the edge computing node reads from the first valid status code backward according to the sampling time. During the execution of S22, the current status code is used as the segment start status code. The first occurrence of the same status code is retrieved. It is required that there is at least one different status code between the two identical status codes. When this relationship is satisfied, the two identical status codes and the corresponding interface records between them are assigned to the same return segment. When two adjacent status codes are the same, the corresponding interface record is included in the status holding area and no separate return segment is established; after the return segment is formed, the search continues from the first valid status code after the last interface record of the return segment, and the return segment sequence is generated in sequence. If the segment start status code does not reappear after reading to the end of the record sequence, the segment start status code and subsequent interface records are written to the waiting area. The search continues after the new interface record arrives. When the return segment spans a running cycle, the running cycle number at the beginning of the segment, the running cycle number at the end of the segment, and the source running cycle number of each interface record are retained; Once inside a single return segment, each accompanying quantity field is assigned a rank in ascending order of its field value. For interface records with the same field value, different ranks are assigned according to the order of sampling time. The time order is only used to eliminate parallelism, and the original field value is still retained in the interface record. S23 calculates the acid-base rank difference, redox potential rank difference, dissolved oxygen rank difference, and biological activity rank difference of the current interface record relative to the first and last interface records of the return segment, and sums the absolute values ​​of the four rank differences. The first item is read in descending order of the summation result and ascending order of the sampling time, and the interface record corresponding to the first item is determined as the turnaround point. When there is no intermediate interface record in the return segment, no return point is generated, and the corresponding interface record is retained for subsequent status code retrieval; when the field summation results are parallel, the one with the sampling time earlier is determined as the return point first. The path organization phase divides the departure process and the return process by using the turnaround point as the boundary; the first and second interfaces of the return segment record up to the turnaround point and arrange them according to the sampling time to form the outgoing path; the interface records from the turnaround point to the last interface of the return segment are first read according to the sampling time and then arranged in reverse order to form the return path; the relevant processing is written into S24. The turnaround point serves as both the last interface record of the outbound path and the first interface record of the return path. Each path retains the interface record index, source operation cycle number, sampling time, status code, original value of the accompanying quantity, rank of the accompanying quantity, and target ion concentration. After the return path is reversed, the status code is not rewritten for the time being. In subsequent mirror return processing, the negative code and positive code are swapped according to the original changed edge between adjacent return interface records, and the code is kept, thereby avoiding a one-bit misconnection between the status code and the interface record. After the above processing, the status code can reflect the direction of the accompanying quantity change between adjacent interface records, the return segment can limit the recording range of the water phase state of the solid-liquid interface after leaving the state at the beginning of the segment and then returning to the state at the beginning of the segment, and the turnaround point is used to separate the state leaving and state returning processes. If, within a certain operating cycle, pH, redox potential, dissolved oxygen, and biological activity together form the sequence of changes of “segment start status code, opposite direction status code, segment start status code”, the edge computing node establishes a return segment based on this sequence and determines the turnaround point from within the return segment. The resulting outgoing and return paths are directly used for subsequent pairing mesh generation and discrete Fraser matching reading.

[0019] S3. The discrete Fraser matching algorithm establishes a distance table. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order. The pairing sequence is formed by backtracking from the last cell. The outbound path and the return path represent the changes in the state of the aqueous phase leaving the beginning of the segment and returning to the beginning of the segment at the solid-liquid interface, respectively. There are differences in the number of records, the duration of state dwell, and the sampling order between the two. This implementation first reconstructs the mirrored return path according to the return change edge, then establishes a pairing grid based on the associated quantity state and relative progress. The pairing sequence is determined by verifying forward cost, backward cost, and bidirectional index. The target ion concentration is only read after pairing is completed to avoid the target ion concentration from participating in the determination of its own migration position; the specific processing is as follows: Before the path enters the pairing process, the edge computing node reads the status code, rank bit of the associated quantity, sampling time and record index of the outbound and return paths. After the return path is reversed, the change edge between each adjacent mirror return record reads the status code formed by the same pair of interface records in the original return path, then swaps the negative code and positive code and retains the hold code. The first record of the mirror return path is written with a status code composed of four hold codes, thus forming the mirror return path. S31 establishes a pairing unit for the i-th outbound record and the j-th mirror return record, compares the two status codes bit by bit according to the fields of pH, redox potential, dissolved oxygen and biological activity, writes the number of different fields as the status code digits, and then takes the absolute value of the rank difference of the four accompanying quantities and sums them; the number of outbound path records is denoted as P, the number of mirror return path records is denoted as R, and the absolute value of the value obtained by multiplying i by R and subtracting the value obtained by multiplying j by P is taken to form the relative progress difference; The status code out-of-digit number, the absolute value of the rank difference of the accompanying quantity, and the relative progress difference are written into the unit cost word in sequence. All paired units form rows according to the outgoing sequence and columns according to the mirrored return sequence, forming a paired grid. After the paired grid is formed, the Discrete Fraser matching algorithm performs monotonic recursion along the outgoing sequence and the mirror return sequence; S32 takes the paired unit formed by the intersection of the first outgoing record and the first mirror return record as the first grid, writes the cost word of the first grid unit into the peak field and the cumulative field, and writes the repetition count to zero. The first row of paired units only accepts the left neighbor paired unit, the first column of paired units only accepts the top neighbor paired unit, and the remaining paired units form three candidate results: left neighbor accept, top neighbor accept, and top left neighbor accept. For each candidate result, the preceding peak field is compared with the current unit cost word by field according to the status code out-of-digit number, the absolute value of the rank difference of the accompanying quantity, and the relative progress difference. The last item in the dictionary order is written as the new peak field. Then, the preceding cumulative field is added to the current unit cost word field by field to form a new cumulative field. The left neighbor and the upper neighbor are added to the preceding repetition count by one, and the left upper neighbor is used as the preceding repetition count. The three candidate results are compared sequentially using the new peak field, the new cumulative field, and the duplicate count. If all fields are the same, the first item is read in the order of top left neighbor, left neighbor, and top neighbor. The reading result and the direction of continuation are written into the current paired cell until the last cell completes the recursion, forming the forward cost table. To verify whether each paired cell is located on the same complete discrete Frescher path, the edge computing node calculates the backward cost from the last cell of the paired grid. The last cell cost word is written to both the last cell peak field and the cumulative field in S33, and the repetition count is written to zero. The last row only accepts the right neighbor pairing unit, the last column only accepts the lower neighbor pairing unit, and the remaining pairing units accept the right neighbor, lower neighbor and lower right neighbor pairing units respectively. The peak field, cumulative field and duplicate count comparison rules of the forward cost table are followed. When all fields are the same, the first item is read in the order of lower right neighbor, right neighbor and lower neighbor to form the backward cost table. For any pairing unit, the merged peak field is the last lexicographical item among the forward peak field, the current unit cost word, and the backward peak field. The merged cumulative field is the sum of the forward cumulative field and the backward cumulative field, minus the current unit cost word once. The merged repetition count is the sum of the forward repetition count and the backward repetition count. The peak field, cumulative field, and repetition count in the last cell of the forward cost table form the path cost word of the last cell. Paired cells whose merged results match the path cost word of the last cell field by field are written into the candidate pairing grid. Other paired cells do not participate in index reading. After the candidate pairing grid is generated, the path divergence caused by parallel costs is eliminated by reading in the opposite direction; when S34 reads from the first cell to the last cell, it only reads candidate pairing cells in the right neighbor, the lower neighbor, and the lower right neighbor, and takes the first item in the order of lower right neighbor, right neighbor, and lower neighbor; When reading from the last cell to the first cell, only candidate paired units are read from the left neighbor, the top neighbor, and the top left neighbor. The first item is taken in the order of top left neighbor, left neighbor, and top neighbor. Then the reverse reading results are arranged in reverse order to form two index sequences consisting of the outgoing sequence and the returning sequence. When there is a difference between two index sequences, the outbound and return sequences are compared item by item starting from the first position. The pairing unit with the later lexicographical order in the first difference position is written into the disable flag. Neither the forward nor backward recursion will accept pairing units with the disable flag. Then, the forward cost table, backward cost table and candidate pairing grid are regenerated. When the selected pairing unit belongs to the first or last grid, the difference pairing unit in the other index sequence is disabled instead. If the monotonic continuity path is lost from the first cell to the last cell after disabling, the disabling mark for this round is removed and another difference pairing unit is disabled; if neither disabling result can restore the monotonic continuity path, the continuity path from the last cell to the first cell in the forward cost table is read and a two-way verification inconsistency mark is written. Each round adds a new disable flag, and the number of paired grid cells limits the number of disable rounds; when two index sequences are identical, the outbound record and mirrored return record corresponding to the index sequence are written as a paired sequence; Taking a return segment containing five outbound records and four return records as an example, the mirrored return path first completes the direction flip based on the original return change edge, and the edge computing node establishes twenty pairing units accordingly; the forward recursion gives the acceptance cost from the first cell to the last cell, the backward recursion verifies whether each pairing unit can accept the last cell, the cost is merged to screen out candidate pairing grids, and then the branches with the same cost are eliminated by reading through bidirectional index; The resulting pairing sequence maintains a monotonically advancing forward and backward order, and the corresponding recording position is determined by the changes in pH, redox potential, dissolved oxygen, and biological activity, providing a definite input for subsequent target ion concentration difference deduction and residual response calculation.

[0020] S4. Read the target ion concentration difference according to the pairing sequence. Subtract the first value of the reverse concentration difference in ascending absolute value. Retain the same concentration difference and form the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. Establish a time unfolding network. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and target ion net response sequence. The paired sequence determines the correspondence between the outgoing record and the mirrored return record in the aqueous phase state at a similar solid-liquid interface. Based on this, the concentration changes that are recovered as the aqueous phase state returns are separated, and the concentration changes that are not recovered in the current operating cycle are transferred to the subsequent operating cycles. This implementation first generates the residual response, then establishes a time-unfolded network and allocates cross-cycle cancellation amounts to obtain the target ion net response sequence; the specific processing is as follows: The concentration response in the same period is first matched within each paired record; the edge computing node reads the outgoing record and the mirrored return record along the pairing sequence, and subtracts the concentration of the previous record in each path from the target ion concentration of the current record to form the outgoing concentration difference and the return concentration difference; the first and second records of the two paths and the repeated inherited records are written to the zero concentration difference. When two non-zero concentration differences have opposite signs in S41, the first value of their absolute values ​​in ascending order is taken as the reversible quantity, and the original signs are restored after subtracting them from their absolute values; when the signs are the same, they are retained respectively; when either concentration difference is zero, only the other concentration difference is retained. The non-zero subtraction margin and the retained concentration difference are written as the residual response, and carry the target ion number, packing position code, operation cycle number, record index, source side, response symbol, response quantity and pairing position. The pairing position is jointly identified by the outbound status code, the mirror return status code and the pairing order. Cross-cycle offsetting relationships are established based on source consistency and time sequence; all remaining responses are processed by S42, negative remaining responses establish supply nodes and write supply quantities, and positive remaining responses establish demand nodes and write demand quantities. Under the same target ion number, filler position code and pairing position, the early supply node points to the late demand node to form a canceling edge; The running cycle interval, status code difference, and record index difference are converted into ascending position numbers respectively. The total number of remaining response nodes plus one is used as the carry base. The three ascending position numbers are multiplied by the square of the carry base, the carry base plus one, and then summed to form the offset edge cost. The ascending first value of the absolute value of supply and demand is written as the offset edge capacity. The time-unfolded network adds a net response node. The balance quantities of the supply node, demand node, and net response node are written as the supply quantity, the negative value of the demand quantity, and the total demand minus the total supply quantity, respectively. The net response edge capacity from the supply node to the net response node and from the net response node to the demand node are equal to the corresponding supply and demand, respectively. The edge cost is formed by adding the cube of the carry-in base to the index of the remaining response node. Network traffic establishes an initial base tree that satisfies node balance through net response edges; the flow from supply nodes to net response nodes and from net response nodes to demand nodes are written as supply and demand in S43, respectively, so that the outflow of each node minus the inflow equals the node balance; the potential of the net response node is written as zero, and the potential of the remaining nodes is calculated along the base tree according to the tree edge direction and cost. The reduced cost of non-tree edges in the lower bound is calculated according to the original direction, and the reduced cost of non-tree edges in the upper bound is calculated according to the reverse remaining direction. In each round, the negative first edge is selected in ascending order according to the reduced cost, source node index, target node index and edge index and swapped into the base tree. The swapped edge and the unique path of the original base tree form a cycle. The loop edge in the same direction as the switching direction forms a positive margin by subtracting the flow rate from the capacity, and the reverse edge forms a reverse margin by the current flow rate. The first value of all margins in ascending order is taken as the augmentation amount. The same-direction edge increases the augmentation amount, and the reverse edge decreases the augmentation amount. Edges that reach capacity are denoted as upper bound non-tree edges, and edges whose flow reaches zero are denoted as lower bound non-tree edges. The first edge is swapped out in ascending order of edge index. Swapping in and out is still completed when the augmentation amount is zero. When the base tree edge number sequence is repeated, the next negative cost edge is selected. Before the network operation ends, check the node conservation, edge capacity and cost status; check the inflow, outflow and node balance of each node one by one in S44, check whether the flow of each edge is between zero and edge capacity, and check whether the reduced cost of the lower bound non-tree edge and the reverse reduced cost of the upper bound non-tree edge are both non-negative values. When node flow is not conserved or edge flow exceeds the limit, the current round of augmentation is cancelled and the augmentation amount is recalculated. When the reduction cost calculation is inconsistent, the node potential is recalculated. After all the checks are confirmed to be consistent, the offset flow is written as the cross-cycle offset amount, the remaining flow from the supply node to the net response node is written back to the source running cycle in a negative sign, the remaining flow from the net response node to the demand node is written back to the source running cycle in a positive sign, and arranged according to the running cycle number and record index to form the target ion net response sequence. For example, the same target ion and filler location produces a negative residual response in the first operating cycle, and a positive residual response in the second and third operating cycles at the same paired location. The time-unfolded network allocates flow according to offsetting edge costs and edge capacity. The portion that has been taken over by subsequent positive responses is written as the cross-cycle offset amount, and the negative responses that have not been taken over and the positive responses that have not been taken over by earlier supply are written into the corresponding operating cycles respectively. Therefore, the reversible quantity within the same period, the offset quantity across periods, and the net response of the target ion can all be retrieved back to the solid-liquid interface aqueous phase detection data along the record index.

[0021] S5. Using the return segment as the state node, connect adjacent return segments with the same first state code and consistent paired state positions, and write the target ion net response to form a solid-liquid interface microenvironment evolution model; when the net response is in the same direction and the paired order is not reversed, record the stable migration stage; when the direction is reversed, record the transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, record the abnormal migration stage, and generate ion migration analysis results. The solid-liquid interface microenvironment evolution model is used to organize the net response of target ions dispersed in each operating cycle into a state sequence with a time succession relationship; This implementation method defines state nodes by combining the return segment and the pairing state position, connects homologous state nodes according to the operating cycle, and then divides the migration stage according to the target ion net response direction, pairing order, and cross-cycle cancellation amount, so that ion migration analysis results can be retrieved back to the corresponding return segment and interface record; the relevant processing is carried out as follows: The model mapping first addresses the state assignment when multiple target ion net responses exist within the same return segment. Edge computing nodes read the return segment number, segment head status code, target ion number, filler position code, pairing sequence, target ion net response, and cross-cycle offset amount. An ordered code pair consisting of a state code of an outgoing record and the corresponding mirrored return record status code is recorded as the pairing state position, and the integer order of the pairing unit in the pairing sequence is recorded as the pairing order. When the same return segment contains only one paired state position, a state node is directly established. When the same return segment contains multiple paired state positions, state nodes are established separately according to the paired state positions, and each state node carries the corresponding return segment number. Multiple target ion net responses belonging to the same operating cycle, the same target ion number, the same packing position code, and the same pairing state position are algebraically accumulated. The sign of the accumulated result is written as the target ion net response direction, and the absolute value is written as the response amount. When the accumulated result is zero, a zero response is written. The above node organization and field writing are completed by S51. Subsequently, the state nodes are arranged according to the running cycle number. Directed edges are established only between adjacent state nodes that have the same target ion number, filler position code, segment start state code, and paired state position, forming a solid-liquid interface microenvironment evolution model. When the corresponding state node is missing in the middle running cycle, the directed edge is not established across the missing cycle, and the unaccepted state is written into the model breakpoint record. The initial division of the migration stage along the directed edge is based on the net response direction of the target ion and the pairing order. For the case where the same pairing state position appears repeatedly in a pairing sequence, a pairing order column is formed according to the order of appearance. When comparing across running cycles, the first position is read one by one. If the first order is the same, the next order is compared. When the pairing order of the next state node is equal to or later than that of the previous state node, it is recorded as the pairing order not being reversed. S52 reads the state nodes arranged according to the running cycle, and merges adjacent state nodes with the same net response direction of the target ion and whose pairing order has not been reversed into the same stable migration stage. Subsequent state nodes continue to be merged when they continue to satisfy this relationship. The current stable migration stage ends when the first change in direction or reversal of pairing order occurs. The state nodes from the last state node of the previous stable migration phase to the first state node of the next stable migration phase in the opposite direction are classified into the transition migration phase. When the non-zero response directions before and after a zero-response state node are the same, it is incorporated into the adjacent stable migration stage; when the directions are different, it is classified into the transition migration stage; when there are only one side of adjacent state nodes, it is written into the corresponding stage along with the state node on that side. Anomaly migration identification is performed after the initial partitioning is completed in the stable migration phase and the transitional migration phase; for internal state nodes that have both forward and backward directed edges, S53 compares the net response direction of the target ion with that of the current state node and the previous and next state nodes. When the current direction is different from the preceding and following directions, the preceding and following directions are the same, and the cross-cycle offset amount corresponding to the current state node is zero, the current state node is removed from the original migration stage and written as an abnormal migration stage independently. The abnormal migration phase takes precedence over the stable migration phase and the transitional migration phase. After the current state node is removed, the adjacency relationship of the remaining state nodes before and after is rechecked and the start and end cycle of the phase is corrected. State nodes located at the beginning or end of a directed state sequence lack bilateral comparison objects and therefore do not perform abnormal transition identification. The state nodes on both sides of the model breakpoint are divided into stages to avoid missing running cycles being interpreted as changes in migration direction. After completing the stage boundary correction, the edge computing nodes are arranged into stable migration stage, transition migration stage and abnormal migration stage according to the running cycle number. For each stage, the target ion number, filler position code, pairing status position, stage type, start running cycle number, end running cycle number, target ion net response direction, response amount, cross-cycle offset amount, return segment number and interface record index range are written to generate the ion migration analysis results of target ions in the aqueous phase of the solid-liquid interface. For example, when three adjacent operating cycles form positive, negative and positive net responses of target ions in the same paired state position, and the positive state nodes in front and behind have the same direction, and there is no cross-cycle cancellation amount in the negative state node in the middle, the middle state node is written as the abnormal migration stage. If the intermediate negative state continues to extend and forms a new sequence of states in the same direction, then the direction transition interval is written as the transitional migration stage, and the subsequent negative state nodes form a new stable migration stage. Therefore, the phase results simultaneously retain the microenvironmental state continuity, the source of the net response of the target ions, and the basis for cross-cycle offsetting.

[0022] exist Figure 1 It should be noted that the diagram shows the overall method flow in the order of record sequence formation, return process identification, path matching, net response calculation, and migration stage determination; rectangles represent processing steps, and solid arrows indicate that the output of the previous step directly enters the next step; exist Figure 2 It should be noted that: This diagram illustrates the structural relationship between the data acquisition module, the edge analysis unit, and the result output. The edge analysis unit includes a state segmentation module, a path pairing module, a response allocation module, and a stage determination module; the outer frame indicates the functional affiliation, and the solid arrows indicate data transfer between modules.

[0023] Furthermore, the present invention also includes an ion migration analysis system based on solid-liquid interface microenvironment modeling, the system comprising a data acquisition module, a state segmentation module, a path pairing module, a response allocation module, and a stage discrimination module: The data acquisition module is used to acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. The state segmentation module calculates the accompanying difference between adjacent records based on the record sequence and generates a status code according to the symbol. The records between two consecutive repetitions of the status code form a return segment. The turnaround point is determined according to the absolute value of the rank difference, forming the outgoing path and the reverse return path. The path pairing module is used to establish a distance table using the discrete Fraser matching algorithm. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order, and the pairing sequence is formed by backtracking from the last cell. The response allocation module is used to read the target ion concentration difference according to the pairing sequence. The reverse concentration difference is subtracted by the first value in ascending absolute value order, and the same-direction concentration difference is retained. The remaining amount forms the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. A time unfolding network is established. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and the target ion net response sequence. The stage discrimination module is used to write the target ion net response to form a solid-liquid interface microenvironment evolution model, with the return segment as the state node, connecting adjacent return segments with the same first state code and consistent paired state positions; when the net response is in the same direction and the paired order is not reversed, it is recorded as a stable migration stage; when the direction is reversed, it is recorded as a transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, it is recorded as an abnormal migration stage, generating ion migration analysis results.

[0024] Overall execution and implementation instructions: This solution first uses an intelligent sensing system to collect the target ion concentration, pH, redox potential, dissolved oxygen and biological activity in the aqueous phase at the solid-liquid interface of the biological deodorizing packing material, and then the edge computing nodes form a record sequence according to the sampling time. Subsequently, a status code is generated based on the direction of change of each accompanying quantity in adjacent records. The return segment of the solid-liquid interface microenvironment after leaving the original state is identified, and the outgoing path and the return path are divided. The discrete Fraser matching algorithm determines the corresponding record based on the status code, rank of the accompanying quantity and relative order in the two paths, so that the changes in target ion concentration under similar microenvironment states can be compared with each other. Then, the outbound and inbound concentration differences are deducted to separate the recoverable changes within the same operating cycle. The cross-cycle offset amount and the net response of the target ion are calculated by using a time-unfolding network and a network simplex algorithm. Finally, a solid-liquid interface microenvironment evolution model is established according to the state succession relationship between the return and return segments to identify the stable migration stage, the transitional migration stage, and the abnormal migration stage, thus forming the ion migration analysis results. Taking the operation monitoring of a biological deodorization tower as an example, the intelligent sensing system continuously collects ammonium ion concentration, pH, redox potential, dissolved oxygen, and biological activity at the same packing location. When changes in aeration or microbial metabolism cause the interfacial water phase state to deviate and then recover, the edge computing node identifies a return segment based on the accompanying quantity changes and matches the state departure process with the state return process. If the ammonium ion concentration recovers to its original change range during the return process, the corresponding part is recorded as a reversible quantity in the same period, and the unrecovered part is entered into subsequent operating cycles for further comparison. When a residual response with the opposite direction and corresponding position appears in the next running cycle, the time-unfolded network calculates the cross-cycle cancellation amount between the two, and the uncancelled part is retained as the net response of the target ion. When the net response direction and pairing order of multiple running cycles are consistent, a stable migration stage is formed. The direction switching process forms a transitional migration stage. When there is a local direction change and no cross-cycle cancellation, an abnormal migration stage is formed, thereby judging the continuous migration, state transition and abnormal changes of ammonium ions at the solid-liquid interface.

[0025] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. An ion migration analysis method based on solid-liquid interface microenvironment modeling, characterized in that, include: S1. Acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. S2. Calculate the accompanying difference between adjacent records based on the record sequence and generate a status code according to the symbol. Form a return segment based on the records between two consecutive repetitions of the status code. Determine the turnaround point according to the absolute value of the rank difference to form the outgoing path and the reverse return path. S3. The discrete Fraser matching algorithm establishes a distance table. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order. The pairing sequence is formed by backtracking from the last cell. S4. Read the target ion concentration difference according to the pairing sequence. Subtract the first value of the reverse concentration difference in ascending absolute value. Retain the same concentration difference and form the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. Establish a time unfolding network. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and target ion net response sequence. S5. Using the return segment as the state node, connect adjacent return segments with the same first state code and consistent paired state positions, and write the target ion net response to form a solid-liquid interface microenvironment evolution model; when the net response is in the same direction and the pairing order is not reversed, record the stable migration stage; when the direction is reversed, record the transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, record the abnormal migration stage, and generate ion migration analysis results.

2. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 1, characterized in that: S1 includes: S11. The intelligent sensing system performs water analysis on the solid-liquid interface aqueous phase at the same packing location, obtains the target ion concentration and dissolved oxygen according to the sampling time, and uses electrochemical variables to study or analyze the material to obtain acidity, alkalinity and redox potential. The bioactivity and sampling time of each field obtained simultaneously are written into the original record. S12. Edge computing nodes form electrochemical change segments with adjacent acidity and redox potentials. Linear interpolation is performed according to the time position of each field sampling time in the electrochemical change segment to generate the electrochemical state corresponding to each field sampling time. The electrochemical state corresponding to the target ion concentration sampling time is recorded as the baseline state. S13. For dissolved oxygen and biological activity in the current operating cycle and adjacent operating cycles, calculate the sum of absolute values ​​of the field differences between the corresponding electrochemical state and the baseline state. Take the first item in ascending order according to the sum of absolute values ​​of field differences, sampling time difference and operating cycle. Write the dissolved oxygen, biological activity, target ion concentration and baseline state corresponding to the first item into the same interface record. Arrange the interface records according to the sampling time of the target ion concentration to form a record sequence.

3. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 2, characterized in that: S2 includes: S21. For adjacent interface records in the record sequence, calculate the difference between the next accompanying quantity and the previous accompanying quantity for each field, write the negative code, hold code or positive code according to the sign of the difference, and concatenate the field codes according to the order of the accompanying quantity fields to form the status code corresponding to the interface record. S22. Starting from the first status code, search for the same status code. The first detected same status code and the interface record between them are divided into a return segment. Then, continue searching from the last status code after the last position of the return segment to form a sequence of return segments arranged according to the sampling time.

4. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 3, characterized in that: S2 also includes: S23. In each return segment, assign ranks according to the ascending order of the values ​​of each field of the accompanying quantity. When the field values ​​are the same, assign ranks according to the order of sampling time. Calculate the sum of the absolute values ​​of the field rank differences between each interface record and the first interface record of the return segment. Determine the first interface record as the turnaround point according to the descending order of the absolute values ​​of the field rank differences and the ascending order of the sampling time. S24. The outgoing path is formed by recording the first and second interfaces of the return segment to the turning point, and the return path is formed by recording the reverse arrangement of the first and second interfaces of the return segment to the turning point. The corresponding relationship between the status code, the rank of the accompanying quantity, and the target ion concentration in each path is retained.

5. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 4, characterized in that: S3 includes: S31. The edge computing node reads the rank and status code of the associated quantity of the aqueous phase at the solid-liquid interface in the outgoing and returning paths. It swaps the negative and positive codes in the status code of the returning path and retains the holding code to form a mirrored returning path. Then, for any outgoing record and any mirrored returning record, it calculates the absolute value of the difference in the status code, the absolute value of the difference in the rank of the associated quantity, and the absolute value of the difference in the cross product of the outgoing and returning order in sequence. The node forms the cell cost word according to the field order and writes it into the paired grid. The target ion concentration does not participate in the calculation of the cell cost word. S32. The Discrete Fraser matching algorithm recursively starts from the first cell of the paired grid. The current cell accepts the left neighbor, the upper neighbor, and the upper left neighbor respectively. The peak cost word of the predecessor path and the current cell cost word are sorted in descending order in lexicographical order and written as the new peak cost field. The cumulative field of the predecessor is added to the current cell cost word field by field and written as the new cumulative field. For the left neighbor or the upper neighbor, a repetition count is added. For the upper left neighbor, the repetition count is maintained. The first item is taken in ascending order according to the peak field, cumulative field, repetition count, and upper left neighbor priority over left neighbor and upper neighbor, forming the forward cost table.

6. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 5, characterized in that: S3 also includes: S33. Perform a recursive process from the last cell of the paired grid to the first cell, in the opposite direction to S32, to form a backward cost table. Merge the forward cost, cell cost word and backward cost cell by cell, and retain the paired cells whose merged result is consistent with the path cost word of the last cell to form candidate paired grids. S34. Read pairing units from the last cell to the first cell and from the first cell to the last cell of the candidate pairing grid to form two index sequences. If the two index sequences are inconsistent, compare them bit by bit according to the index field and delete the first difference pairing unit from the index sequence that is later in the sequence. Then, re-execute S32 and S33 until the two index sequences are consistent. Write the outbound record and return record corresponding to the consistent index sequence as a pairing sequence.

7. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 6, characterized in that: S4 includes: S41. Read the target ion concentration of each outgoing record, return record and its previous record according to the pairing sequence. Subtract the concentration of the previous record from the current record concentration to form the outgoing concentration difference and the return concentration difference. When the two concentration differences have opposite signs, take the first value in ascending order of absolute value as the reversible quantity and subtract it from the two concentration differences. When the signs are the same, retain the two concentration differences. Write the subtracted balance and the retained concentration difference as the residual response with the running cycle, packing position, pairing position, response sign and response quantity. S42. Establish supply nodes with negative residual responses and demand nodes with positive residual responses. Only under the same target ion, the same packing position, and the same pairing position, establish offset edges from the earlier supply node to the later demand node. Use the total number of residual response nodes plus one as the carry base. Multiply the operating cycle interval, the number of status codes, and the difference in record index by the square of the carry base, the carry base, and one, and then add them to form the offset edge cost. Take the first value in ascending order of the absolute values ​​of supply and demand as the offset edge capacity, and add net response nodes. Write the total demand minus the total supply as the net response node balance. Use the cube of the carry base plus the index of the residual response node to form the net response edge cost, and establish a time-unfolded network.

8. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 7, characterized in that: S4 also includes: S43. The edge computing node adopts the network simplex algorithm. The initial base tree is established by the net response edges of each supply node pointing to the net response node and the net response edges of the net response node pointing to each demand node. The initial flow is written according to the balance of the net response node and the node potential is calculated based on the tree edge cost. The non-tree edge reduction cost is calculated round by round. The negative first edge is selected in ascending order according to the reduction cost, source node index and target node index and replaced into the base tree. The first value of the loop formed by the replaced edge is taken in ascending order according to the remaining capacity to perform augmentation. Then, the first edge of the loop edge index that the flow reaches the capacity is taken in ascending order and replaced out of the base tree until the non-tree edge reduction cost is non-negative. S44. Based on the time, unfold the network and check the inflow, outflow and net response node balance of each node, and check the flow and edge capacity of each edge. If any check result is inconsistent, recalculate the potential of the corresponding node and continue to execute S43. When all check results are consistent, write the offset edge flow as the cross-cycle offset amount, and arrange the net response edge flow according to the running cycle and record index to form the target ion net response sequence.

9. The ion migration analysis method based on solid-liquid interface microenvironment modeling according to claim 8, characterized in that: S5 includes: S51. Read the segment head status code, pairing status position, target ion net response and cross-cycle cancellation amount corresponding to each return segment, establish a status node with the return segment, establish a directed edge between adjacent status nodes with the same segment head status code and the same pairing status position, and write the target ion net response direction, response amount and pairing order into the status node to form a solid-liquid interface microenvironment evolution model. S52. Compare the target ion net response direction and pairing order of adjacent state nodes along the directed edge. Merge state nodes with the same target ion net response direction and whose pairing order is not earlier than that of the previous state node into a stable migration stage. Write the first state node interval from the change of target ion net response direction to the new direction as the transition migration stage. S53. For any internal state node, when the current target ion net response direction is different from the previous and subsequent state nodes, the previous and subsequent state nodes have the same direction, and the current cross-cycle cancellation amount is zero, the current state node is separated from the corresponding migration stage and written as the abnormal migration stage. Then, the stable migration stage, transitional migration stage, and abnormal migration stage are arranged in the order of the running cycle to generate the ion migration analysis results of the target ion in the aqueous phase at the solid-liquid interface.

10. An ion migration analysis system based on solid-liquid interface microenvironment modeling, used to implement the ion migration analysis method based on solid-liquid interface microenvironment modeling as described in any one of claims 1-9, the system comprising a data acquisition module, a state segmentation module, a path pairing module, a response allocation module, and a stage discrimination module, characterized in that: The data acquisition module is used to acquire multi-cycle data of the solid-liquid interface aqueous phase of biological deodorization. The intelligent sensing system performs water analysis to obtain the target ion concentration and uses electrochemical variables to study or analyze materials to obtain acidity, alkalinity, and redox potential. Simultaneously, dissolved oxygen and biological activity are obtained. Edge computing nodes form a record sequence according to the sampling time. The state segmentation module calculates the accompanying difference between adjacent records based on the record sequence and generates a status code according to the symbol. The records between two consecutive repetitions of the status code form a return segment. The turnaround point is determined according to the absolute value of the rank difference, forming the outgoing path and the reverse return path. The path pairing module is used to establish a distance table using the discrete Fraser matching algorithm. The current cost is taken as the descending first value of the current distance and the first value of the predecessor cost in ascending order, and the pairing sequence is formed by backtracking from the last cell. The response allocation module is used to read the target ion concentration difference according to the pairing sequence. The reverse concentration difference is subtracted by the first value in ascending absolute value order, and the same-direction concentration difference is retained. The remaining amount forms the residual response. The early residual response of the same target ion, filler position and pairing state position points to the late reverse residual response. A time unfolding network is established. The edge computing nodes use the network simplex algorithm to perform negative reduction cost edge replacement, loop augmentation and saturation edge replacement to form the cross-period offset amount and the target ion net response sequence. The stage discrimination module is used to write the target ion net response to form a solid-liquid interface microenvironment evolution model, with the return segment as the state node, connecting adjacent return segments with the same first state code and consistent paired state positions; when the net response is in the same direction and the paired order is not reversed, it is recorded as a stable migration stage; when the direction is reversed, it is recorded as a transitional migration stage; when it is different from the preceding and following nodes and there is no cross-cycle cancellation amount, it is recorded as an abnormal migration stage, generating ion migration analysis results.

Images