A digital management and evaluation system for the treatment of soil acidification in a honey pomelo garden

Abstract

The application discloses a kind of honey pomelo garden soil acidification treatment digital management and evaluation system, specifically related to orchard soil treatment digital management technical field, including the soil pH value of each sampling position, exchangeable acid amount, soil layer identification, sampling time, change record, irrigation and drainage record and rainfall record, output original treatment data;Receive original treatment data, execute position binding, layer position binding, time sorting and history storage, output continuous treatment data;Receive continuous treatment data, the detection record of the same tree tray position, the same soil layer, the same treatment round is combined with change record, generates and actually treats the digital twin site of one-to-one mapping of position, output digital twin site set;By mapping treatment position as digital twin site, and to the favorable change detected successively executes interlayer continuity, change range consistency, boundary external interference exclusion and subsequent round persistence review, to distinguish real improvement and phase appearance improvement.

Description

Technical Field

[0001] This invention relates to the field of digital management technology for orchard soil remediation, and more specifically, to a digital management and evaluation system for soil acidification remediation in pomelo orchards. Background Technology

[0002] In the practice of information management for orchard soil acidification control, existing solutions mainly focus on how to determine the effectiveness of control based on the results of phased testing, and use the results of this judgment to make subsequent investment arrangements, adjust the order of control, and determine the end of control. In engineering, soil pH value, organic matter content, application records, irrigation and drainage records, and data from previous retests are generally collected by plot, zone, or management unit. The difference, direction of change, and magnitude of change between two or more consecutive test results are then used as the basis for the control effect.

[0003] Taking the scenario of continuous soil acidification control in a mountain pomelo orchard as an example, there are obvious differences in water and fertilizer migration between different slopes in the orchard. The responses inside and outside the tree basin, and between the surface layer and the main root distribution layer are not consistent. At the same time, it is not possible to repeatedly conduct high-density deep sampling to improve the accuracy of the judgment. It is also necessary to comply with the fruit harvesting schedule, rainfall process, irrigation rhythm and existing fertilization cycle. Under such conditions, the existing treatment method is prone to the following phenomenon: some areas have shown a decrease in acidification index, a local pH increase or a short-term trend improvement in one or several retests. Based on this, the system lowers the treatment level, reduces subsequent investment, or even removes it from the key treatment area. However, after entering the subsequent management cycle, the area will again show a decline in acidification, failure to improve the deep root zone synchronously, or reintroduction of surrounding influences.

[0004] The root of this problem is that the existing assessment criteria usually assume that as long as the test results show a positive change, the change can be directly written into the subsequent management chain as a governance achievement, without further distinguishing whether the positive change is a real improvement that can continuously support subsequent decisions, or just a temporary apparent improvement brought about by superficial short-term relief, changes in local sampling performance, or disturbances in the adjacent area.

[0005] The technical problem this application aims to solve is: how to effectively identify positive changes detected during the digital management and evaluation of soil acidification control in pomelo orchards, and avoid misjudging temporary apparent improvements that lack sustained decision-making significance as genuine control effectiveness. Summary of the Invention

[0006] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a digital management and evaluation system for the treatment of soil acidification in pomelo orchards. By mapping the treatment location to a digital twin, and sequentially performing interlayer continuity, consistency of treatment scope, exclusion of interference outside the boundary, and continuous review of subsequent rounds on the detected positive changes, the system distinguishes between real improvement and staged apparent improvement, thereby solving the problems mentioned in the background art.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a digital management and evaluation system for soil acidification control in pomelo orchards, comprising:

[0008] The data acquisition terminal is used to collect soil pH value, exchangeable acidity, soil layer identification, sampling time, treatment records, irrigation and drainage records, and rainfall records at each sampling location, and output raw treatment data.

[0009] The processing host is used to receive raw governance data, perform location binding, hierarchical binding, time sorting and historical storage, and output continuous governance data.

[0010] The digital twin generation module is used to receive continuous governance data, combine the detection records and construction records of the same tree basin location, the same soil layer, and the same governance round, generate digital twin bits that are mapped one-to-one with the actual governance location, and output the digital twin bit set;

[0011] The improvement extraction module is used to receive the digital twin bit set, perform a same-position comparison between the current detection record of the same digital twin bit and the detection record of the previous treatment round, and determine the existence of candidate improvement when the current pH value is higher than that of the previous treatment round and the current exchangeable acidity is lower than that of the previous treatment round, and output the candidate improvement set.

[0012] The qualification review module is used to receive the candidate improvement set and the corresponding improvement records, perform continuous verification of adjacent soil layers, same-direction verification within the improvement range, reverse exclusion verification outside the improvement range, and continuous verification in the next retesting round. Candidate improvements that pass all verifications are determined as effective improvements, and the rest are determined as apparent improvements. The improvement review results are then output.

[0013] In a preferred embodiment, it further includes:

[0014] The management and evaluation module receives improvement review results and continuous governance data, writes effective improvements into the digital twin effectiveness record and performs governance sequence updates, input arrangement updates and exit determination generation, writes apparent improvements into the observation record and performs supplementary retesting arrangements and subsequent review arrangements, and outputs governance management results and evaluation results.

[0015] In a preferred embodiment, the data acquisition terminal includes:

[0016] Based on the tree basin location marker, soil layer marker, and sampling time corresponding to the current sampling location, soil pH and exchangeable acidity are bound at the same location, soil layer, and time, and the test records are output.

[0017] For the current sampling location, based on the application time, application method, application coverage boundary, irrigation and drainage records, and rainfall records, perform location-based treatment association on the detection records and output the treatment association records;

[0018] Call the detection record and governance association record that are the same as the current sampling position in the previous governance round, perform pairing between the current detection record and the corresponding record in the previous governance round, and output the continuous sampling record.

[0019] In a preferred embodiment, the processing host includes:

[0020] Based on the tree basin location markers, soil layer markers, and sampling times in the original treatment data, soil pH, exchangeable acidity, treatment records, irrigation and drainage records, and rainfall records are merged according to the same location and the same stratum, and the corresponding records of the location and stratum are output.

[0021] Based on the sampling time, application time, irrigation and drainage time and rainfall occurrence time in the location-stratum corresponding records, the records are sorted and arranged in rounds to output the treatment time sequence record;

[0022] The system retrieves historical records corresponding to the same tree basin location and soil layer from the governance timeline records, performs sequential storage between the current cycle record and the historical cycle records, and outputs continuous governance data.

[0023] In a preferred embodiment, the digital twin generation module includes:

[0024] Receive detection records and remediation records from continuous treatment data, perform same-value screening according to tree basin location identifier, soil layer identifier, and treatment round, construct a correspondence graph with detection records as nodes on one side and remediation records as nodes on the other side, calculate a corresponding cost quadruple consisting of location deviation, layer deviation, time interval, and coverage of the remediation effect range for each detection record and each remediation record, retain feasible record pairs under the constraints of the same tree basin location identifier, the same soil layer identifier, the same treatment round, and the remediation time being earlier than the sampling time, and output the initial correspondence graph.

[0025] In a preferred embodiment, the digital twin generation module further includes:

[0026] For each feasible record pair in the initial correspondence map, the consistency of soil pH and exchangeable acidity in the detection records with the application method, application coverage boundary, and application time in the application records is checked. A consistency result group is formed, consisting of consistent results in the direction of acidity change, consistent results in the direction of pH change, and consistent results in the application coverage. The consistency result group is written back to the corresponding cost quadruple and confidence updates are performed round by round. When the same detection record is connected to multiple application records, the record pair with the first position in the updated cost quadruple is retained in the order of position deviation, layer deviation, time interval, and coverage gap. When the same application record is connected to multiple detection records, the record pair with continuous tree basin position, continuous soil layer, and the most recent sampling time is retained. Confidence updates and conflict resolution are performed cyclically until the record pairs retained in the previous and next rounds are completely consistent, and a stable correspondence map is output.

[0027] Based on each pair of retained records in the stable correspondence diagram, the corresponding detection record and the implementation record execution field are concatenated to generate a digital twin bit including tree basin location identifier, soil layer identifier, treatment round, detection record identifier, implementation record identifier, corresponding cost quadruple and consistency result group. All digital twin bits are encoded in the execution order of tree basin location identifier, soil layer identifier and treatment round, and a set of digital twin bits that are mapped one-to-one with the actual treatment location is output.

[0028] In a preferred embodiment, the improved extraction module includes:

[0029] Receive the digital twin bit of the current treatment round and the digital twin bit of the previous treatment round from the digital twin bit set, perform same-position pairing according to tree basin location identifier, soil layer identifier and treatment record identifier, calculate the pH difference value and exchangeable acidity difference value for the current detection record and the previous treatment round detection record respectively, and output the improvement difference value record.

[0030] For each digital twin bit in the improvement difference record, call the digital twin bits of adjacent soil layers at the same tree basin location and the digital twin bits of adjacent tree basin locations in the same treatment round, and perform inter-layer same-direction verification and adjacent reverse investigation on the pH difference and exchangeable acidity difference. When the pH difference of the current digital twin bit is positive, the exchangeable acidity difference is negative, the adjacent soil layers have the same-direction change and the adjacent tree basin locations do not form reverse traction, it is determined that there is a candidate improvement for the digital twin bit, and the candidate improvement set is output.

[0031] In a preferred embodiment, the qualification review module includes:

[0032] Receive the candidate improvement set and corresponding improvement records. According to the tree basin location identifier, soil layer identifier, and treatment round, call the adjacent soil layer digital twins, digital twins within the improvement coverage boundary, and adjacent digital twins outside the improvement coverage boundary of the current digital twin. Calculate the interlayer continuation difference sequence between the current digital twin and the adjacent soil layer digital twin, the intra-boundary same-direction sequence with the intra-boundary digital twin, and the extra-boundary reverse sequence with the extra-boundary adjacent digital twin. Under the conditions that the improvement time is earlier than the sampling time, the improvement coverage boundary includes the current tree basin location, and the improvement method corresponds to the current soil layer, combine the interlayer continuation difference sequence, the intra-boundary same-direction sequence, and the extra-boundary reverse sequence into the initial review record.

[0033] In a preferred embodiment, the qualification review module further includes:

[0034] For the initial review record, the digital twin of the retest corresponding to the current candidate improvement in the next retest round is called. The round continuation sequence and round back-off sequence between the current candidate improvement and the next retest round are obtained. The round continuation sequence and round back-off sequence are written back to the corresponding initial review record to form the retest review record. Then, each candidate improvement is checked in four rounds in the order of layer continuation, same direction within the boundary, reverse direction outside the boundary, and round continuation. When a round of check is not valid, the candidate improvement is directly recorded as an apparent improvement candidate. When all four rounds of check are valid, the candidate improvement is recorded as a valid improvement candidate, and a stable review record is output.

[0035] Based on the stability review records, candidate improvements that simultaneously meet the following criteria are identified as effective improvements: continuity between adjacent soil layers, same-direction improvement within the improvement area, reverse exclusion improvement outside the improvement area, and continued improvement in the next retesting round. The remaining candidate improvements are identified as apparent improvements. The corresponding interlayer continuity difference sequence, same-direction sequence within the boundary, reverse sequence outside the boundary, round continuation sequence, round regression sequence, and review conclusion are combined and written into the improvement review results.

[0036] In a preferred embodiment, the management evaluation module includes:

[0037] Receive the improvement review results and continuous governance data, call the tree basin location identifier, soil layer identifier, governance round, implementation record and review conclusion corresponding to each digital twin, write the digital twin with the review conclusion of effective improvement into the digital twin effectiveness record, write the digital twin with the review conclusion of apparent improvement into the observation record, and output the classification record set;

[0038] Based on the classification record set, the system retrieves the digital twin effectiveness records, observation records, and continuous management data of all digital twins within the same management round. For each tree basin location, it performs item-by-item statistics based on the number of digital twins corresponding to effective improvement, the number of digital twins corresponding to apparent improvement, and the number of digital twins corresponding to no improvement. For each soil layer, it assigns corresponding markers based on the continuous distribution of effective improvement and the recurrence of apparent improvement. Finally, it generates management sequence update results, input arrangement update results, supplementary retesting arrangements, and subsequent verification arrangements in the order of first zero effective improvement, then non-zero apparent improvement, and finally non-zero non-improvement.

[0039] For the same tree basin location and the same soil layer, the digital twin records of effective and observational data from two consecutive treatment cycles are retrieved. When effective improvement is recorded in both consecutive treatment cycles but no apparent improvement is recorded, an exit judgment is generated. When apparent improvement is recorded in the current treatment cycle, the original treatment status is maintained and supplementary retesting and subsequent review arrangements are written. The treatment sequence update results, input arrangement update results, exit judgment, supplementary retesting arrangements, and subsequent review arrangements are combined and written into the treatment management results and evaluation results.

[0040] The technical effects and advantages of this invention are as follows:

[0041] 1. This plan separates the phased apparent improvement from the real improvement by sequentially performing continuous verification of adjacent soil layers, same-direction verification within the scope of the improvement, reverse exclusion verification outside the scope of the improvement, and continuous verification in the next retesting round. This helps to reduce the situation of mistakenly writing short-term positive changes into the governance effectiveness chain.

[0042] 2. Generate digital twins between the test records and the treatment records according to the tree basin location, soil layer, and treatment cycle, so that the test results and the actual treatment location form a stable correspondence, which helps to improve the consistency of subsequent co-location comparison, boundary verification, and cycle continuation;

[0043] 3. In the improvement extraction stage, digital twins of adjacent soil layers and adjacent tree basins are introduced to perform inter-layer same-direction verification and adjacent reverse investigation on single-point positive changes, which helps to suppress the direct impact of local sampling fluctuations or neighboring disturbances on improvement judgment.

[0044] 4. Effective improvements will be written into the digital twin effectiveness record, and apparent improvements will be written into the observation record. These will trigger updates to the governance sequence, investment arrangements, supplementary retesting arrangements, and subsequent review arrangements, thereby helping to improve the correspondence between subsequent management actions and the nature of the improvements.

[0045] 5. By combining the digital twin records of effectiveness and observation records of two consecutive treatment cycles for the same tree basin location and the same soil layer, the exit conditions can be determined, which helps to mitigate the risk of subsequent acidification decline caused by prematurely exiting treatment based solely on the positive results of a single cycle. Attached Figure Description

[0046] Figure 1 This is a schematic diagram of the system module structure of the present invention. Detailed Implementation

[0047] 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.

[0048] Refer to the instruction manual appendix Figure 1 The present invention provides a digital management and evaluation system for soil acidification control in pomelo orchards, comprising:

[0049] The data acquisition terminal is used to collect soil pH value, exchangeable acidity, soil layer identification, sampling time, treatment records, irrigation and drainage records, and rainfall records at each sampling location, and output raw treatment data.

[0050] This implementation method illustrates how the data acquisition terminal organizes the scattered detection and treatment information collected on-site into continuous sampling records that can be directly continued by the subsequent processing host. The purpose of this implementation process is to first fix the detection value formed at the same sampling location at the current moment to a unique tree basin location and a unique soil layer, then incorporate the treatment background and environmental background corresponding to the sampling location into the detection value, and finally establish a sequential relationship between the current sampling result and the corresponding result of the previous treatment round, thereby providing consistent input for subsequent continuous treatment data generation, digital twin generation, and improvement extraction. This implementation process includes the following steps:

[0051] At the start of the current sampling task, the acquisition terminal first reads the sampling location number, tree basin location identifier, soil layer identifier, and sampling time from the sampling work order, and simultaneously receives the soil pH value and exchangeable acidity (EAA) sampling results. The tree basin location identifier indicates the pomelo tree basin management unit to which the sampling location belongs, the soil layer identifier indicates the soil layer to which the sampling depth belongs, and the sampling time is written using the actual sampling completion time. The soil pH value and AEA are obtained from the same sampling location and the same sampled soil sample. The acquisition terminal performs same-location, same-soil-layer, and same-moment binding on the tree basin location identifier, soil layer identifier, sampling time, soil pH value, and AEA, generating a detection record and writing it to the current sampling buffer for subsequent treatment association calls. When the sampling location number lacks a tree basin location identifier, the acquisition terminal calls the location mapping table to look up the tree basin location identifier according to the sampling location number before performing the binding. When the soil layer identifier is missing, the sampling result is recorded as a missing layer record, and the subsequent before-and-after pairing process is paused.

[0052] After the detection record is generated, the data acquisition terminal continues to read the application time, application method, application coverage boundary, irrigation and drainage records, and rainfall records corresponding to the current sampling location, and uses the detection record as the basic record for governance association. Specifically, the application time is the completion time of the most recent preceding application action at the sampling location; the application method is the application category corresponding to the application action; the application coverage boundary is the boundary of the registered effective area of ​​the application action; the irrigation and drainage records are the irrigation and drainage events at the corresponding tree basin location between the application time and the sampling time; and the rainfall records are the rainfall events in the corresponding orchard area within the same time period. The data acquisition terminal first determines the current sampling location based on the tree basin location identifier. The system checks whether the location falls within the coverage boundary of the modification. Then, based on the chronological relationship between the modification time and the sampling time, it determines whether the modification action is a preceding governance action to the current sampling result. When the coverage relationship is established and the modification time is earlier than the sampling time, the modification time, modification method, modification coverage boundary, irrigation and drainage records, and rainfall records are written into the detection record to generate a governance association record. This governance association record is then written into the governance association cache for subsequent rounds of pairing and retrieval. When no preceding modification action covering the current sampling location is found, the acquisition terminal retains the detection record and writes a "no modification" flag into the governance association record, so that the subsequent processing host can identify and observe the round data accordingly.

[0053] After the governance association record is formed, the acquisition terminal calls the detection record and governance association record of the same sampling location in the previous governance round and performs before-and-after pairing on the current governance association record. Specifically, the acquisition terminal uses the same tree basin location identifier, the same soil layer identifier and adjacent governance rounds as pairing conditions. First, it retrieves the corresponding detection record of the previous governance round, then retrieves the governance association record corresponding to the detection record, and forms a detection before-and-after pair with the detection record of the previous governance round. It also forms a governance before-and-after pair with the governance association record of the previous governance round. The detection before-and-after pair and the governance before-and-after pair are combined into a continuous sampling record and written to the continuous sampling storage area for the processing host to read. When there are multiple records that meet the conditions in the previous governance round, the acquisition terminal retains the record closest to the current governance round as the unique preceding record according to the sampling time from the nearest to the furthest. When there are no records that meet the conditions in the previous governance round, the acquisition terminal writes the first round sampling identifier into the continuous sampling record and outputs the current governance association record separately, so as to ensure that subsequent steps can still be executed according to the unified field structure.

[0054] Through the above processing, the output of the acquisition terminal is no longer a scattered sample value, but a continuous sampling record that has completed location attribution, soil layer attribution, treatment background writing, and cycle continuation. The subsequent processing host can directly continue to perform location binding, layer binding, time sorting, and historical storage according to the tree basin location identifier, soil layer identifier, and treatment cycle, without needing to re-trace the original sampling relationship on site. At the same time, since the treatment time, treatment method, treatment coverage boundary, irrigation and drainage records, and rainfall records have been associated with the treatment location during the acquisition stage, the corresponding treatment background fields can be directly read during subsequent digital twin generation and qualification review calls, reducing intermediate conversion ambiguity.

[0055] In practical applications: Taking the collection of surface soil samples from a tree basin in a mountain pomelo orchard as an example, the collection terminal first binds the surface pH value and exchangeable acidity of the tree basin with the current sampling time to form a detection record. Then, it reads the application time of the previous round of conditioner application, the application coverage boundary, and the irrigation, drainage, and rainfall events from the application to the sampling to form a treatment association record. Subsequently, it retrieves the detection records and treatment association records of the same tree basin and the same surface soil layer in the previous treatment round to complete the pairing. Finally, it outputs continuous sampling records, which can be directly connected to the continuous treatment data by the subsequent processing host.

[0056] The processing host is used to receive raw governance data, perform location binding, hierarchical binding, time sorting and historical storage, and output continuous governance data.

[0057] This implementation method illustrates how the processing host organizes the raw governance data output from the acquisition terminal into continuous governance data that can directly support the generation and improvement review of digital twins. The core of this implementation process lies in first merging the detection information, remediation information, and environmental information corresponding to the same tree basin location and soil layer into a unified field structure; then forming governance cycles according to the order of governance actions and retesting; and finally establishing a sequential relationship between the current cycle record and historical cycle records, thereby providing stable input for subsequent peer comparison, boundary verification, and continuous cycle verification. This implementation process includes the following steps:

[0058] The processing host first performs location and stratification merging on the original treatment data to form a unified input record for subsequent rounds of arrangement. Specifically, the processing host receives detection records, treatment association records, and continuous sampling records from the original treatment data, reads the tree basin location identifier, soil layer identifier, sampling time, soil pH value, exchange acidity, treatment records, irrigation and drainage records, and rainfall records, and uses the tree basin location identifier as the location merging key and the soil layer identifier as the stratification merging key to merge records with the same tree basin location identifier and the same soil layer identifier into the same location and stratification record group.

[0059] During the merging process, soil pH and exchangeable acidity are written as detection fields into the location-stratum record group. The application time, application method, application coverage boundary, and target soil layer in the application record are written as treatment fields into the location-stratum record group. The event time and event type in the irrigation and drainage records, as well as the occurrence time and rainfall in the rainfall records, are written as environmental fields into the location-stratum record group. This forms a location-stratum corresponding record and writes it into the current location-stratum buffer for round-by-round allocation and retrieval. When there are multiple detection records with the same sampling time in the same tree basin location and under the same soil layer, the processing host performs deduplication according to the detection record identifier and retains one record. When the tree basin location identifier or soil layer identifier is missing, the corresponding record is written into the missing field record table and does not enter the location-stratum corresponding record generation process.

[0060] After the processing host forms the location-layer corresponding record, it performs time expansion and round arrangement on the treatment and environmental fields within the same location-layer record group to form a treatment time sequence record with a clear order between each round. Specifically, the processing host calls the sampling time, modification time, irrigation and drainage time and rainfall occurrence time in the location-layer corresponding record, first performs unified sorting according to time order to form a time sequence under the same tree basin location and the same soil layer, and then uses the modification time as the starting point of the treatment round, arranges the sampling records after a certain modification time and before the next modification time as the retest records in the same treatment round, and arranges the irrigation and drainage records and rainfall records within the time interval as the environmental accompanying records of the treatment round.

[0061] When there is no new treatment time for the same tree basin location and soil layer between adjacent sampling times, the processing host merges the subsequent sampling records into the previous treatment round. When there is no treatment time before the current sampling record, the processing host assigns the sampling record as an observation round record and writes the observation round identifier, thereby obtaining a treatment time sequence record containing tree basin location identifier, soil layer identifier, treatment round, round start treatment time, sampling time sequence within the round, irrigation and drainage time sequence within the round, and rainfall time sequence within the round. The treatment time sequence record is then written to the round time sequence storage area for subsequent use. When there are multiple treatment records at the same time, the processing host assigns them according to the order of completion time registered in the treatment records. If the completion times are the same, they are assigned according to the order in which the records were written, without introducing other sorting conditions.

[0062] After obtaining the governance timeline records, the processing host performs sequential storage of the current round record and the historical round records to form continuous governance data that can be directly read by the subsequent digital twin generation module. Specifically, the processing host calls the tree basin location identifier and soil layer identifier corresponding to each current round record in the governance timeline records, retrieves the historical round records with the same tree basin location, the same soil layer and governance round before the current round in the historical storage area, and sequentially appends the current round record to the most recent preceding historical round record according to the order of governance rounds to form the round chain under the current position layer.

[0063] During the continuation process, the processing host writes the detection field, governance field, and environment field from the current round record into the current node of the round chain. At the same time, it writes the round identifier of the most recent preceding historical round record into the preceding round field of the current node, and writes the current round identifier into the subsequent round field of the most recent preceding historical round record, thereby forming continuous governance data with bidirectional correlation between the preceding and following rounds and writing it into the continuous governance data area for the digital twin generation module to read. When there is no preceding historical round record with the same tree plate location and the same soil layer in the historical storage area, the processing host records the current round record as the first round node and writes it separately into the continuous governance data area. When the historical round record chain is interrupted, the processing host writes a breakpoint identifier at the interruption position and keeps the current round record continuing to enter the chain to ensure that subsequent steps can identify the first round node and the breakpoint node.

[0064] Through the above processing, the continuous governance data output by the processing host has achieved unified location, unified layer, unified cycle, and historical continuity. The subsequent digital twin generation module does not need to trace back the original sampling records and governance-related records again. It can directly extract the corresponding digital twin bits of the detection records and the modification records according to the tree basin location identifier, soil layer identifier, and governance cycle to generate digital twin bits. At the same time, the cycle time sequence relationship has been fixed in the continuous governance data. When the subsequent improvement extraction module calls the previous governance cycle record and the qualification review module calls the next retest cycle record, it can directly retrieve the data by unified fields, avoiding the problems of inconsistent cycle values ​​and unclear correspondence between cycles.

[0065] In practical applications: Taking the topsoil layer of the same pomelo tree basin as an example, the processing host first merges the soil pH value, exchangeable acidity, application records, irrigation and drainage records, and rainfall records of the topsoil layer of the tree basin into corresponding records for the same location layer. Then, taking the completion time of a certain application of conditioner as the starting point of the treatment cycle, the host matches the two subsequent retest records and the corresponding irrigation, drainage, and rainfall events into the same treatment cycle. Subsequently, the topsoil records of the previous treatment cycle are called to establish the successor-follower relationship, and finally, continuous treatment data of the topsoil layer of the tree basin is formed, which can be continuously called for subsequent digital twin generation, candidate improvement extraction, and improvement qualification review.

[0066] The digital twin generation module is used to receive continuous governance data, combine the detection records and construction records of the same tree basin location, the same soil layer, and the same governance round, generate digital twin bits that are mapped one-to-one with the actual governance location, and output the digital twin bit set;

[0067] This implementation method illustrates how the digital twin generation module establishes a unique, traceable, and directly accessible digital twin for each actual governance location within continuous governance data. The core of this process is not simply storing detection and remediation records side-by-side, but rather first identifying corresponding record pairs under constraints of the same tree basin location, soil layer, and governance cycle. Then, based on the consistency between the detection results and the remediation background, the record pairs are iteratively verified and uniquely retained. Finally, the retained detection and remediation records are concatenated into a complete digital twin, ensuring that subsequent location comparisons, boundary checks, and cycle-based continuous checks all correspond to a unique actual governance location. This implementation process includes the following steps:

[0068] The digital twin generation module first performs same-value screening and corresponding mapping on the detection records and remediation records in the continuous governance data to determine the feasible record pairs that can enter the digital twin mapping in the current governance round. Specifically, the digital twin generation module reads the detection record identifier, remediation record identifier, tree basin location identifier, soil layer identifier, governance round, sampling time, remediation time, and remediation coverage boundary from the continuous governance data, and constructs an initial candidate node set with the detection record as the detection side node and the remediation record as the remediation side node. Subsequently, the digital twin generation module connects the corresponding detection side node and remediation side node as candidate record pairs only under the conditions that the tree basin location identifier, soil layer identifier, governance round, and remediation time are the same and the remediation time is earlier than the sampling time. For each candidate record pair, the corresponding cost quadruple is calculated. The position deviation is the difference between the tree basin location where the detection record is located and the tree basin location where the remediation record is applied. When the tree basin locations are completely consistent, the position deviation is recorded as zero. When the tree basin locations are inconsistent, the candidate record pair is not retained.

[0069] The layer deviation is measured by the difference between the soil layer to which the detection record belongs and the target soil layer of the modification record. When the soil layers are consistent, the layer deviation is recorded as zero. When the soil layers are inconsistent, the candidate record pair is not retained. The time interval is measured by the time difference between the sampling time and the modification time. The modification coverage is measured by the coverage result of the modification coverage boundary on the current tree basin position. When the current tree basin position falls within the modification coverage boundary, the modification coverage is recorded as coverage is successful. Otherwise, the candidate record pair is not retained. After completing the above calculations, the digital twin generation module writes the candidate record pairs that have not been eliminated into the correspondence graph, forms the initial correspondence graph, and writes it into the initial correspondence graph storage area for subsequent action consistency verification. When no modification record that meets the conditions is found for the same detection record in the current governance round, the digital twin generation module writes the detection record into the unmapped record table and does not enter the digital twin generation process of this round.

[0070] After the initial correspondence map is formed, the digital twin generation module performs consistency verification and confidence updates for each feasible record pair to eliminate mapping conflicts caused by one detection record corresponding to multiple treatment records or one treatment record corresponding to multiple detection records. Specifically, the digital twin generation module reads the soil pH, exchangeable acidity, treatment method, treatment coverage boundary, and treatment time from each feasible record pair, and determines the expected direction of action based on the treatment method. When the treatment method is acidification treatment, the expected direction of exchangeable acidity is decreasing, and the expected direction of soil pH is increasing. Subsequently, the digital twin generation module compares the current detection record with the detection records of the same tree basin location and the same soil layer in the previous treatment round. When the direction of change of exchangeable acidity is consistent with the expected direction, a result of consistent acidity change direction is generated. When the direction of change of soil pH is consistent with the expected direction, a result of consistent pH change direction is generated. When the current tree basin location is still within the treatment coverage boundary and the sampling time is after the treatment time, a result of successful treatment coverage is generated, and the three results are combined into a consistency result group.

[0071] The digital twin generation module then writes the consistency result group back to the corresponding cost quadruple to perform round-by-round confidence updates. The execution method of round-by-round confidence updates is as follows: For multiple modification records connected to the same detection record, the record pairs are retained first by the position deviation, then by the layer deviation, then by the time interval, and then by the coverage of the modification effect area. If there is a difference in the previous item, the subsequent item will not participate in the current round of retention. For multiple detection records connected to the same modification record, the record pairs with tree basin positions that are continuous tree basin positions within the same modification coverage boundary are retained first, and the record pairs with soil layers that are the modification target soil layers and their directly adjacent soil layers are retained. If there are still multiple record pairs, the record pair with the shorter interval between the sampling time and the modification time is retained.

[0072] After completing the current round of retention, the digital twin generation module compares the retention results of the current round with those of the previous round one by one. When the retained detection record identifier and the modification record identifier are completely consistent, the correspondence is determined to be stable and the update stops. When there is an inconsistency, the retention result of the current round is used as the input for the next round to re-execute the consistency result group write-back and conflict resolution until the retained record pairs of the two rounds are completely consistent. The stable correspondence graph is output and written to the stable correspondence graph storage area for use by the digital twin bit generation module. When a feasible record pair in the initial correspondence graph loses its retention qualification after any round of update, the digital twin generation module writes the feasible record pair into the conflict removal table and no longer participates in subsequent rounds of update.

[0073] After obtaining the stable correspondence map, the digital twin generation module performs field concatenation and sequential encoding on each retained record pair to form a digital twin bit set that maps one-to-one with the actual governance location. Specifically, the digital twin generation module reads the detection record and the implementation record corresponding to each retained record pair in the stable correspondence map, and writes the tree basin location identifier, soil layer identifier, governance cycle, detection record identifier, implementation record identifier, corresponding cost quadruple, and consistency result group into the same digital object in a fixed field order. The fixed field order is tree basin location identifier, soil layer identifier, governance cycle, detection record identifier, implementation record identifier, corresponding cost quadruple, and consistency result group, thereby generating a single digital twin bit.

[0074] After all digital twin bits are generated, the digital twin generation module encodes them sequentially according to the tree basin location identifier, soil layer identifier, and treatment cycle. First, the tree basin location identifiers are arranged in order of management number; then, the soil layers are arranged from shallow to deep; and finally, the treatment cycles are arranged from first to last. This order is written into the digital twin bit encoding field. The resulting digital twin bit set is then written into the digital twin bit storage area for direct reading by the improvement extraction module. When a certain retained record pair in the stable corresponding diagram is missing a detection record identifier or an application record identifier, the digital twin generation module does not generate the corresponding digital twin bit and writes the retained record pair into the missing field retention table, waiting for the data to be completed before reassembling.

[0075] Through the above processing, the digital twin bit set output by the digital twin generation module has completed the one-to-one correspondence, conflict resolution, and field solidification between the detection records and the remediation records. When the subsequent improvement extraction module reads the digital twin bits, it can directly call the current detection record and the detection record of the previous remediation cycle according to the tree plate location identifier, soil layer identifier, and remediation cycle to perform a positional comparison, without needing to re-determine which remediation action the detection result belongs to. At the same time, since the corresponding cost quadruple and consistency result group have been solidified within the digital twin bits, the subsequent qualification review module can directly use this mapping relationship to continue to perform the verification within and outside the remediation scope and the continuous verification of the cycle, avoiding the problems of object switching and inconsistent attribution in subsequent steps.

[0076] In practical applications: Taking a detection record of the surface soil layer at a certain pomelo tree basin location in the second treatment round as an example, the digital twin generation module first filters out two treatment records from the continuous treatment data that have the same tree basin location identifier, soil layer identifier, and treatment round as the detection record, and whose treatment time is earlier than the sampling time. After calculating the corresponding cost quadruples for each record, the module writes them into the initial correspondence map. Then, based on the soil pH increase and exchangeable acidity decrease results of the detection record relative to the previous treatment round, the module performs consistency verification and confidence update round by round on the two treatment records. It retains the treatment record whose treatment coverage boundary includes the current tree basin location, has the same target soil layer, and whose treatment time is closer to the current sampling time. Finally, the detection record and the retained treatment record are spliced ​​together to form a unique digital twin. Together with the digital twins formed by other tree basin locations, other soil layers, and other treatment rounds in the same orchard, the module forms a digital twin set for direct use in subsequent candidate improvement extraction.

[0077] The improvement extraction module is used to receive the digital twin bit set, perform a same-position comparison between the current detection record of the same digital twin bit and the detection record of the previous treatment round, and determine the existence of candidate improvement when the current pH value is higher than that of the previous treatment round and the current exchangeable acidity is lower than that of the previous treatment round, and output the candidate improvement set.

[0078] This implementation method illustrates how the improvement extraction module extracts candidate improvements with continued review value from digital twins that have already completed actual governance location mapping within the digital twin set. The purpose of this implementation process is to first establish a one-to-one correspondence between the digital twins of the current governance round and those of the previous governance round, based on the same tree basin location, the same soil layer, and the same remediation background. Then, based on the changing direction of the detection values ​​between the two rounds, an improvement difference record is formed. Subsequently, considering the changes in adjacent soil layers and adjacent tree basin locations, it is determined whether the current digital twin has candidate improvements that can enter the eligibility review process, thereby avoiding direct entry into subsequent effective improvement identification based solely on a single positive value. This implementation process includes the following steps:

[0079] The improvement extraction module first establishes a comparison relationship between the current digital twin and the digital twin of the previous treatment round through positional pairing, so as to obtain the improvement difference record that can be used for subsequent verification. Specifically, the improvement extraction module receives the digital twins of the current treatment round and the digital twins of the previous treatment round from the digital twin set, reads the tree basin location identifier, soil layer identifier, application record identifier, detection record identifier, soil pH value and exchange acidity from each digital twin, and uses the same tree basin location identifier, the same soil layer identifier and the same application record identifier as the positional pairing conditions to form a pair of digital twins between the current treatment round and the digital twin of the previous treatment round.

[0080] The purpose of introducing the modification record identifier is to separate two detection chains that are in the same tree basin location and soil layer but have different modification backgrounds, thus avoiding mismatching detection results generated under different modification actions as identical records. After pairing, the improvement extraction module calculates the soil pH difference and exchangeable acidity difference between the current detection record and the previous treatment round detection record in each paired digital twin. The soil pH difference is calculated by subtracting the soil pH value from the previous treatment round from the current soil pH value, and the exchangeable acidity difference is calculated by subtracting the exchangeable acidity value from the previous treatment round from the current exchangeable acidity value. The tree basin location identifier, soil layer identifier, application record identifier, current test record identifier, previous treatment round test record identifier, soil pH difference, and exchangeable acidity difference are combined into an improvement difference record and written into the improvement difference storage area for subsequent inter-layer same-direction verification and adjacent reverse investigation. When the digital twin of the current treatment round does not find a digital twin of the previous treatment round that meets the same-position matching condition, the improvement extraction module records the digital twin of the current treatment round as a missing preceding position and does not generate an improvement difference record, waiting for subsequent rounds to complete it or for the management evaluation module to process it as an unimproved position.

[0081] After the improvement extraction module generates the improvement difference record, it performs inter-layer same-direction verification and adjacent reverse screening around each digital twin position to screen out candidate improvements that only show improvement in the current digital twin position itself and are not negated by the surrounding change direction. Specifically, the improvement extraction module calls the adjacent soil layer digital twin position that has the same tree basin position identifier as the current digital twin position and whose soil layer order is directly adjacent to the current soil layer, and calls the adjacent tree basin position digital twin position that is in the same governance round as the current digital twin position and whose tree basin position is directly adjacent to the current tree basin position in the tree basin position mapping table.

[0082] The adjacent soil layers are determined according to a preset soil layer sequence table. The layer above and the layer below the current soil layer are considered directly adjacent soil layers. The positions of adjacent tree basins are determined according to the positions of the left and right adjacent tree basins within the same management zone. When the current tree basin position is located at the edge of the management zone, only the existing adjacent tree basin position on one side is called. The improvement extraction module then reads the soil pH difference and exchangeable acidity difference in the digital twins of adjacent soil layers and adjacent tree basin positions. It performs inter-layer same-direction verification and adjacent reverse check on the current digital twin. The establishment rule for inter-layer same-direction verification is: at least one adjacent soil layer digital twin satisfies that the soil pH difference is positive and the exchangeable acidity difference is negative. The passing rule for adjacent reverse check is: there is no record in the adjacent tree basin position digital twin that simultaneously satisfies that the soil pH difference is positive and the exchangeable acidity difference is negative, and its corresponding modification record identifier is different from the modification record identifier of the current digital twin.

[0083] When the current digital twin site satisfies the following conditions: the soil pH difference is positive, the exchangeable acidity difference is negative, and the interlayer unidirectional verification is successful and the adjacent reverse check is passed, the improvement extraction module identifies the current digital twin site as a candidate for improvement and writes its tree basin location identifier, soil layer identifier, improvement record identifier, current detection record identifier, soil pH difference, and exchangeable acidity difference into the candidate improvement set for the qualification review module to read. When the adjacent soil layer digital twin site is missing, the improvement extraction module records the digital twin site as a site to be verified by interlayers and does not write it directly into the candidate improvement set. When the adjacent tree basin location digital twin site is missing, only the existing adjacent tree basin locations are reverse checked and the adjacent missing identifier is written into the candidate improvement set simultaneously for the subsequent qualification review module to continue to verify in combination with the relationship between the improvement scope and the outside.

[0084] Through the above processing, the candidate improvement set output by the improvement extraction module has completed the initial screening of comparison between previous and subsequent rounds and consistency with the surrounding environment. When the subsequent qualification review module reads the candidate improvement set, it can directly perform continuous verification of adjacent soil layers, same-direction verification within the scope of the improvement, reverse exclusion verification outside the scope of the improvement, and continuous verification in the next retesting round around the digital twin position corresponding to the candidate improvement, without having to search for comparison relationships between previous and subsequent rounds again from within the digital twin position set. At the same time, since the generation of candidate improvements has excluded positive changes that only appear locally on a single digital twin position but are not supported by adjacent soil layers or have heterogeneous traction in the nearby tree basin, the subsequent qualification review faces candidate objects that are closer to the real improvement, which can reduce the risk of apparent improvement directly entering the management chain.

[0085] In practical applications: Taking the digital twin of the middle soil layer of a certain pomelo tree basin in the third treatment round as an example, the improvement extraction module first retrieves the digital twin of the same tree basin location, the same middle soil layer, and the same treatment record identifier in the previous treatment round. It calculates that the current soil pH difference is positive and the exchangeable acidity difference is negative. Then, it calls the surface digital twin and deep digital twin of the tree basin location to perform inter-layer same-direction verification. It calls the middle layer digital twin of the left and right adjacent tree basin locations in the same treatment round to perform adjacent reverse screening. When there is at least one layer of same-direction change in the surface or deep layer and there are no traction records with different treatment backgrounds but the same direction of change in the left and right adjacent tree basin locations, it is determined that the middle layer digital twin has a candidate improvement and writes it into the candidate improvement set for the subsequent qualification review module to continue review.

[0086] The qualification review module is used to receive the candidate improvement set and the corresponding improvement record, perform continuous verification of adjacent soil layers, same-direction verification within the improvement range, reverse exclusion verification outside the improvement range, and continuous verification in the next retesting round. Candidate improvements that pass all verifications are determined as effective improvements, and the rest are determined as apparent improvements. The improvement review results are then output.

[0087] This implementation method illustrates how the qualification review module performs continuity, scope consistency, and cross-round continuity reviews on candidate improvements, and accordingly categorizes candidate improvements into effective improvements or apparent improvements. The purpose of this implementation process is to avoid directly entering the governance effectiveness chain based solely on localized positive changes in the current round. Instead, it verifies candidate improvements layer by layer through three dimensions: adjacent soil layers, areas outside the scope of the improvement, and the next retesting round. This ensures that effective improvements ultimately entering the management evaluation module simultaneously meet four criteria: layer-by-layer continuity, consistency within the scope of the improvement, exclusion of areas outside the scope of the improvement, and continuation in subsequent rounds. This implementation process includes the following steps:

[0088] The qualification review module first constructs an initial review record based on the current candidate improvement to fix the spatial and implementation relationships required for the subsequent four rounds of verification. Specifically, the qualification review module receives the candidate improvement set and the corresponding implementation records, reads the tree basin location identifier, soil layer identifier, treatment round, current detection record identifier, implementation record identifier, sampling time, implementation time, implementation method, and implementation coverage boundary corresponding to the candidate improvement, and calls the adjacent soil layer digital twins, the digital twins within the implementation coverage boundary, and the adjacent digital twins outside the implementation coverage boundary of the current digital twin. Among them, the adjacent soil layer digital twins are determined according to the preset soil layer sequence table as the direct upper and lower adjacent layer digital twins of the current soil layer, the digital twins within the implementation coverage boundary are determined as the digital twins that are in the same treatment round as the current digital twin and whose tree basin location falls within the implementation coverage boundary, and the adjacent digital twins outside the implementation coverage boundary are determined as the digital twins that are directly adjacent to the outside of the implementation coverage boundary and whose treatment round is the same as the current digital twin.

[0089] The qualification review module then calculates the interlayer continuation difference sequence, the intra-boundary unidirectional sequence, and the extra-boundary reverse sequence. The interlayer continuation difference sequence is composed of the soil pH and exchangeable acidity differences between the current digital twin and its directly upper and lower adjacent digital twins, arranged in soil layer order. The intra-boundary unidirectional sequence is composed of the soil pH and exchangeable acidity differences of each digital twin within the application / cover boundary, arranged in tree basin position order. The extra-boundary reverse sequence is composed of the soil pH and exchangeable acidity differences of adjacent digital twins outside the application / cover boundary, arranged in adjacent order. Under the conditions that the time is earlier than the sampling time, the modification coverage boundary includes the current tree basin location, and the modification method corresponds to the current soil layer, the qualification review module combines the interlayer continuum difference sequence, the intra-boundary same-direction sequence, and the extra-boundary reverse sequence into an initial review record and writes it into the initial review storage area for use in the next retest round; when both the directly upper adjacent layer digital twin and the directly lower adjacent layer digital twin are missing, the qualification review module writes an interlayer missing identifier in the initial review record; when there is no adjacent digital twin outside the modification coverage boundary, the qualification review module writes an extra-boundary missing identifier in the initial review record and retains it for subsequent rounds of review.

[0090] After generating the initial review record, the qualification review module introduces the next retesting round information for the current candidate improvement and performs four rounds of verification in a fixed order to fix the review affiliation of the candidate improvement. Specifically, the qualification review module calls the digital twin of the next retesting round corresponding to the current candidate improvement. The correspondence is based on the tree basin location identifier being the same, the soil layer identifier being the same, and the treatment round being the next round directly after the current treatment round. The module also calculates the round continuation sequence and the round regression sequence. The round continuation sequence is composed of the results of the soil pH value continuing to rise and the exchangeable acidity continuing to fall between the digital twin of the current candidate improvement and the next retesting round, arranged in the order of the detection fields. The round regression sequence is composed of the results of the soil pH value turning to fall or the exchangeable acidity turning to rise in the next retesting round, arranged in the order of the detection fields.

[0091] The qualification review module writes the round continuation sequence and round rollback sequence back to the corresponding initial review record. After forming the review record, it performs four rounds of verification on each candidate improvement in the order of layer continuation, intra-boundary same-direction, extra-boundary reverse, and round continuation. The rules for layer continuation verification are: at least one directly adjacent soil layer in the layer continuation difference sequence maintains the same direction of soil pH difference and exchangeable acidity difference as the current candidate improvement. The rules for intra-boundary same-direction verification are: the intra-boundary same-direction sequence records no changes in the opposite direction to the current candidate improvement in each boundary's digital twin. The rules for extra-boundary reverse verification are: the extra-boundary reverse sequence... There are no adjacent digital twins that are in the same direction as the current candidate improvement and have different implementation record identifiers. The rule for the validity of the round continuation verification is that the round continuation sequence is valid and the round rollback sequence has not recorded the rollback result. When a round of verification fails, the qualification review module immediately stops the subsequent verification of the candidate improvement and records it as an apparent improvement candidate. When all four rounds of verification are valid, the candidate improvement is recorded as a valid improvement candidate, and the verification result is written into the stability review record for final attribution. When the digital twin of the next retesting round is missing, the qualification review module records the round continuation verification as incomplete and records the candidate improvement as an apparent improvement candidate. It will re-enter the review after the data of the next retesting round is supplemented.

[0092] After obtaining the stability review records, the qualification review module performs final attribution writing for candidate improvements to form the improvement review results that are directly called by the subsequent management and evaluation module. Specifically, the qualification review module reads the interlayer continuous difference sequence, the same-direction sequence within the boundary, the reverse sequence outside the boundary, the round continuation sequence, the round regression sequence, and the four-round verification results from each stability review record. Candidate improvements that simultaneously meet the requirements of adjacent soil layer continuity, same-direction within the modification range, reverse exclusion outside the modification range, and continued validity in the next retest round are identified as effective improvements. The remaining candidate improvements are identified as apparent improvements. The tree basin location identifier, soil layer identifier, treatment round, current detection record identifier, modification record identifier, interlayer continuous difference sequence, same-direction within the boundary, reverse sequence outside the boundary, round continuation sequence, round regression sequence, and review conclusion are combined and written into the improvement review results according to a fixed field order.

[0093] The fixed fields are ordered as follows: tree basin location identifier, soil layer identifier, treatment round, current detection record identifier, construction record identifier, inter-layer continuation difference sequence, intra-boundary same-direction sequence, extra-boundary reverse sequence, round continuation sequence, round regression sequence, and review conclusion. The generated improvement review results are written to the improvement review result storage area for the management and evaluation module to call. When a candidate improvement has both an inter-layer missing identifier and a missing identifier for the next retest round in the stability review record, the qualification review module will still classify it as an apparent improvement and write a supplementary review identifier in the review conclusion so that it can be called in subsequent supplementary retest arrangements.

[0094] Through the above processing, the improvement review results output by the qualification review module have categorized candidate improvements into effective improvements and apparent improvements according to unified rules, and have also solidified and written the spatial sequence and round sequence that formed the review conclusion. The subsequent management and evaluation module no longer needs to re-trace the original data of adjacent soil layers, the area inside and outside the improvement coverage, and the next retest round. It can directly execute the writing of digital twin effectiveness records, observation records, updating of governance order, updating of investment arrangements, and generation of exit judgment based on the review conclusion. At the same time, since the basis for the formation of apparent improvements has been clearly fixed, the subsequent supplementary retest arrangements and subsequent review arrangements can also be directly carried out for the verification links that have not passed, avoiding the mistaken inclusion of positive changes that do not yet have sustainable decision-making significance into the governance effectiveness chain.

[0095] In practical applications: Taking the candidate improvement formed in the middle soil layer of a certain pomelo tree basin during the third treatment round as an example, the qualification review module first retrieves the surface and deep digital twins of the middle soil layer to form an interlayer continuum difference sequence. Then, it retrieves the middle soil layer digital twins of other tree basin locations within the same treatment coverage boundary to form a sequence in the same direction within the boundary. It then retrieves the middle soil layer digital twins of tree basin locations directly adjacent to the treatment coverage boundary outside the boundary to form a sequence in the opposite direction outside the boundary. The module also considers the treatment time to be earlier than the current sampling time, the treatment coverage boundary to include the current tree basin location, and the treatment method to be... Under the condition of corresponding modification of the middle layer, an initial review record is formed; then, the digital twins of the same tree basin position and the same middle soil layer in the next retesting round are called to generate the round continuation sequence and the round regression sequence. Then, four rounds of verification are completed in the order of interlayer continuation, same direction within the boundary, reverse direction outside the boundary and round continuation. When all four rounds of verification are valid, the candidate improvement is determined as a valid improvement and written into the improvement review result. When any one of the verifications is invalid, it is determined as an apparent improvement and written into the review conclusion to be reviewed, so that the subsequent management and evaluation module can continue to arrange supplementary retesting and subsequent review.

[0096] The management and evaluation module is used to receive improvement review results and continuous governance data, write effective improvements into the digital twin effectiveness record and perform governance sequence updates, input arrangement updates and exit judgment generation, write apparent improvements into the observation record and perform supplementary retesting arrangements and subsequent review arrangements, and output governance management results and evaluation results.

[0097] This implementation method illustrates how the management assessment module transforms improvement review results into actionable governance management and assessment results. The purpose of this process is to first categorize completed digital twins into effective and apparent improvements and assign them to different record chains. Then, it combines the location of each tree basin and the distribution of each soil layer within the same governance round to update the governance sequence and input arrangements. Finally, based on the record status of consecutive governance rounds, it generates exit criteria, supplementary retesting arrangements, and subsequent review arrangements. This ensures that only improvements with sustainable management significance enter the digital twin's effective record, while apparent improvements only enter the observation chain and continue to undergo subsequent review. This implementation process includes the following steps:

[0098] The management assessment module first performs classification and writing of the improvement review results to establish the base records used for subsequent statistics, sorting, and exit determination. Specifically, the management assessment module receives the improvement review results and continuous governance data, reads the tree basin location identifier, soil layer identifier, governance round, modification record identifier, review conclusion, inspection record identifier, and round-related fields in the continuous governance data corresponding to each digital twin, and performs record distribution based on the review conclusion. When the review conclusion is a valid improvement, the management assessment module writes the corresponding digital twin into the digital twin effective record. The digital twin effective record includes at least the tree basin location identifier, soil layer identifier, governance round, modification record identifier, inspection record identifier, and valid improvement identifier.

[0099] When the review conclusion is apparent improvement, the management assessment module writes the corresponding digital twin bit into the observation record. The observation record includes at least the tree basin location identifier, soil layer identifier, treatment round, implementation record identifier, testing record identifier, apparent improvement identifier, and pending review identifier. After completing the classification writing, the management assessment module combines all digital twin effective records and observation records within the current treatment round into a classification record set and writes it into the classification record storage area for subsequent statistical retrieval. When the improvement review result contains a missing tree basin location identifier, a missing soil layer identifier, or a missing treatment round digital twin bit, the management assessment module does not perform classification writing but writes the digital twin bit into the missing field classification table to wait for upstream data to complete it.

[0100] After forming a categorized record set, the management assessment module performs statistics and marking on the location of each tree basin and each soil layer within the same treatment round to generate updated treatment sequence results, updated input arrangement results, supplementary retesting arrangements, and subsequent review arrangements. Specifically, the management assessment module calls up all digital twin effective records, observation records, and continuous treatment data within the same treatment round, and establishes a tree basin location statistics table according to the tree basin location identifier. For each tree basin location, it counts the number of digital twins corresponding to effective improvement, the number of digital twins corresponding to apparent improvement, and the number of digital twins corresponding to no improvement.

[0101] The number of digital twins corresponding to unimproved areas is the remaining number after subtracting the number of digital twins corresponding to effective improvements and the number of digital twins corresponding to apparent improvements from the total number of digital twins in the same treatment round and the same tree basin location. Subsequently, the management assessment module establishes a soil layer status table according to the tree basin location identifier and soil layer identifier, and performs status marking on each soil layer in each tree basin location. The rule for the establishment of continuous distribution of effective improvements is that two or more consecutive soil layers in the adjacent soil layer sequence record effective improvements. The rule for the establishment of repeated appearance improvements is that the same tree basin location... The same soil layer is recorded as having apparent improvement in both the current treatment round and the previous treatment round. After the statistics and marking are completed, the management assessment module generates treatment order update results and input arrangement update results in a fixed order. The fixed order is as follows: first, screen out tree basin locations where the number of digital twins corresponding to effective improvement is zero; then, screen out tree basin locations where the number of digital twins corresponding to apparent improvement is not zero; then, screen out tree basin locations where the number of digital twins corresponding to no improvement is not zero; and form treatment order update results according to the above screening order.

[0102] For the updated results of the investment arrangement, the management assessment module records tree locations with continuous distribution of effective improvement as maintenance investment targets, tree locations with repeated occurrences of apparent improvement as review investment targets, and tree locations with non-zero corresponding digital twins and no recorded continuous distribution of effective improvement as priority treatment investment targets. Simultaneously, the management assessment module writes the tree locations and soil layers with recorded apparent improvement in the current treatment round into the supplementary retesting arrangement, and writes the tree locations and soil layers with continuous apparent improvement in both the current and previous treatment rounds into the subsequent review arrangement. These results are then written into the treatment sequence update table, investment arrangement update table, supplementary retesting table, and subsequent review table for exit judgment and result write-back. When a tree location does not have a complete total number of digital twins in the current treatment round, the management assessment module records this tree location as an incomplete statistical target and does not include it in the current round's treatment sequence update and investment arrangement update.

[0103] After completing the classification statistics and arrangement generation within the current treatment round, the management assessment module performs an exit judgment on the records of consecutive treatment rounds for the same tree basin location and the same soil layer, and combines all management results into the treatment management results and assessment results. Specifically, for each tree basin location and each soil layer, the management assessment module calls the digital twin effective records and observation records of the two consecutive treatment rounds, and reads the review conclusion status of the previous treatment round and the review conclusion status of the current treatment round, respectively. When both consecutive treatment rounds record effective improvement in the digital twin effective records and neither treatment round records apparent improvement, the management assessment module generates an exit judgment for that tree basin location and soil layer, and writes it into the exit judgment table.

[0104] When the observation record of the current treatment round shows apparent improvement, the management assessment module maintains the original treatment status of the tree basin location and soil layer. The original treatment status includes keeping the tree basin location and soil layer in the treatment object set and not writing it into the exit judgment. At the same time, it is written into the supplementary retesting arrangement and subsequent review arrangement. When there are two consecutive treatment rounds, the previous treatment round recorded effective improvement, but the current treatment round did not record effective improvement or apparent improvement, the management assessment module does not generate an exit judgment and writes the tree basin location and soil layer into the key retesting object table of the next treatment round.

[0105] Finally, the management and evaluation module will combine the updated results of the governance sequence, the updated results of the input arrangement, the exit judgment, the supplementary retesting arrangement, and the subsequent review arrangement into the governance management results and the evaluation results in a fixed field order. The fixed field order is as follows: tree basin location identifier, soil layer identifier, governance round, governance sequence status, input arrangement status, exit judgment status, supplementary retesting status, and subsequent review status. When any governance round in two consecutive governance rounds is missing a digital twin effective record or observation record, the management and evaluation module will record the tree basin location and the soil layer as exit pending judgment objects, without generating an exit judgment, and wait for the next result to be supplemented before re-judging.

[0106] Through the above processing, the management assessment module further transforms the improvement review results into management results that can be directly used by the governance execution side. This allows effective improvements to enter the digital twin effectiveness record and participate in governance sequence updates, input arrangement updates, and exit determinations. Apparent improvements are entered into the observation record and trigger supplementary retesting arrangements and subsequent review arrangements. This ensures that the judgment results on whether the improvement has sustainable management significance are truly written into the subsequent governance process. At the same time, the exit determination is based on the digital twin effectiveness record and observation record of two consecutive governance rounds, which can avoid premature exit from the governance chain based solely on a single round of positive results.

[0107] In practical application: Taking the topsoil and middle soil layers of a pomelo tree basin as an example, the management assessment module first writes the digital twins of the topsoil layer that are recorded as effective improvements in the current treatment round into the digital twin effectiveness record, and writes the digital twins of the middle soil layer that are recorded as apparent improvements in the current treatment round into the observation record. Then, according to the tree basin location, the number of digital twins corresponding to effective improvements, apparent improvements, and no improvements in the current treatment round is counted. Combined with the records of the previous treatment round, it is determined whether the middle soil layer constitutes a repeated occurrence of apparent improvement, and thus the middle soil layer is written into the supplementary retesting arrangement and subsequent review arrangement. Subsequently, the records of the topsoil layer that have recorded effective improvements in two consecutive treatment rounds but have not recorded apparent improvements are retrieved, and an exit judgment is generated for the topsoil layer. Finally, the treatment sequence update result, investment arrangement update result, topsoil layer exit judgment, and middle soil layer supplementary retesting arrangement and subsequent review arrangement corresponding to the tree basin location are written into the treatment management result and assessment result.

[0108] The working principle of this scheme is as follows: First, the soil pH value, exchangeable acidity, and corresponding application, irrigation, drainage, and rainfall information obtained from each sampling in the pomelo orchard are compiled into continuous treatment data. Then, the detection records and application records that can be matched under the same tree basin location, the same soil layer, and the same treatment cycle are combined into digital twins, so that each digital twin corresponds to an actual treatment location in the orchard. Based on this, the system first compares the detection changes of the same digital twin in the current treatment cycle with those in the previous treatment cycle, extracts candidate improvements that show a positive trend, and then further examines whether these candidate improvements can be continuously reflected in adjacent soil layers. The system considers several factors: whether the improvement shows consistent change within the coverage area, whether it is not subject to reverse pull from other locations outside the coverage area, and whether it can maintain this improvement in the next retesting round. Only candidate improvements that pass all these checks are recognized as effective improvements and recorded in the digital twin's effectiveness record; the rest are only recorded as apparent improvements and recorded in the observation record. Finally, based on the distribution of effective and apparent improvements in different tree basin locations, different soil layers, and different treatment rounds, the system updates the subsequent treatment sequence, investment arrangements, supplementary retesting arrangements, subsequent review arrangements, and exit criteria, thereby avoiding mistaking superficial improvements for real treatment effectiveness.

[0109] For example, in a mountain pomelo orchard, if the surface soil at a certain tree basin undergoes an acidification treatment and the pH value increases while the exchangeable acidity decreases during a subsequent retest, the system will not immediately determine that the treatment has been effective. Instead, it will first correlate the current test results with those from the previous treatment cycle at that location, then check whether adjacent soil layers at that tree basin also show similar changes, whether other tree basins within the treatment coverage boundary show consistent changes, and whether adjacent locations outside the treatment coverage boundary show any potentially disruptive reverse changes. It will then continue to check whether the changes at that location will be maintained during the next retest. If all these conditions are met... Only when the improvement is confirmed will the system record the location as an effective improvement, and accordingly reduce subsequent investment or determine whether to exit treatment if there is stable improvement in two consecutive treatment rounds. If the current improvement is only temporary on the surface and the deeper levels have not kept up, or if the surrounding locations show that the change may not be the real result of this treatment, the system will record it as an apparent improvement and continue to arrange supplementary retesting and subsequent verification. In this way, orchard managers will not only see where things have improved, but also where there is real improvement, where there is only temporary improvement, where to treat first, and where treatment cannot be withdrawn yet, thus making the entire acidification treatment process more stable and reliable.

[0110] 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. A digital management and evaluation system for the acidification treatment of pomelo orchard soil, characterized by, include: The data acquisition terminal is used to collect soil pH value, exchangeable acidity, soil layer identification, sampling time, treatment records, irrigation and drainage records, and rainfall records at each sampling location, and output raw treatment data. The processing host is used to receive raw governance data, perform location binding, hierarchical binding, time sorting and historical storage, and output continuous governance data. The digital twin generation module is used to receive continuous governance data, combine the detection records and construction records of the same tree basin location, the same soil layer, and the same governance round, generate digital twin bits that are mapped one-to-one with the actual governance location, and output the digital twin bit set; The improvement extraction module is used to receive the digital twin bit set, perform a same-position comparison between the current detection record of the same digital twin bit and the detection record of the previous treatment round, and determine the existence of candidate improvement when the current pH value is higher than that of the previous treatment round and the current exchangeable acidity is lower than that of the previous treatment round, and output the candidate improvement set. The qualification review module is used to receive the candidate improvement set and the corresponding improvement records, perform continuous verification of adjacent soil layers, same-direction verification within the improvement range, reverse exclusion verification outside the improvement range, and continuous verification in the next retesting round. Candidate improvements that pass all verifications are determined as effective improvements, and the rest are determined as apparent improvements. The improvement review results are then output.

2. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 1, characterized in that: Also includes: The management and evaluation module receives improvement review results and continuous governance data, writes effective improvements into the digital twin effectiveness record and performs governance sequence updates, input arrangement updates and exit determination generation, writes apparent improvements into the observation record and performs supplementary retesting arrangements and subsequent review arrangements, and outputs governance management results and evaluation results.

3. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 2, characterized in that: The data acquisition terminal includes: Based on the tree basin location marker, soil layer marker, and sampling time corresponding to the current sampling location, soil pH and exchangeable acidity are bound at the same location, soil layer, and time, and the test records are output. For the current sampling location, based on the application time, application method, application coverage boundary, irrigation and drainage records, and rainfall records, perform location-based treatment association on the detection records and output the treatment association records; Call the detection record and governance association record that are the same as the current sampling position in the previous governance round, perform pairing between the current detection record and the corresponding record in the previous governance round, and output the continuous sampling record.

4. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 3, characterized in that: The processing host includes: Based on the tree basin location markers, soil layer markers, and sampling times in the original treatment data, soil pH, exchangeable acidity, treatment records, irrigation and drainage records, and rainfall records are merged according to the same location and the same stratum, and the corresponding records of the location and stratum are output. Based on the sampling time, application time, irrigation and drainage time and rainfall occurrence time in the location-stratum corresponding records, the records are sorted and arranged in rounds to output the treatment time sequence record; The system retrieves historical records corresponding to the same tree basin location and soil layer from the governance timeline records, performs sequential storage between the current cycle record and the historical cycle records, and outputs continuous governance data.

5. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 4, characterized in that: The digital twin generation module includes: Receive detection records and remediation records from continuous treatment data, perform same-value screening according to tree basin location identifier, soil layer identifier, and treatment round, construct a correspondence graph with detection records as nodes on one side and remediation records as nodes on the other side, calculate a corresponding cost quadruple consisting of location deviation, layer deviation, time interval, and coverage of the remediation effect range for each detection record and each remediation record, retain feasible record pairs under the constraints of the same tree basin location identifier, the same soil layer identifier, the same treatment round, and the remediation time being earlier than the sampling time, and output the initial correspondence graph.

6. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 5, characterized in that: The digital twin generation module also includes: For each feasible record pair in the initial correspondence map, the consistency of soil pH and exchangeable acidity in the detection records with the application method, application coverage boundary, and application time in the application records is checked. A consistency result group is formed, consisting of consistent results in the direction of acidity change, consistent results in the direction of pH change, and consistent results in the application coverage. The consistency result group is written back to the corresponding cost quadruple and confidence updates are performed round by round. When the same detection record is connected to multiple application records, the record pair with the first position in the updated cost quadruple is retained in the order of position deviation, layer deviation, time interval, and coverage gap. When the same application record is connected to multiple detection records, the record pair with continuous tree basin position, continuous soil layer, and the most recent sampling time is retained. Confidence updates and conflict resolution are performed cyclically until the record pairs retained in the previous and next rounds are completely consistent, and a stable correspondence map is output. Based on each pair of retained records in the stable correspondence diagram, the corresponding detection record and the implementation record execution field are concatenated to generate a digital twin bit including tree basin location identifier, soil layer identifier, treatment round, detection record identifier, implementation record identifier, corresponding cost quadruple and consistency result group. All digital twin bits are encoded in the execution order of tree basin location identifier, soil layer identifier and treatment round, and a set of digital twin bits that are mapped one-to-one with the actual treatment location is output.

7. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 6, characterized in that: The improved extraction module includes: Receive the digital twin bit of the current treatment round and the digital twin bit of the previous treatment round from the digital twin bit set, perform same-position pairing according to tree basin location identifier, soil layer identifier and treatment record identifier, calculate the pH difference value and exchangeable acidity difference value for the current detection record and the previous treatment round detection record respectively, and output the improvement difference value record. For each digital twin bit in the improvement difference record, call the digital twin bits of adjacent soil layers at the same tree basin location and the digital twin bits of adjacent tree basin locations in the same treatment round, and perform inter-layer same-direction verification and adjacent reverse investigation on the pH difference and exchangeable acidity difference. When the pH difference of the current digital twin bit is positive, the exchangeable acidity difference is negative, the adjacent soil layers have the same-direction change and the adjacent tree basin locations do not form reverse traction, it is determined that there is a candidate improvement for the digital twin bit, and the candidate improvement set is output.

8. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 7, characterized in that: The qualification review module includes: Receive the candidate improvement set and corresponding improvement records. According to the tree basin location identifier, soil layer identifier, and treatment round, call the adjacent soil layer digital twins, digital twins within the improvement coverage boundary, and adjacent digital twins outside the improvement coverage boundary of the current digital twin. Calculate the interlayer continuation difference sequence between the current digital twin and the adjacent soil layer digital twin, the intra-boundary same-direction sequence with the intra-boundary digital twin, and the extra-boundary reverse sequence with the extra-boundary adjacent digital twin. Under the conditions that the improvement time is earlier than the sampling time, the improvement coverage boundary includes the current tree basin location, and the improvement method corresponds to the current soil layer, combine the interlayer continuation difference sequence, the intra-boundary same-direction sequence, and the extra-boundary reverse sequence into the initial review record.

9. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 8, characterized in that: The qualification review module also includes: For the initial review record, the digital twin of the retest corresponding to the current candidate improvement in the next retest round is called. The round continuation sequence and round back-off sequence between the current candidate improvement and the next retest round are obtained. The round continuation sequence and round back-off sequence are written back to the corresponding initial review record to form the retest review record. Then, each candidate improvement is checked in four rounds in the order of layer continuation, same direction within the boundary, reverse direction outside the boundary, and round continuation. When a round of check is not valid, the candidate improvement is directly recorded as an apparent improvement candidate. When all four rounds of check are valid, the candidate improvement is recorded as a valid improvement candidate, and a stable review record is output. Based on the stability review records, candidate improvements that simultaneously meet the following criteria are identified as effective improvements: continuity between adjacent soil layers, same-direction improvement within the improvement area, reverse exclusion improvement outside the improvement area, and continued improvement in the next retesting round. The remaining candidate improvements are identified as apparent improvements. The corresponding interlayer continuity difference sequence, same-direction sequence within the boundary, reverse sequence outside the boundary, round continuation sequence, round regression sequence, and review conclusion are combined and written into the improvement review results.

10. The digital management and evaluation system for soil acidification control in pomelo orchards according to claim 9, characterized in that: The management assessment module includes: Receive the improvement review results and continuous governance data, call the tree basin location identifier, soil layer identifier, governance round, implementation record and review conclusion corresponding to each digital twin, write the digital twin with the review conclusion of effective improvement into the digital twin effectiveness record, write the digital twin with the review conclusion of apparent improvement into the observation record, and output the classification record set; Based on the classification record set, the system retrieves the digital twin effectiveness records, observation records, and continuous management data of all digital twins within the same management round. For each tree basin location, it performs item-by-item statistics based on the number of digital twins corresponding to effective improvement, the number of digital twins corresponding to apparent improvement, and the number of digital twins corresponding to no improvement. For each soil layer, it assigns corresponding markers based on the continuous distribution of effective improvement and the recurrence of apparent improvement. Finally, it generates management sequence update results, input arrangement update results, supplementary retesting arrangements, and subsequent verification arrangements in the order of first zero effective improvement, then non-zero apparent improvement, and finally non-zero non-improvement. For the same tree basin location and the same soil layer, the digital twin records of effective and observational data from two consecutive treatment cycles are retrieved. When effective improvement is recorded in both consecutive treatment cycles but no apparent improvement is recorded, an exit judgment is generated. When apparent improvement is recorded in the current treatment cycle, the original treatment status is maintained and supplementary retesting and subsequent review arrangements are written. The treatment sequence update results, input arrangement update results, exit judgment, supplementary retesting arrangements, and subsequent review arrangements are combined and written into the treatment management results and evaluation results.

Images