A method for dynamically regulating and controlling external water pressure of a deep buried water conveying tunnel and designing a drainage system

Abstract

This invention relates to the field of water conservancy engineering or tunnel engineering technology, and particularly to a method for dynamic control of external water pressure and drainage system design in deeply buried water conveyance tunnels. The method includes: acquiring hydrogeological zoning information, lining structure design parameters, and water resource constraint parameters; completing field mapping, terminology registration, division of pressure-controlled tunnel sections and pressure-controlled sectors, and constraint assembly and caliber unification to form a pressure budget table; constructing a controllable drainage network topology, completing the assembly of controllable pressure relief branches, valve group mapping, and monitoring orchestration; assembling monitoring data frames under monitoring orchestration constraints and performing consistency checks and coupled state updates to generate a state vector package; completing the assembly of control constraints and interlocking rules based on the state vector package, implementing valve group collaborative control, and performing anomaly diagnosis and degradation handling to form a closed-loop update package. This invention effectively improves the consistency and feasibility of the tunnel external water pressure control and drainage design process.

Description

Technical Field

[0001] This invention relates to the field of water conservancy engineering or tunnel engineering technology, and in particular to a method for dynamic control of external water pressure and design of drainage system for deeply buried water conveyance tunnels. Background Technology

[0002] In the field of deep-buried water conveyance tunnel engineering, dynamic control of external water pressure and drainage system design are crucial for ensuring the structural safety and long-term stable operation of the tunnel. Existing solutions typically rely on hydrogeological zoning information obtained from engineering survey data, combined with lining structure design parameters, to reduce the external water pressure behind the lining by setting drainage holes or blind drainage pipes. While these methods can achieve drainage and pressure reduction to some extent, they suffer from common limitations due to static design thinking and segmented, isolated control models, such as a disconnect between design parameters and operational status, a lack of coordination between control constraints and the objects of execution, and difficulty in tracing abnormal processes.

[0003] Existing methods are mostly based on geological survey conclusions during the design phase to determine drainage schemes. During construction and operation, when hydrogeological conditions change or structures are altered, adjustments can often only be made locally based on human experience.

[0004] During the design phase, engineering survey data, structural design documents, and water resource constraint information are usually in the form of scattered documents, lacking a unified field mapping and terminology registration mechanism. This leads to ambiguity in the interpretation of data from different sources during subsequent assembly, making it difficult to form a standardized input package that can be directly called by the automated system.

[0005] In the drainage system construction phase, existing technologies generally adopt a fixed drainage structure layout, failing to establish a one-to-one mapping relationship between pressure control units and adjustable capacity items. This makes it impossible to dynamically bind valve group groups to pressure control objects during subsequent operation, and control commands are difficult to accurately implement to specific branches. Regarding monitoring data acquisition, traditional solutions often focus on real-time values ​​of single measuring points, lacking unified time reference registration and field specification binding for multiple measuring points such as seepage pressure, flow rate, and valve opening. This leads to problems such as timestamp misalignment and unit specification confusion when aligning data frames across measuring points and valve groups.

[0006] In the data quality processing stage, existing technologies typically only perform simple outlier removal, lacking integrated processing for missing data distribution description, temporal consistency verification, and physical constraint consistency verification. This makes it difficult for downstream control modules to distinguish whether data anomalies originate from acquisition link failures or actual physical state changes. In the control decision-making stage, traditional drainage control is mostly triggered by single valve thresholds or manually operated remotely. It cannot jointly integrate the upper and lower thresholds and rate of change constraints in the design pressure budget table with the groundwater recharge boundary state, permeability parameter zoning state, and pressure relief branch blockage state in the operational state vector package. This results in control commands that may both exceed the lining safety boundary and cause surrounding rock stability risks.

[0007] In terms of anomaly diagnosis and closed-loop recording, existing systems typically only record the final alarm information, lacking multi-source aggregation and root cause classification of evidence from the monitoring side, status side, execution side, and rule side. This leads to the repeated occurrence of the same abnormal event without being able to locate the root cause, and makes it even more difficult to form a closed-loop update package containing version chain fields and traceability index fields to drive the system to self-correct. Summary of the Invention

[0008] To address the aforementioned technical problems, this invention provides a method for dynamically controlling external water pressure and designing a drainage system for deeply buried water conveyance tunnels, comprising: S100: Obtain the partition information field set, structural parameter field set, and water resource constraint field set; perform multi-source boundary merging and priority adjudication processing; and generate a pressure budget table. S200. Based on the pressure budget table, perform valve grouping and mapping relationship registration processing, and perform measurement point type assembly and field caliber binding processing to generate a monitoring arrangement package; S300. Based on the monitoring and orchestration package, perform data quality inspection, slot assembly, and coupling state update processes to generate a state vector package. S400. Based on the state vector package, perform control constraint assembly and interlocking rule assembly processing, and execute valve group collaborative control decision processing to generate a closed-loop update package.

[0009] Furthermore, the process of multi-source boundary merging and priority adjudication includes: The multi-source boundary merging and priority adjudication process includes extracting hydrogeological zoning boundaries from the zoning information field set, extracting lining joint and structural change boundaries from the structural parameter field set, extracting sensitive section boundaries that require limiting drainage intensity from the water resource constraint field set, and setting priority adjudication rules to merge conflicting boundaries.

[0010] Furthermore, the process of registering valve group grouping and mapping relationships includes: The valve group grouping and mapping relationship registration process includes generating valve group identifiers according to grouping rules and assigning branch numbers to the corresponding valve group identifiers. The grouping rules include constraints such as the inability to cross the boundary of the pressure control tunnel, the independent addressability of the pressure control sector, and the operation and maintenance reachability constraints. The process also includes performing intra-group consistency checks and cross-tunnel boundary checks to generate a valve group mapping matrix containing valve group identifiers, branch numbers, and pressure control unit indexes.

[0011] Furthermore, the process of assembling measurement point types and binding field specifications includes: The process of assembling measurement point types and binding field calibers includes assembling pressure measurement points, flow measurement points, and opening status measurement point types. It maps measurement point installation location entries to branch pre-buried location entries and pressure control sector numbers, plans acquisition channels, binds field calibers, registers a unified time reference, and generates a monitoring arrangement package containing a set of measurement point types, measurement point installation location entries, field caliber binding record entries, unified time reference entries, and data quality rule entries.

[0012] Furthermore, the process of performing data quality inspection and slot assembly includes: The monitoring data frame set is divided into time windows, and the pressure segment, flow segment and aperture segment are extracted. The consistency tests of caliber, time series, topology and physical constraints are performed. The consistency inspection includes consistency of unit diameter, consistency of measurement range, and consistency of missing measurement placeholders. The temporal consistency check includes sampling time offset detection, duplicate sampling detection, and timestamp reversal detection; The topology consistency check includes whether the branch number exists, whether the pressure control unit index exists, and whether the valve group identifier and the branch number are consistent. The physical constraint consistency test includes consistency tests of the rate of change of seepage pressure with the direction of opening action, consistency tests of the change of flow rate with the change of opening, and consistency tests of the change of seepage pressure difference between adjacent pressure control tunnel sections with the longitudinal pressure gradient constraint entries. It generates a state recognition input slot set containing time window index field, seepage pressure segment field, flow rate segment field, opening segment field and tag set field.

[0013] Furthermore, the coupling state update process includes: The coupled state update process includes loading prior configuration and performing quality gating and action association gating. The quality gating is based on the timing anomaly marker field and the caliber conflict marker field in the consistency inspection summary field; The action association gating is based on the action mismatch flag field; Integrating execution constraint consistency with credibility; The constraint unification reads the state constraint entries and performs boundary clipping and rate of change clipping on the three types of state fields; The credibility fusion input includes the fragment quality summary field of the slot entry, the missing measurement distribution description field, the tag set field, and the frame skeleton version tag in the reference version reference field, and generates a state vector package containing the groundwater recharge boundary state field, the permeability parameter zoning state field, the pressure relief branch blockage state field, and the state credibility field.

[0014] Furthermore, the process of performing controlled and constrained assembly includes: Assemble the constraint entries in the stress budget table into a constraint expression structure that includes a static constraint area and a dynamic constraint area. The static constraint area includes constraint object field, constraint boundary field, and constraint priority field, while the dynamic constraint area includes time window index field reference, state reference slot field, rate of change threshold field, gradient threshold field, trigger condition field, and release condition field. Map the state fields in the state vector package to the state reference slots in the dynamic constraint area, establish a one-to-one mapping relationship based on the pressure control unit index field, and establish a one-to-many mapping relationship based on the branch number or valve group identifier field.

[0015] Furthermore, the process of assembling interlocking rules includes: Arrange interlocking rule entries to generate an interlocking rule diagram containing rule node fields, condition node fields, action node fields, and rollback node fields; Execute access control; when the state credibility field is lower than the gate control threshold, write the gate rejection flag field and trigger the fallback node field. Perform conflict resolution by prioritizing multiple constraint conflicts based on the constraint priority field and rule version number.

[0016] Furthermore, the process of implementing coordinated control decision-making for the valve group includes: The valve group collaborative control decision processing includes a set of assembly control objects; Read the set of dynamic constraint entries corresponding to the index field of this time window from the control constraint package; Extract the index field of the pressure control unit to form a set of control objects; Read the mapping relationship from the valve group mapping matrix, map the set of controlled objects to the set of valve group identifier fields and the set of branch number fields, and generate the valve group control context package.

[0017] Furthermore, the closed-loop update package includes: The closed-loop update package includes an anomaly classification result field, a handling action summary field, and a version chain field.

[0018] The following are its main beneficial effects: (1) By generating a pressure budget table through multi-source boundary merging and priority adjudication, the problem of scattered and inconsistent data in exploration, structure and water resources is solved, so that the safety boundary and rate of change constraints in the design period can be directly called by the subsequent control link in the form of standardized fields, avoiding constraint mismatch caused by data ambiguity.

[0019] (2) By registering valve group grouping and mapping relationship, as well as assembling measurement point type and binding field caliber, a one-to-one mapping relationship between pressure control unit and adjustable branch was established, and the spatiotemporal reference of monitoring data was unified, so that control commands can be accurately sent to specific branches, while ensuring that data frames across measurement points and across valve groups can be aligned on the time axis and parsed semantically.

[0020] (3) By checking data quality, assembling slots and updating coupled state, the original monitoring data containing missing measurements and abnormal markers are transformed into state vector packages with confidence components, enabling the downstream control module to distinguish the reliability of the data source. Even when the data quality is poor, the continuity of the state output can still be maintained by weighted update or alternative estimation, which enhances the system's adaptability to complex on-site environments. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a design method for dynamic control of external water pressure and drainage system in a deeply buried water conveyance tunnel, as provided in an embodiment of this application. Detailed Implementation

[0022] Example 1: Refer to Figure 1 This is a flowchart illustrating a method for dynamic control of external water pressure and design of a drainage system for a deeply buried water conveyance tunnel, provided by an embodiment of the present invention. The process may include at least steps S100-S400: S100: Obtain the partition information field set, structural parameter field set, and water resource constraint field set; perform multi-source boundary merging and priority adjudication processing; and generate a pressure budget table. S200. Based on the pressure budget table, perform valve grouping and mapping relationship registration processing, and perform measurement point type assembly and field caliber binding processing to generate a monitoring arrangement package; S300. Based on the monitoring and orchestration package, perform data quality inspection, slot assembly, and coupling state update processes to generate a state vector package. S400. Based on the state vector package, perform control constraint assembly and interlocking rule assembly processing, and execute valve group collaborative control decision processing to generate a closed-loop update package.

[0023] Step S100 includes at least steps S110-S130: S110. Obtain hydrogeological zoning information, lining structure design parameters, and water resource constraint parameters, perform field mapping and terminology registration processing, and obtain the engineering input package.

[0024] Specifically, the hydrogeological zoning information comes from the fusion of engineering survey data along the tunnel route, supplementary survey data during construction, and monitoring archive data during operation. The fusion results can be composed of Geographic Information System (GIS) base map, borehole columnar section, pumping test results, groundwater level observation records, fault fracture zone and karst development section identification, water-bearing capacity level and permeability level, etc., and the unified standard of "mileage-coordinates-tunnel axis" is completed before entering this step, so that each zoning record can be uniquely located to the tunnel mileage range.

[0025] To avoid ambiguities in subsequent assembly due to differences in terminology, units, coordinate references, and time references from different data sources, this step establishes a processing link for field mapping and terminology registration. Field mapping refers to: The hydrogeological zoning information, the lining structure design parameters, and the water resource constraint parameters are respectively converted into standard field sets that can be directly called in subsequent steps, and the field naming, unit caliber, value range, default value processing, and source traceability marking are completed under the constraints of the same field mapping table entries. Terminology registration refers to registering unified terms, terminology definitions, and field definitions that will be repeatedly referenced in subsequent steps of this invention in the terminology dictionary field set. The terminology dictionary field set and the field mapping table entries together serve as the basic data structure for auditing and version evolution.

[0026] Furthermore, the lining structure design parameters are derived from the archived results of structural design documents, construction drawings, and change records. The lining structure design parameters include at least the lining type identifier, lining thickness item, lining concrete strength grade item, steel reinforcement information item, lining joint and construction joint arrangement item, lining waterproofing and drainage construction reserved conditions item, and stress calculation diameter item for external water pressure analysis.

[0027] The water resource constraint parameters are derived from engineering water use permits, ecological discharge constraints, construction period water use plans, operation period fire fighting and cleaning water use plans, and groundwater environmental impact assessment constraints. These water resource constraint parameters are uniformly converted into a water resource constraint field set during field mapping, and the composition of the "water resource constraint field set" and its correspondence with subsequent control constraint assembly are clearly defined in the terminology registration to avoid confusion between policy clauses and executable engineering parameters.

[0028] In terms of implementation, field mapping and terminology registration processing can be performed by a "field mapping and terminology registration module," which can be deployed in the engineering data management terminal during the design phase or in the project data platform during the construction phase, and run under trigger conditions. When a new exploration zone record, structural change record, or water resource constraint update record is received, the field mapping and terminology registration module performs integrity checks, duplication checks, and conflict checks on the input data. When the verification passes, the input data is written into the corresponding field set and a source traceability mark is generated; when the verification fails, the reason for the abnormality is recorded and an item to be supplemented is generated. The item to be supplemented is written into the engineering input package along with the quality mark item, so that subsequent steps can perform default processing or rejection processing when dividing the pressure control tunnel section and pressure control sector.

[0029] Therefore, the output of this step is organized into an engineering input package, which includes a partition information field set, a structural parameter field set, a water resource constraint field set, and a terminology dictionary field set. The engineering input package is explicitly stated in natural language within the paragraph as the input for the next step S120, and is called to extract mileage range entries and circumferential partition rule entries. At the same time, the terminology dictionary field set will be referenced in the constraint entry standardization stage of the subsequent S130 to complete the standardization of spelling and standardization.

[0030] S120. Extract mileage range entries and circumferential partitioning rule entries from the engineering input package, perform pressure control tunnel segment division and pressure control sector division, and generate pressure control unit mesh.

[0031] The mileage range entries can be obtained by combining the starting and ending points of the tunnel axis mileage, the mileage of the zone boundary, the mileage of the fault fracture zone, the mileage of the water-rich abrupt change, and the mileage of the construction joint. The circumferential partitioning rule entries are defined as sector division rules along the cross-sectional direction of the tunnel. The rules include at least sector number entries, sector angle range entries, sector naming entries, and sector-to-structural part mapping entries. The structural parts may specifically correspond to the arch crown, arch waist, sidewall, and invert arch, etc.

[0032] To ensure that the pressure control object has an executable spatial granularity and forms a one-to-one mapping with the subsequent controllable drainage network topology assembly, this step proposes an operable processing link for the division of pressure control tunnel sections and pressure control sectors: First, the entire line is divided into several pressure control tunnel sections based on the mileage range entries as the main line. Then, each pressure control tunnel section is divided into pressure control sectors according to the circumferential zoning rules, and the pressure control tunnel section number, pressure control sector number, and pressure control unit index are combined and registered.

[0033] Furthermore, the implementation of pressure-controlled tunnel section division can adopt the mechanism of "multi-source boundary merging and priority adjudication", that is, extracting hydrogeological zoning boundaries from the zoning information field of the engineering input package, extracting lining joint and structural change boundaries from the structural parameter field, extracting sensitive section boundaries that need to limit drainage intensity from the water resource constraint field, and setting priority adjudication rules for various types of boundaries. When different boundaries conflict or the spacing is less than the preset minimum segment length, boundary merging is performed and the merging reason is recorded. The merging reason is written into the source tracing mark of the pressure control tunnel segment mileage range.

[0034] The implementation of pressure-controlled sector division can adopt a mechanism of "fixed template + local correction". That is, an initial sector set is generated based on the standard sector template given by the circumferential partitioning rule entries, and then the sector boundaries are locally corrected by combining the water-rich directionality, structural fracture dominance direction or bias risk direction in the partitioning information field set. The local correction does not change the sector quantity entries and sector naming entries, but only updates the sector angle range entries and records the correction reasons to avoid terminology drift when "the pressure-controlled sector" is referenced in the future.

[0035] To improve automation, this step can be automatically executed by the "Pressure Control Unit Mesh Generation Module" during the design phase. The triggering conditions can be set as follows: when the version number of the engineering input package output by S110 is updated or new / changed entries appear in the partition information field set, the Pressure Control Unit Mesh Generation Module will re-execute the pressure control tunnel segment division and pressure control sector division; when the version number does not change but the data quality mark entry changes, the Pressure Control Unit Mesh Generation Module will only perform incremental updates on the affected mileage range entries, thereby reducing redundant calculations and manual intervention.

[0036] Regarding anomaly handling, if gaps or overlaps exist in the mileage range entries, this step performs gap filling and overlap trimming: for gapped segments, default inheritance is performed according to the partition information field set of adjacent pressure-controlled tunnel segments, and a default inheritance flag is written; for overlapping segments, trimming is performed according to priority adjudication rules, and a trimming flag is written. If the circumferential partition rule entry does not provide a sector angle range entry or a sector naming entry, this step calls the pre-registered default rule entry from the terminology dictionary field set to generate a default sector set and writes a default rule flag so that subsequent steps can identify this sector as the default configuration source.

[0037] After completing the above processing, this step organizes the results into a pressure control unit grid. The pressure control unit grid includes the pressure control tunnel segment number, the pressure control tunnel segment mileage range, the pressure control sector number, the circumferential range of the pressure control sector, and the pressure control unit index, and specifies its destination within the segment. The pressure control unit grid serves as the input for the "pressure control unit grid" in the next step S130 and is used for lining safety constraint assembly, surrounding rock stability constraint assembly, and constraint item standardization. At the same time, the pressure control unit index will be used for one-to-one mapping between the branch and the pressure control object in the controllable pressure relief branch assembly stage in the subsequent S210, and will serve as an index field in the status identification input slot set in the subsequent S320 to receive the assembly of the seepage pressure segment, flow segment, and opening segment.

[0038] S130. Perform lining safety constraint assembly, surrounding rock stability constraint assembly and constraint item caliber unification on the pressure control unit mesh, and generate a pressure budget table.

[0039] Specifically, the lining safety constraint assembly refers to: for each pressure control unit index in the pressure control unit grid, extracting the lining type identifier, lining thickness entry, lining concrete strength grade entry, steel reinforcement information entry, and stress calculation diameter entry corresponding to the pressure control tunnel segment number from the structural parameter field set of the engineering input package; and generating an upper limit threshold entry that can be directly used for the assembly of the "external water pressure budget zone" based on the same stress calculation diameter entry without introducing formula expression, according to the preset bearing verification rules or verification result referencing rules. The upper limit threshold item is defined in terms of terminology as "the upper limit control item of the external water pressure acceptable to the pressure control unit". Its value can be obtained from the allowable external pressure item in the structural design verification conclusion, the external pressure limit item corresponding to the crack control item, or the control item confirmed by the design unit, and is written into the pressure budget table along with the source traceability mark.

[0040] The surrounding rock stability constraint assembly refers to: for the same pressure control unit index, extracting water-bearing level, permeability level, fault fracture zone identifier, karst development identifier, grouting ring condition item and groundwater level change record item from the partition information field of the engineering input package, and generating lower limit threshold item based on the surrounding rock stability discrimination caliber item; The lower limit threshold item is defined in terms of terminology as "the lower limit control item of external water pressure that the pressure control unit should not be lower than". Its engineering meaning is that the drainage and pressure reduction process should not cause the surrounding rock to be in an unfavorable seepage condition or generate a risk of seepage damage. The lower limit threshold item is also written into the pressure budget table and carries a source traceability mark and a data quality mark.

[0041] Furthermore, in order to ensure that the subsequent valve group coordinated control decision has executable time and space constraints, this step generates external water pressure change rate constraint items and longitudinal pressure gradient constraint items in addition to assembling upper limit threshold items and lower limit threshold items.

[0042] The external water pressure change rate constraint entry is defined as a constraint entry for the allowable change range of internal and external water pressure within a preset time window for the same pressure control unit index. Its source can be the comprehensive assembly result of the surrounding rock stability discrimination entry, the lining stress sensitivity entry, and the construction and operation management entry. The longitudinal pressure gradient constraint entry is defined as the constraint entry for the external water pressure difference between adjacent pressure-controlled tunnel sections. Its source can be the hydraulic connection strength, fault water-blocking entry, and lining joint arrangement entry reflected in the zoning information fields.

[0043] All the above constraint items undergo a standardization process during assembly. This standardization process is performed by the "Constraint Item Standardization Module." This module verifies the standardized terms in the terminology dictionary field set, and ensures consistency in the field standards of units, value ranges, default value rules, and source traceability markers. It also adjudicates conflicting constraint items: when the upper limit threshold item is less than the lower limit threshold item, a conflict marker is recorded and the item is adjusted to an executable budget band according to preset adjudication rules. At the same time, the adjustment record is written into the change record corresponding to the constraint version number. When there are missing constraints for external water pressure change rate and longitudinal pressure gradient, they are supplemented according to the default rule items in the terminology dictionary field set and a default marker is written to facilitate the identification of the constraint source in subsequent steps.

[0044] Regarding the operation process and automation triggering, this step can be executed periodically or by event triggering by the "Pressure Budget Table Generation Module" in the engineering data platform: when the pressure control unit grid version output by the pressure control unit grid generation module is updated, or the structural parameter field set of the engineering input package is changed, or a sudden change in the water abundance level appears in the zoning information field set, the pressure budget table generation module re-executes the lining safety constraint assembly and surrounding rock stability constraint assembly; when only the water resource constraint field set is updated, the pressure budget table generation module performs incremental assembly of the water resource constraint items and updates the constraint version number without changing the pressure control unit grid, so that the subsequent zoning controllable drainage network topology assembly of S200 can call the latest pressure budget table and maintain version traceability.

[0045] This step clarifies the core parameters of the minimum set: the pressure control unit index, upper limit threshold entries, lower limit threshold entries, external water pressure change rate constraint entries, longitudinal pressure gradient constraint entries, and constraint version numbers constitute the minimum set of the pressure budget table; water resource constraint entries, source traceability markers, and data quality markers are preferred extended fields, used to connect with the subsequent S410 control constraint assembly and support long-term operation and maintenance audits.

[0046] Step S200 includes at least steps S210-S230: S210. Obtain the pressure budget table and drainage structure library, perform controllable pressure relief branch assembly processing, and obtain a controllable drainage network topology.

[0047] Specifically, this step follows the pressure budget table formed in the previous step S130. The pressure budget table includes a pressure control unit index, upper limit threshold entries, lower limit threshold entries, external water pressure change rate constraint entries, longitudinal pressure gradient constraint entries, and constraint version numbers at the data structure level. It also has source traceability markers and data quality markers. The pressure budget table is accessed by the engineering data platform or the engineering database in the design phase. When reading, consistency checks are performed according to the constraint version number to prevent mismatches between the pressure control object and the constraint entries.

[0048] The drainage structure library, which is obtained in parallel with the pressure budget table, is derived from the fusion archive of existing tunnel drainage structure standard drawings, project construction drawings, feasible structure list, and material and equipment ledger. The drainage structure library is defined as a set of structure entries that include drainage structure type entries, geometric parameter entries, connection interface entries, construction and installation constraint entries, operation and maintenance constraint entries, adjustable capacity entries, and blockage sensitivity entries. Among them, the adjustable capability item is used to characterize whether a certain structural item supports opening adjustment, bypass switching, and backwashing action sequence, and the clogging sensitivity item is used to characterize the risk classification of channel cross-sectional area attenuation of the structural item under conditions of sand, scaling, or microbial adhesion.

[0049] Furthermore, the controllable pressure relief branch assembly process is performed by the controllable pressure relief branch assembly module, which consists of a structure screening unit, a branch synthesis unit, an interface verification unit, a redundant assembly unit, and a version registration unit. The structure screening unit retrieves candidate structure entries from the drainage structure database according to the pressure budget band width, external water pressure change rate constraint entries, and longitudinal pressure gradient constraint entries corresponding to the pressure control unit index, and filters out structure entries that cannot be constructed or cannot be pre-embedded according to the construction and installation constraint entries. The branch synthesis unit synthesizes candidate structural items into branch items in the order of "pressure relief inlet - adjustable unit - metering interface - drainage connection port", and assigns branch numbers to the branch items. The branch numbers are generated using traceable numbering rules, which are written into the version registration field of the branch item. The interface verification unit performs diameter verification on the connection interface items of the branch items and the drainage channel. The verification content includes interface diameter consistency, material compatibility, construction sequence compatibility and maintenance space constraints. If the verification fails, a conflict mark is recorded and the process is backed to the structural screening unit to reselect structural items. Under the premise of meeting the operation and maintenance constraints, the redundant assembly unit assembles bypass channel entries or backwash interface entries for branch entries with higher blockage sensitivity levels, and writes the redundant assembly method into the operation and maintenance action field of the branch entry for subsequent abnormal diagnosis and degradation handling entries to call.

[0050] The version registration unit generates a topology version number for the assembly result of the controllable pressure relief branch, and establishes an association between the topology version number and the constraint version number of the pressure budget table, thereby supporting subsequent steps to trigger incremental topology updates when the constraint version is updated.

[0051] Understandably, the controllable pressure relief branch assembly process does not simply involve directly laying out drain holes or blind pipes. Instead, it establishes a one-to-one mapping relationship between the pressure control unit index and the adjustable capability item, so that each pressure control unit index corresponds to at least one branch number, and the branch number serves as an addressable execution object in the subsequent valve group collaborative control decision link.

[0052] During the assembly process, spatial constraint mapping is also performed: the pressure control unit index is associated with the pressure control tunnel section number and the pressure control sector number, and then associated with the lining joint and construction joint layout entries, thereby obtaining the branch road pre-embedded location entries and crossing structure entries; The pre-embedded location entries for branch roads are written into the topology node attribute field, and the crossing construction entries are written into the topology edge attribute field, so that the measurement point installation location information can be synchronously assembled in the subsequent measurement point type assembly and field caliber binding stages.

[0053] For pressure-controlled tunnel sections with abrupt changes in water-bearing levels or fault fracture zone markings, this step adds a "sudden surge condition linkage interface" attribute field to the branch entries within the tunnel section. This attribute field is used to mark that the branch can be included in the rapid pressure relief linkage set during the interlocking rule assembly process. However, this step does not involve the decision-making content of the interlocking rule assembly; it only completes the writing of the interface attribute and version registration.

[0054] Regarding the triggering conditions, the controllable pressure relief branch assembly module has two types of triggering links: design stage triggering and construction stage triggering. The design stage triggering is driven by the topology generation task scheduling unit. When it is detected that the constraint version number of the pressure budget table has changed, or the set of construction entries in the drainage structure library has been added, or the construction and installation constraint entries have been updated, the topology generation task scheduling unit triggers the reassembly of the controllable pressure relief branch assembly module. The construction phase is triggered by on-site change events. When the material and equipment ledger shows a change in the supply specifications of a certain type of adjustable unit, or a change in the construction space constraints, the assembly module performs branch entry replacement without changing the pressure control unit index set, and adds a change record under the topology version number, so that the same topology version number can be traced back to every change event.

[0055] In terms of anomaly handling, if a pressure control unit index does not find an adjustable capacity entry that meets the construction and installation constraints in the drainage structure database, this step writes the pressure control unit index into the non-assembly list field and records the reason code. The reason code is also written into the quality mark field of the controllable drainage network topology for subsequent steps to identify and to adopt a downgraded mapping strategy in the valve group grouping and mapping relationship registration stage.

[0056] After assembly, the output of this step is a controllable drainage network topology. The controllable drainage network topology is defined as a network data structure consisting of a set of topology nodes and a set of topology connections. The set of topology nodes includes pressure control unit index nodes, branch number nodes, and drainage channel nodes. The set of topology connections includes mapping connections from pressure control unit indexes to branch numbers, connection connections from branch numbers to drainage channels, and redundant connections of bypass channels. The destination of the fields is indicated at the end of this paragraph using natural language: The controllable drainage network topology serves as the input for the next step S220, "Controllable Drainage Network Topology," and is used to extract branch numbers and pressure control unit indexes, and to complete valve grouping and mapping relationship registration. At the same time, the branch pre-buried location entries and interface attribute fields in the controllable drainage network topology are called in the subsequent S230 stage of measuring point type assembly and field caliber binding, forming a monitoring arrangement package and crossing the main step to enter the measuring point acquisition channel binding and data frame assembly processing link of S300.

[0057] S220. Extract the branch number and pressure control unit index from the controllable drainage network topology, register the valve group grouping and mapping relationship, and generate the valve group mapping matrix.

[0058] Specifically, this step takes the controllable drainage network topology output by S210, reads the branch number node set and the pressure control unit index node set from the topology node set, and reads the mapping connection from the pressure control unit index to the branch number from the topology connection set, thereby forming a basic mapping relationship that can be grouped.

[0059] The valve group grouping and mapping relationship registration is performed by the valve group grouping and mapping relationship registration module, which consists of a grouping strategy unit, an intra-group consistency verification unit, a cross-hole section boundary verification unit, a mapping registration unit, and a version anchoring unit.

[0060] The grouping strategy unit receives the pressure-controlled tunnel section number, pressure-controlled sector number, branch pre-embedded location entry, adjustable capacity entry, and blockage sensitivity entry, generates valve group identifiers according to preset grouping rules, and assigns branch numbers to the corresponding valve group identifiers. The core constraint minimum set of the grouping rules includes the constraint that the boundary of the pressure-controlled tunnel section cannot be crossed, the constraint that the pressure-controlled sector can be independently addressed, and the constraint that operation and maintenance reachability. The preferred extension of the grouping rules includes the constraint that segments of the same water-rich level should be grouped together first, the constraint that the procedures of the same construction section should be consistent, and the constraint that bypass channels should be shared. The preferred extension is used to reduce the complexity of operation and maintenance scheduling without changing the basic feasibility of the closed-loop link.

[0061] Furthermore, the consistency verification unit within the group performs consistency verification on the branch numbers under the same valve group identifier. The verification content includes consistency of adjustable capacity items, consistency of opening status measuring point interfaces, consistency of metering interfaces, and consistency of bypass channel assembly methods.

[0062] When the consistency check fails, the module removes the conflicting branch number from the valve group identifier and returns it to the grouping strategy unit for reallocation. At the same time, it writes a conflict record field into the mapping registration unit. The conflict record field contains the branch number, conflict type entry, and processing action entry for review during the subsequent anomaly diagnosis entry execution phase.

[0063] The cross-tunnel boundary verification unit performs boundary verification on the range of pressure-controlled tunnel numbering covered by the valve group identifier. The boundary verification follows the "pressure-controlled tunnel boundary cannot be crossed constraint". When it is detected that a valve group identifier contains multiple pressure-controlled tunnel numbers, the module generates a split action entry and writes it into the mapping registration unit. The split action entry drives the grouping strategy unit to split the valve group identifier into multiple valve group identifiers and re-register the mapping relationship, thereby maintaining the structural consistency between the valve group and the pressure-controlled tunnel.

[0064] Understandably, the mapping relationship registration is not just a textual record of "which group the valve belongs to", but rather a structured product that generates a valve group mapping matrix.

[0065] The valve group mapping matrix is ​​defined as a matrix-type record structure containing valve group identifier, branch number, pressure control unit index, pressure control tunnel number, pressure control sector number, branch pre-embedded location entry, adjustable capacity entry, and topology version number. The matrix-type record structure can be implemented as a relational table structure or a key-value index structure, but it is uniformly referred to as valve group mapping matrix at the terminology level.

[0066] When generating the valve group mapping matrix, the mapping registration unit performs field alignment: the valve group identifier adopts a unified naming rule, the branch number references the number in the controllable drainage network topology, the pressure control unit index references the index field in the pressure control unit grid, and the topology version number references the topology version number generated in S210, thereby establishing cross-step numbering consistency.

[0067] The version anchoring unit establishes an anchoring relationship between the valve group mapping matrix and the constraint version number and topology version number. The anchoring relationship is written into the version anchoring field of the valve group mapping matrix, so that the pressure budget table version corresponding to the valve group mapping matrix can be traced in subsequent steps when processing control constraint assembly and interlocking rule assembly.

[0068] Regarding automated triggering conditions, there are three types of triggering entry points in this step.

[0069] The first type is triggered by a change in the topology version number. When the topology version number output by S210 changes, the valve group grouping and mapping relationship registration module re-executes the grouping strategy unit and updates the valve group mapping matrix. The second type is triggered by changes in operation and maintenance constraints. When the operation and maintenance constraint entries or maintenance space constraint entries in the drainage structure library are changed, the module updates only the valve group identifier and branch number allocation relationship without changing the pressure control unit index set, and writes the change record into the version anchoring field. The third type is triggered by on-site status feedback. When the monitoring archive data during operation shows that the blockage sensitivity item of a certain tunnel section has been upgraded or the bypass channel is unavailable, the module performs valve group identification reorganization and updates the shared bypass channel constraints. The update is also recorded through the version anchor field, thereby supporting the auditability of the closed-loop iteration process.

[0070] In terms of anomaly handling, when multiple branch numbers corresponding to the same pressure control unit index are detected and the adjustable capacity entries are inconsistent, this step marks the pressure control unit index as a multi-branch heterogeneous entry and writes the heterogeneous mark field into the valve group mapping matrix for the subsequent valve group collaborative control decision-making stage to adopt a hierarchical decision-making strategy. However, this step does not enter the control decision-making process, but only completes the marking and registration.

[0071] Upon completion, the output of this step is a valve group mapping matrix, and its destination is indicated at the end of this section using natural language: The valve group mapping matrix serves as the input for the next step S230, and is used for measuring point type assembly and field caliber binding to generate a monitoring orchestration package. At the same time, the valve group mapping matrix is ​​called in the subsequent control constraint assembly and interlocking rule assembly process of S410, serving as an index bridging field between the pressure control unit index and the execution object. In the valve group collaborative control decision stage of S420, it is input together with the control constraint package to generate the valve group opening instruction set, thereby maintaining the cross-main step connection link from S200 to S300 and S400.

[0072] S230. Assemble the measurement point types and bind the field calibers of the valve group mapping matrix to generate a monitoring orchestration package.

[0073] Specifically, this step takes over the valve group mapping matrix output by S220. The valve group mapping matrix already includes valve group identifier, branch number, pressure control unit index, pressure control tunnel number, pressure control sector number, branch pre-embedded location entry and topology version number, and is associated with the constraint version number of the pressure budget table through a version anchoring field.

[0074] The measurement point type assembly in this step is performed by the measurement point type assembly and field caliber binding module, which consists of a measurement point template assembly unit, an installation pose mapping unit, an acquisition channel planning unit, a field caliber binding unit, a time reference registration unit, and a quality mark assembly unit.

[0075] The measuring point template assembly unit assembles a set of measuring point types that satisfy closed-loop observability based on the mapping relationship between the pressure control unit index and the branch number. The set of measuring point types includes at least three categories: seepage pressure measuring points, flow measuring points, and opening status measuring points. Seepage pressure measuring points are defined as pore water pressure acquisition measuring points installed on the outside of the lining or the outside of the waterproof layer of the lining. Flow measuring points are defined as flow acquisition measuring points installed at the metering interface of the controllable pressure relief branch. Opening status measuring points are defined as status acquisition measuring points installed at the actuator end of the adjustable unit to characterize the opening position or action status. The measuring point template assembly unit can also be equipped with differential pressure measuring points or turbidity measuring points as preferred extensions. Differential pressure measuring points are used to characterize the local pressure drop of the branch, and turbidity measuring points are used to characterize the sand content and blockage risk of the drainage. However, the preferred extensions do not change the minimum set structure of the input slot set for subsequent state identification. The minimum set is still composed of seepage pressure segment, flow segment and opening segment.

[0076] Furthermore, the installation pose mapping unit maps the measurement point type to the implementable installation position field according to the branch pre-embedded position entries and pressure control sector numbers in the valve group mapping matrix, and writes them into the measurement point installation position entries.

[0077] In terms of terminology, the measurement point installation location entry is defined as a location description field that includes mileage positioning entries, circumferential positioning entries, structural layer entries, and construction procedure window entries. The mileage positioning entry references the mileage range entry corresponding to the pressure control tunnel section number, the circumferential positioning entry references the circumferential zoning rule entry corresponding to the pressure control sector number, the structural layer entry references the embedded condition entry in the lining structure design parameters, and the construction procedure window entry references the procedure node archive entry in the construction plan.

[0078] The acquisition channel planning unit generates an acquisition channel binding scheme based on the measurement point installation location entries and the channel resource entries of the field data acquisition system. However, this step does not perform the binding; it only generates channel planning fields that can be called by S310. The field data acquisition system can be a Supervisory Control and Data Acquisition (SCADA) system or an equivalent data acquisition terminal network. The first English abbreviations are given in Chinese and English above. The channel resource entries include channel number, sampling granularity entries, synchronization trigger entries, and power supply and protection level entries.

[0079] The field caliber binding unit performs unified caliber binding on the data fields collected by the measurement points. The unified caliber binding includes consistency in field naming, unit consistency, timestamp field caliber consistency, missing measurement mark caliber consistency, and quality mark caliber consistency. The field caliber binding is checked against the terminology dictionary field set formed by S110. If the check passes, it is written into the field caliber binding record entry. If the check fails, the conflict is recorded and the entry to be corrected is output. The entry to be corrected is written into the monitoring orchestration package along with the quality mark for identification in subsequent steps.

[0080] Understandably, this step predefines and writes the frame structure required for "data frame assembly processing" into the monitoring orchestration package, so that it can be directly referenced in the binding of measurement point acquisition channels and data frame assembly processing of S310.

[0081] To this end, the time reference registration unit registers a unified time reference entry in the monitoring orchestration package. The unified time reference entry is defined as the timestamp caliber and synchronization strategy entry followed when the monitoring data frame set is time-aligned across measuring points and valve groups. The synchronization strategy entry includes a synchronization trigger source entry and an alignment window entry. At the same time, the quality mark assembly unit assembles data quality rule entries in the monitoring orchestration package. The data quality rule entries are defined as rule fields for the missing rate, drift discrimination, and abnormal jump discrimination of pressure measuring points, flow measuring points, and opening status measuring points. The rule fields are used for consistency verification of S320 and discrimination actions in the slot assembly stage, but this step only completes the writing of the rule fields.

[0082] To maintain the level of closed-loop automation and auditability, this step also writes an orchestration version number into the monitoring orchestration package. The orchestration version number is associated with the version anchor field of the valve group mapping matrix and the topology version number, so that when valve group reorganization or topology replacement occurs, the monitoring orchestration package has a replayable historical version link.

[0083] Upon completion, the output of this step is a monitoring orchestration package. This package, at the data structure level, includes a set of measurement point types, measurement point installation location entries, data acquisition channel planning fields, field caliber binding record entries, unified time reference entries, data quality rule entries, orchestration version number and version association fields, and the destination of these fields is indicated at the end of this section using natural language. The monitoring orchestration package, as the input to the "monitoring orchestration package" in S310, is used for binding measurement point acquisition channels and assembling data frames to output a set of monitoring data frames. At the same time, the data quality rule entries are called in the consistency check and slot assembly stage in S320. The measurement point type set and field caliber binding record entries are referenced in the control constraint assembly and interlocking rule assembly process in S410 to define the correspondence between the calibers of monitoring data fields and control constraint fields, thereby connecting the closed loop of cross-main steps from S200 to S300 and S400.

[0084] Step S300 includes at least steps S310-S330: S310. Obtain the monitoring arrangement package, perform measurement point acquisition channel binding and data frame assembly processing to obtain the monitoring data frame set.

[0085] Specifically, this step follows the monitoring orchestration package output by the preceding S230. The monitoring orchestration package includes a set of measurement point types, measurement point installation location entries, acquisition channel planning fields, field caliber binding record entries, unified time reference entries, data quality rule entries, orchestration version number and version association fields, and has a version anchoring relationship with the valve group mapping matrix and the controllable drainage network topology.

[0086] During this step, the monitoring orchestration package is read from the engineering data storage medium or configuration management system by the monitoring operation orchestrator. During the reading process, the orchestration version number is validated, and the topology version number pointed to by the version association field is compared for consistency. When a discrepancy is detected between the topology version number and the version of the controllable unit already deployed on-site, the monitoring and orchestration machine writes a version inconsistency event record and enters a waiting state or enters an incremental re-orchestration process. The incremental re-orchestration process updates only the channel resource occupancy relationship between the acquisition channel planning field and the measurement point installation location entry without changing the control unit index set, so that the subsequent acquisition link is consistent with the status of the on-site equipment.

[0087] Furthermore, the measurement point acquisition channel binding and data frame assembly processing is performed by the measurement point acquisition channel binding and data frame assembly module, which consists of a channel resource discovery unit, a binding verification unit, a sampling granularity coordination unit, a time alignment pre-arrangement unit, a data frame skeleton generation unit, and an abnormal acquisition processing unit.

[0088] The channel resource discovery unit discovers channel resources for the field data acquisition system. The input sources for channel resource discovery include a list of channel numbers, sampling granularity entries, synchronization trigger entries, and power supply and protection level entries, and are matched with the acquisition channel planning field in the monitoring orchestration package. After matching is completed, channel binding candidate relationships are formed and written into the binding candidate table fields. The binding verification unit verifies the binding candidate table fields, and the verification content includes the consistency between the measurement point installation location entries and the channel physical wiring list, the consistency between the field caliber binding record entries and the acquisition protocol field mapping, and the compatibility between the synchronization trigger entries and the unified time reference entries. When the verification is successful, the binding verification unit will upgrade the candidate binding table field to a channel binding table field and write a binding confirmation flag. When the verification fails, the binding verification unit writes the binding conflict record field. The binding conflict record field includes the conflict measurement point identifier, conflict channel number, conflict type entry and rollback action entry. The rollback action entry triggers the channel resource discovery unit to rediscover or triggers the generation of on-site maintenance work order. However, this step only writes the record and completes the rollback, and does not enter the external execution of maintenance actions.

[0089] Understandably, the sampling granularity coordination unit performs sampling granularity coordination after the channel binding is completed. The objects of sampling granularity coordination are the pressure measurement points, flow measurement points and opening status measurement points corresponding to the set of measurement point types.

[0090] Since the sampling granularity entries at different measurement points may differ, the sampling granularity coordination unit determines the frame period entries and sub-sampling index entries of the monitoring data frame set based on the alignment window entries in the unified time reference entries, and writes the frame period entries into the frame template field of the data frame skeleton generation unit.

[0091] The time alignment pre-arrangement unit generates time alignment pre-arrangement table fields based on the synchronization trigger entries and alignment window entries. The time alignment pre-arrangement table fields record the sampling time index of each channel number in the frame period entry, and also record the missing measurement placeholders required by the missing measurement mark caliber consistency rule. When the on-site synchronous trigger entry supports hardware triggering, the time alignment pre-programming unit writes the hardware trigger flag; When the on-site synchronous triggering entry only supports software polling, the time alignment pre-arrangement unit writes the polling scheduling flag and generates a polling drift tolerance window entry. The polling drift tolerance window entry is used in conjunction with the abnormal jump discrimination rule in the data quality rule entry to attribute timing jitter to subsequent consistency verification stages.

[0092] Furthermore, the data frame skeleton generation unit generates a data frame skeleton based on the set of measurement point types and field caliber binding record entries in the monitoring orchestration package. The data frame skeleton is defined as a field layout description of a single frame of data within the monitoring data frame set. This field layout description includes a frame timestamp field, a pressure control unit index field, a branch number field, a valve group identifier field, a pressure sampling field, a flow sampling field, an opening status sampling field, a quality marker field, and a missing measurement marker field. The frame timestamp field uses the timestamp caliber specified by the unified time reference entry. The pressure control unit index field and the branch number field reference the index caliber within the valve group mapping matrix. The valve group identifier field references the named caliber within the valve group mapping matrix. The pressure sampling field, flow sampling field, and opening status sampling field reference the field name and unit caliber within the field caliber binding record entries. The quality marker field is equipped with quality label bits according to the data quality rule entries. The missing measurement marker field is equipped with missing measurement placeholders according to the missing measurement rate rules.

[0093] The abnormal acquisition processing unit performs abnormal acquisition processing when there is a momentary interruption, noise spike, or range overflow in the acquisition channel. The abnormal acquisition processing is written to the abnormal acquisition event field in the event logging method. The abnormal acquisition event field is bound to the frame timestamp field and associated with the channel number, thereby supporting the tracing of the source of the abnormality in the subsequent consistency verification stage.

[0094] Regarding the triggering conditions, this step is initiated in two types of triggering methods during the project's operation.

[0095] The first type is periodic triggering, which is driven by frame period entries. The measurement point acquisition channel binding and data frame assembly module generates a monitoring data frame and writes it into the monitoring data frame set when each frame period entry arrives. The second type is event-triggered. Event triggering is caused by abnormal acquisition event fields, binding conflict record fields, or changes in the status of field equipment. When the event level of an abnormal acquisition event field exceeds the preset level or when a change in the status of the equipment causes the channel binding table field to become invalid, the module triggers the incremental binding process and regenerates the data frame skeleton. At the same time, a frame skeleton version mark is written into the monitoring data frame set, so that the same monitoring data frame set can carry multiple versions of frame data and maintain auditability.

[0096] In terms of version management strategy, when the monitoring data frame set is written, it is accompanied by an orchestration version number and a frame skeleton version mark, and maintains a corresponding relationship with the version association field of the monitoring orchestration package; when subsequent steps read the monitoring data frame set to perform consistency checks, the corresponding data quality rule entries can be retrieved based on the orchestration version number, thereby avoiding the drift of the judgment caliber caused by the intersection of different version rules.

[0097] Upon completion, the output of this step is a monitoring data frame set. This monitoring data frame set is called as the input of the "monitoring data frame set" in the subsequent S320. S320 extracts the seepage pressure segment, flow rate segment and opening segment from the monitoring data frame set and performs consistency check and tank assembly. Meanwhile, the abnormal acquisition event field and frame skeleton version marker in the monitoring data frame set are referenced in the subsequent abnormal diagnosis entry execution and downgrade handling entry execution stages of S430 to determine whether the abnormal source belongs to the acquisition side, execution side or environment side, so as to maintain consistency with the recording caliber of the closed-loop update package.

[0098] S320. Extract the pressure segment, flow segment and opening segment from the monitoring data frame set, perform consistency check and tank assembly, and generate a status identification input tank set.

[0099] Specifically, this step receives the monitoring data frame set output by S310. The monitoring data frame set includes a frame timestamp field, a pressure control unit index field, a branch number field, a valve group identifier field, a seepage pressure sampling field, a flow rate sampling field, an opening status sampling field, a quality mark field, a missing measurement mark field, and may also include an abnormal acquisition event field and a frame skeleton version mark.

[0100] This step is performed by the data quality and slot orchestrator. The data quality and slot orchestrator reads the orchestration version number from the monitoring data frame set and backtracks to retrieve the corresponding data quality rule entry. Then, according to the alignment window entry given by the unified time base entry, the monitoring data frame set is divided into several time window segments. The time window segments are written into the time window index field. The time window index field has a reversible mapping relationship with the frame timestamp field, which facilitates the playback of the original frame data by window in the subsequent coupled state update stage.

[0101] Furthermore, the process of extracting pressure, flow, and valve opening segments is completed by the segment extraction unit. Within each time window segment, the segment extraction unit aggregates pressure sampling fields according to the pressure control unit index field to form a pressure segment, aggregates flow sampling fields according to the branch number field to form a flow segment, and aggregates valve opening status sampling fields according to the valve group identifier field to form an opening segment. During aggregation, the quality mark field and the missing measurement mark field are carried simultaneously, and the aggregated segment is written into the segment container field.

[0102] The segment container field is defined as a segment carrying structure that includes a time window index field, a pressure control unit index field, a branch number field, a valve group identifier field, a pressure segment field, a flow segment field, an opening segment field, a segment quality summary field, and a segment missing measurement summary field.

[0103] The fragment quality summary field is obtained by summarizing the quality mark field within the time window. The summarization rules follow the missing test rate and abnormal jump discrimination rules in the data quality rules. The fragment missing test summary field is obtained by summarizing the missing test mark field within the time window. After summarization, a missing test distribution description field is formed. The missing test distribution description field contains the missing test continuous segment position and missing test length entries, which are used by the consistency verification stage to determine whether it belongs to the sampling link momentary interruption or environmental fluctuation.

[0104] Furthermore, the consistency check is performed by the consistency check unit, which includes a caliber consistency check subunit, a timing consistency check subunit, a topology consistency check subunit, and a physical constraint consistency check subunit.

[0105] The consistency inspection subunit performs consistency inspection on the pressure segment field, flow segment field, and opening segment field within the segment container field. The inspection is based on the record entries bound to the field consistency, including unit consistency, range consistency, and missing measurement placeholder consistency. When a segment is found to have inconsistent unit calibers, the caliber consistency check sub-unit writes the caliber conflict flag field and transfers the segment to the correction queue field. The correction queue field is also written to the metadata area of ​​the state recognition input slot set, so that the subsequent coupled state update stage can identify and choose to ignore or use alternative mapping.

[0106] The timing consistency check subunit performs timing consistency checks on the segment sampling time index in the time window index field. The check is based on the polling drift tolerance window entry or hardware trigger flag in the unified time base entry, including sampling time offset judgment, duplicate sampling judgment and timestamp reverse order judgment. When a timestamp reversal or offset is detected to exceed the tolerance window entry, the timing consistency check subunit writes the timing anomaly flag field and appends the timing anomaly source entry to the fragment quality summary field. The source entry can point to the channel number or the synchronization trigger entry, which facilitates the attribution of subsequent anomaly diagnosis entry execution phase.

[0107] The topology consistency check subunit performs a topology consistency check on the pressure control unit index field, branch number field, and valve group identifier field within the segment container field. The topology consistency check is based on the version anchoring field between the valve group mapping matrix and the controllable drainage network topology, including whether the branch number exists, whether the pressure control unit index exists, and whether the attribution relationship between the valve group identifier and the branch number is consistent. When a fragment of data is detected to originate from an unknown branch number or an unknown valve group identifier, the topology consistency check subunit writes the unknown node marker field and generates a topology rollback action entry. The topology rollback action entry triggers the data quality and slot orchestrator to backtrack and read the latest channel binding table field and regenerate the fragment container field. However, this step does not change the original data of the monitoring data frame set; it only records the rollback action entry in the status identification input slot set.

[0108] The physical constraint consistency verification subunit performs physical constraint consistency verification on the pressure segment field, flow segment field, and opening segment field. The physical constraint consistency verification is based on the external water pressure change rate constraint entry and the longitudinal pressure gradient constraint entry in the pressure budget table, and also refers to the action state change of the opening state sampling field. This includes consistency verification of the pressure change rate with the opening action direction, consistency verification of the flow rate change with the opening change, and consistency verification of the pressure difference change between adjacent pressure control tunnel segments with the longitudinal pressure gradient constraint entry. When a significant pressure change rate is detected but the opening state sampling field shows no action state change, the physical constraint consistency verification subunit writes an action mismatch flag field. The action mismatch flag field can trigger sensor abnormality candidate flags or actuator abnormality candidate flags in the subsequent coupling state update stage, but this step still retains the original segment and records the mismatch entry.

[0109] Understandably, slot assembly is performed after the conformity check is completed, and slot assembly is completed by the slot assembly unit.

[0110] The slot assembly unit, based on the minimum set requirements for status identification, establishes an association assembly relationship between the pressure segment field, flow segment field, and opening degree segment field within each time window index field, according to the pressure control unit index field and the branch number field, and generates a status identification input slot set. The status identification input slot set is defined as a slot-based data structure containing the time window index field, pressure control unit index field, branch number field, valve group identifier field, pressure segment field, flow segment field, opening degree segment field, consistency check summary field, tag set field, and version reference field.

[0111] The consistency check summary field records the summary of the judgment conclusions for consistency checks of caliber, timing, topology, and physical constraints; the tag set field records caliber conflict tag fields, timing anomaly tag fields, unknown node tag fields, action mismatch tag fields, etc. The version reference field records the reference values ​​of the orchestration version number, frame skeleton version tag, topology version number, and valve group mapping matrix version anchor field, so that the slotted data structure can maintain a traceable relationship with the upstream configuration.

[0112] Regarding the triggering conditions, this step triggers slot assembly by rolling the time window index field. The rolling method is determined by the window length and window shift entries specified by the alignment window entries. When there is an abnormal acquisition event field in the monitoring data frame set and the event level exceeds the preset level, the data quality and slot orchestrator marks the time window index field as a high-concern window entry and adds a high-concern window mark field to the status identification input slot set. The high-concern window mark field can trigger a higher frequency of status updates or trigger a bypass processing branch in the subsequent coupled status update stage.

[0113] In terms of exception handling, when the missing measurement distribution description field of the pressure segment field in a certain time window index field shows that the continuous missing measurement length exceeds the threshold of the data quality rule entry, the slot assembly unit marks the slot entry corresponding to the pressure control unit index field as the missing measurement dominant slot and writes the missing measurement dominant mark in the mark set field. The subsequent coupled state update phase can switch to an alternative estimation strategy or reduce the weight of the slot in the state vector packet update based on the missing data dominant flag, but this step only completes the flag writing and keeps the slot structure intact.

[0114] Upon completion, the output of this step is a state recognition input slot set. This state recognition input slot set is called as the input for the subsequent S330. S330 performs coupled state update processing on the state recognition input slot set and generates a state vector package. Meanwhile, the consistency check summary field and tag set field in the state recognition input slot set are referenced in the subsequent S430 abnormal diagnosis item execution and degradation handling item execution stages. This is used to bind the abnormal diagnosis item with the discrimination caliber at the slot level, thereby maintaining consistency with the record structure of the closed-loop update package, and providing the input basis for state credibility for the control constraint assembly and interlocking rule assembly processing of S410.

[0115] S330. Perform coupled state update processing on the state recognition input slot set to generate a state vector package.

[0116] Specifically, this step receives the status identification input slot set output by S320. The status identification input slot set includes a time window index field, a pressure control unit index field, a branch number field, a valve group identifier field, a pressure segment field, a flow segment field, an opening segment field, a consistency check summary field, a tag set field, and a version reference field.

[0117] This step is executed by the coupled state update module, which consists of a state definition unit, a prior configuration loading unit, a slot-driven update unit, a constraint consistency unit, a credibility fusion unit, a state persistence unit, and a playback audit unit.

[0118] The state definition unit defines the field structure of the state vector package. In terms of terminology, the state vector package is defined as a state carrying structure that includes a time window index field, a pressure control unit index field, a groundwater recharge boundary state field, a permeability parameter zoning state field, a pressure relief branch blockage state field, a state credibility field, an update source summary field, and a version reference field. The groundwater recharge boundary state field is used to characterize the state description of the change in water head at the outer boundary of the tunnel segment. The permeability parameter zoning state field is used to characterize the state description of the change in permeability parameters corresponding to the pressure control tunnel segment number or the pressure control unit index field. The pressure relief branch blockage state field is used to characterize the state description of the change in the effectiveness of the channel corresponding to the branch number. The State Credibility field is used to characterize the availability of the state update to the control link. The Update Source Summary field is used to record a summary of the slot entry set and tag set fields used in this update. The Version Reference field inherits the Version Reference field of the state identification input slot set and appends the state model version number of this step.

[0119] Furthermore, the prior configuration loading unit reads the coupled state update prior configuration package from the engineering configuration management system. The coupled state update prior configuration package includes initial state snapshot entries, state update strategy entries, state constraint entries, anomaly degradation strategy entries, and state model version number.

[0120] Initial state snapshot entries can be derived from geological survey data during the design phase, grouting ring construction records, blockage sensitivity classification of drainage structure reservoirs, and archived data from trial operation monitoring; state update strategy entries specify the update frequency, update step size, and noise suppression entries for trench-driven updates. The state constraint entries specify the value boundaries and change rate boundaries of the groundwater recharge boundary state field, the permeability parameter zoning state field, and the pressure relief branch blockage state field. The change rate boundaries have a consistent mapping relationship with the external water pressure change rate constraint entries in the pressure budget table. The abnormal degradation strategy entry specifies the degradation update path when the tag set field contains a missing dominant tag, an action mismatch tag, or an unknown node tag; The state model version number is used to ensure consistency with the rule version for subsequent control constraint assembly and interlocking rule assembly processing.

[0121] Furthermore, the slot-driven update unit performs a slot-driven update on the index field of each time window within the state recognition input slot set.

[0122] The inputs for slot-driven updates are the pressure segment field, flow segment field, and aperture segment field, and refer to the consistency check summary field and the tag set field. The update unit first performs quality gating on the osmotic pressure segment field. The quality gating is based on the time sequence anomaly marker field and the caliber conflict marker field in the consistency inspection summary field. When the caliber conflict marker field exists, the slot entry is marked as unusable for updating the slot and written to the update source summary field. When a timing anomaly flag exists but does not exceed the preset level, the slot entry is marked as a de-weighted update slot, and the credibility component in the status credibility field is reduced.

[0123] Subsequently, the update unit performs action association gating on the flow segment field and the opening segment field. The action association gating is based on the action mismatch flag field. When the action mismatch flag field exists, the update unit generates an actuator abnormality candidate flag and writes it into the update source summary field. At the same time, the update of the pressure relief branch blockage status field adopts a conservative path. The conservative path adopts the update rule of maintaining the blockage status or slowly increasing it, so that the blockage status changes do not have abrupt changes.

[0124] For slot entries with missing dominant markers, the update unit calls the alternative estimation path in the abnormal degradation strategy entry. The alternative estimation path can be inferred from the seepage pressure segment field of the adjacent pressure control unit index field, or from the flow rate segment field and opening degree segment field of the same valve group identifier field. The alternative estimation source is written into the update source summary field, and the alternative estimation confidence component is marked in the status confidence field.

[0125] Understandably, the coupled state update process is a process of converting monitoring trench layer data into state layer data that the control link can directly consume. In this step, the three types of states—groundwater recharge boundary state field, permeability parameter zoning state field, and pressure relief branch blockage state field—are jointly updated within the same time window index field, and the constraint consistency unit performs consistency processing on the joint update results.

[0126] The constraint unification unit reads the state constraint entries and performs boundary pruning and change rate pruning on the three types of state fields. The pruning action is written to the pruning record field and merged into the update source summary field to prevent non-physical jumps in state updates under noise spikes or local missing measurement conditions. At the same time, the constraint unification unit establishes a mapping between the state change rate boundary and the external water pressure change rate constraint entry in the pressure budget table. The mapping relationship is written into the constraint reference field of the state vector package, so that when the control constraint package is assembled in S410, the state layer constraints can be directly referenced without repeated mapping.

[0127] The credibility fusion unit performs fusion calculations on the state credibility field. The fusion input includes the fragment quality summary field of the slot entry, the missing test distribution description field, the tag set field, and the frame skeleton version tag in the reference version reference field. The fusion output is written to the status credibility field, which also includes a credibility source entry. The credibility source entry records the main influencing factors as belonging to the categories of missing test, timing, caliber, topology, or action mismatch, thereby supporting the rapid location of the problem category during the subsequent anomaly diagnosis entry execution stage.

[0128] Furthermore, the state persistence unit writes the state vector package into the state database and establishes an index relationship between the time window index field and the state snapshot entries. The state snapshot entries are bound to the state model version number and written into the persistence version field. When subsequent steps require replaying a control event, the replay audit unit can retrieve the corresponding state update strategy entry and state constraint entry based on the persistent version field, and replay the state identification input slot set metadata area at that time, thereby forming the auditability of the closed-loop iterative process.

[0129] Regarding automated triggering conditions, this step supports two triggering methods: periodic triggering and exception triggering. Periodic triggering is consistent with the S320's time window index field scrolling, triggering slot-driven updates and forming a state vector package after each time window index field is formed; Anomaly triggering is initiated by the high-concern window marker field of the state recognition input slot set. When the high-concern window marker field exists and the consistency check summary field shows a significant deviation in the physical constraint consistency check, the coupled state update module can perform an additional update within the same time window index field and write to the additional update marker field. The additional update marker field is bound to the update source summary field, which is used by the subsequent interlocking rule assembly processing stage to identify the state as a high-risk state snapshot.

[0130] Upon completion, the output of this step is a state vector package. The state vector package is called as the input of the "state vector package" in the subsequent S410. S410 obtains the state vector package and assembles it together with the pressure budget table into a control constraint package. Then, S420 extracts the pressure control unit index from the control constraint package and uses the valve group mapping matrix to execute valve group collaborative control decision to generate a valve group opening instruction set. Meanwhile, the state credibility field, update source summary field, and constraint reference field in the state vector package are referenced in the subsequent S430 abnormal diagnosis item execution and downgrade disposal item execution stages. They are used to bind the diagnosis conclusion with the state update caliber and write it into the closed-loop update package, thereby forming a traceable closed-loop link with the upstream monitoring orchestration package, monitoring data frame set, and state identification input slot set.

[0131] Step S400 includes at least steps S410-S430: S410. Obtain the state vector package and the pressure budget table, and perform control constraint assembly and interlocking rule assembly processing to obtain the control constraint package.

[0132] Specifically, this step receives the state vector package output by S330 and connects it to the pressure budget table output by S130. The state vector package includes at least the time window index field, the pressure control unit index field, the permeability parameter zoning status field, the groundwater recharge boundary status field, the pressure relief branch blockage status field, the status reliability field, the update source summary field, and the version reference field. The pressure budget table includes at least the pressure control tunnel segment number, the pressure control sector number, the pressure control unit index field, the lining safety constraint item, the surrounding rock stability constraint item, the external water pressure change rate constraint item, the longitudinal pressure gradient constraint item, and the constraint item caliber field.

[0133] This step is performed by the control constraint assembly and interlocking rule assembly module, which consists of a constraint assembler, a state mapper, an interlocking rule orchestrator, a gating decision maker, a conflict resolver, a version anchor, and an audit logger. The constraint assembler assembles the constraint entries in the pressure budget table into a constraint expression structure that can be invoked by the controllable link. The state mapper maps the state fields in the state vector package to the state reference slots of the constraint expression structure. The interlocking rule orchestrator assembles interlocking rule entries and generates an executable interlocking rule graph. The gating decision maker performs access control based on the state confidence field and the constraint entry caliber field. The conflict resolver handles priority conflicts when multiple constraints are executed on the same object. The version anchor registers the version number of the control constraint package and establishes an association with the version reference field. The audit logger records the input version, assembly action, and output version of the assembly process in this step.

[0134] Furthermore, after the constraint assembler reads the pressure budget table, it first performs a consistency check on the constraint item caliber field. The check includes the consistency of the naming caliber of the pressure control unit index field, the granularity caliber of the pressure control hole segment number, and the granularity caliber of the pressure control sector number with the index caliber of the pressure control unit mesh generated by S120. When a discrepancy is detected, the constraint assembler writes the discrepancy conflict event field and triggers a backtracking action. The backtracking action points to the terminology registration record and field mapping record of the engineering input package to locate the source of the difference. However, this step only writes the record and enters the conservative assembly branch.

[0135] Subsequently, the constraint assembler assembles the lining safety constraint items and surrounding rock stability constraint items into a static constraint area of ​​the constraint expression structure, and assembles the external water pressure change rate constraint items and longitudinal pressure gradient constraint items into a dynamic constraint area of ​​the constraint expression structure. The static constraint area includes a constraint object field, a constraint boundary field, a constraint priority field, and a constraint source field. The dynamic constraint area includes a time window index field, a status reference slot field, a change rate threshold field, a gradient threshold field, a trigger condition field, and a release condition field. The constraint object field is used to point to the pressure control unit index field or the valve group identifier field. The constraint boundary field is used to describe the allowed seepage pressure segment change, flow segment change, or opening change boundary within the range of the pressure control unit index field. The trigger condition field is used to describe the timing of dynamic constraint activation, and the release condition field is used to describe the timing of dynamic constraint deactivation. All of the above fields are written into the constraint source field during assembly. The constraint source field records at least the pressure budget table version number and the constraint item row number index for easy subsequent audit playback.

[0136] Furthermore, the state mapper accesses the state vector package, maps the permeability parameter partition state field, the groundwater recharge boundary state field, and the pressure relief branch blockage state field to the state reference slot field of the dynamic constraint area, and maps the state credibility field to the credibility gating field of the gating decision-maker; During the mapping process, the state mapper establishes a one-to-one mapping relationship based on the pressure control unit index field and a one-to-many mapping relationship based on the branch number or valve group identifier field. This allows multiple controllable pressure relief branches under the same pressure control unit index field to share the same groundwater recharge boundary state field but have their own pressure relief branch blockage state field.

[0137] To support the subsequent assembly of interlocking rules, the state mapper also generates a state snapshot summary field. The state snapshot summary field contains the time window index field, the control unit index field, the summary value of the state credibility field, and the summary value of the update source summary field, and writes it into the state reference summary area of ​​the control constraint package. The state reference summary area serves as the input to the interlocking rule orchestrator, used for interlocking rule start / stop determination and conflict resolution determination.

[0138] Understandably, the interlocking rule orchestrator assembles interlocking rule entries after the constraint expression structure is completed. These interlocking rule entries are derived from the interlocking rule library and the engineering configuration management system. The interlocking rule library is associated with the drainage structure library, and the interlocking rule entries have an index reference relationship with the valve group mapping matrix.

[0139] The interlocking rule orchestrator parses interlocking rule entries into an interlocking rule graph, which includes rule node fields, condition node fields, action node fields, and fallback node fields. The rule node field records the rule identifier and rule version number; the condition node field references the trigger condition field of the dynamic constraint area and the status reference slot field; the action node field references the valve group identifier field and defines the allowed opening change direction and upper limit of the change range; the backoff node field defines the degradation action when the status confidence field is lower than the gate threshold or when the caliper conflict event field exists.

[0140] After generating the interlocking rule diagram, the interlocking rule orchestrator writes it into the interlocking rule area of ​​the control constraint package and generates an interlocking rule index field. The interlocking rule index field is used by the S420 to quickly locate the set of executable rules when making valve group coordinated control decisions.

[0141] Furthermore, the gating decision maker performs admission control on the control constraint package. The admission control input includes a state confidence field, a caliber conflict event field, a state reference summary area, and a constraint entry caliber field. When the state confidence field is lower than the gating threshold, the gating decision maker writes the gating rejection flag field and triggers the backoff node field of the interlocking rule graph, tightening the upper bound of the allowable change range of the action node field to a conservative value, and simultaneously writes the gating reason field. The gating reason field references the summary value of the update source summary field and records the rejection time window index field.

[0142] After gating is completed, the conflict resolver handles constraint conflicts. The main constraint conflicts are that the same valve group identification field is simultaneously constrained by multiple dynamic constraint zone action boundaries, or the same pressure control unit index field simultaneously hits different priority boundaries of lining safety constraint entries and surrounding rock stability constraint entries. The conflict resolver performs priority adjudication based on the constraint priority field and rule version number, and writes the adjudication result to the conflict resolution record field. The conflict resolution record field includes the conflict object field, the candidate constraint set summary field, the adjudication constraint identifier field, and the adjudication basis field.

[0143] At the end of this step, the version anchorer generates a control constraint package version number and registers a version anchor field. The version anchor field is associated with the version reference field of the state vector package and the version number of the pressure budget table. At the same time, the range of the assembly time window index field is written to facilitate the location of the effective interval when the closed-loop update package is written back.

[0144] Upon completion, the output of this step is a control constraint package. This control constraint package is called as the input of the "control constraint package" in subsequent S420 and serves as the rule reference basis when executing the abnormal diagnosis items in S430. At the same time, the state reference summary area in the control constraint package can be traced back to the state vector package of S330 in the cross-step link for playback auditing of the audit logger.

[0145] S420. Extract the pressure control unit index and valve group mapping matrix from the control constraint package, make valve group collaborative control decisions, and generate valve group opening instruction set.

[0146] Specifically, this step takes over the control constraint package output by S410 and references the valve group mapping matrix generated by the preceding S220; The control constraint package includes at least a static constraint area, a dynamic constraint area, an interlocking rule area, an interlocking rule index field, a gate veto flag field, a conflict resolution record field, and a version anchoring field. The valve group mapping matrix includes at least a valve group identifier field, a branch number field, a pressure control unit index field, a grouping flag field, and a mapping relationship registration field.

[0147] This step is executed by the valve group collaborative control decision module, which consists of a control target assembly unit, a candidate action generation unit, a constraint screening unit, a collaborative solution unit, an instruction framing unit, an instruction issuance and arrangement unit, and an execution receipt acquisition unit. The control target assembly unit is used to assemble the set of control objects within the index field of this time window. The candidate action generation unit is used to generate candidate opening change actions according to the valve group identifier field. The constraint filtering unit is used to filter feasible actions according to the control constraint package. The collaborative solution unit is used to output a consistent opening combination under the coupling constraints of multiple valve groups. The instruction framing unit is used to convert the opening combination into an instruction frame. The instruction issuance and arrangement unit is used to arrange the issuance order according to communication and execution resources. The execution receipt acquisition unit is used to acquire valve group execution receipts and generate a receipt summary field for use in subsequent closed-loop update packages.

[0148] Furthermore, the control target assembly unit reads the set of dynamic constraint area entries corresponding to the index field of this time window from the control constraint package, and extracts the control unit index field to form a set of control objects; Simultaneously, the control target assembly unit reads the mapping relationship registration field from the valve group mapping matrix, maps the control object set to the valve group identifier field set and the branch number field set, and generates a valve group control context package. The valve group control context package includes a time window index field, a pressure control unit index field, a valve group identifier field, a branch number field, a constraint reference field, and a rule reference field. The constraint reference field references the state reference slot field and the rate of change threshold field of the dynamic constraint area, and the rule reference field references the interlocking rule index field pointing to the interlocking rule diagram node.

[0149] The candidate action generation unit generates a candidate action set based on the valve group control context package. The candidate action set consists of several action entries, and each action entry includes at least the action object field, action direction field, action amplitude field, action timestamp field, and action basis field. The action object field points to the valve group identifier field, the action direction field describes the discrete direction of opening increase or decrease, the action amplitude field is jointly limited by the valve group equipment capacity parameters and the upper limit of the allowable change amplitude of the action node field, and the action basis field references the adjudication constraint identifier field of the dynamic constraint area trigger condition field and the conflict resolution record field.

[0150] Understandably, the constraint filtering unit performs a feasibility determination on the candidate action set. The determination input includes a static constraint area, a dynamic constraint area, a gating veto flag field, an interlocking rule diagram, and a conflict resolution record field.

[0151] The determination of the static constraint zone mainly targets the boundary constraints formed by the lining safety constraint items and the surrounding rock stability constraint items. The constraint screening unit projects the action items onto the pressure control unit index field and checks whether the seepage pressure change trend and longitudinal gradient change trend caused by the action are compatible with the constraint boundary field. The determination of the dynamic constraint zone mainly targets the time-varying constraints formed by the external water pressure change rate constraint item and the longitudinal pressure gradient constraint item. The constraint filtering unit reads the status reference slot field and combines it with the trigger condition field to determine whether the constraint is enabled. When the constraint is enabled, the change rate threshold field and gradient threshold field are applied to the candidate action item to restrict it. The determination of the gate veto flag field is mainly used to trigger the degradation branch. When the gate veto flag field exists, the constraint filtering unit tightens the action range field to the conservative range defined by the fallback node field and writes the action basis field into the gate reason field summary, so as to facilitate the differentiation between the source of "gate degradation" and "equipment limitation" in the subsequent abnormal diagnosis item execution stage.

[0152] After the filtering is completed, the constraint filtering unit outputs a subset of actionable actions and generates a feasibility summary field. The feasibility summary field contains a summary of the number of action items removed, the category of removal reason field, and the effective constraint identifier field.

[0153] Furthermore, the collaborative solving unit performs collaborative control decisions on a subset of actionable actions. The collaborative solving input includes the adjacency relationships of the valve group grouping label field and the pressure control unit index field, the coupling constraint description of the dynamic constraint region, and the mutual exclusion relationships of the interlocking rule diagram. The collaborative solving unit first constructs valve group collaborative groups according to the valve group grouping label field, and then constructs a candidate opening degree combination space for each valve group collaborative group. Subsequently, the collaborative solver unit performs constraint propagation processing on the combination space. The constraint propagation processing projects the longitudinal pressure gradient constraint entries to the valve group collaborative group boundary of the adjacent pressure control unit index field, and projects the external water pressure change rate constraint entries to the action timestamp field range within the time window index field, thereby eliminating combinations that violate the coupling boundary in advance within the combination space.

[0154] For the mutual exclusion relationship in the interlocking rule graph, the collaborative solving unit parses the mutual exclusion pairs between the rule node field and the action node field. The mutual exclusion pairs are used to restrict the simultaneous issuance of multiple mutually exclusive actions within the same time window index field. When there is a mutual exclusion conflict, the collaborative solving unit references the adjudication basis field of the conflict resolution record field, outputs a single action or outputs time-sharing actions according to the rule priority, and writes them into the time-sharing arrangement field.

[0155] The collaborative solution unit outputs the valve group opening combination results and generates a collaborative solution summary field. The collaborative solution summary field includes the valve group collaborative group identifier, the selected action item set summary, the mutual exclusion resolution summary, and the time-sharing arrangement field summary.

[0156] Furthermore, the instruction framing unit converts the valve group opening combination result into a valve group opening instruction set. The valve group opening instruction set is defined as an instruction set structure containing a time window index field, a valve group identifier field, a target opening field, an action direction field, an action amplitude field, an instruction sequence number field, an instruction version number field, a constraint reference field, and a rule reference field. The target opening field is obtained by superimposing the current opening state sampling field and the action amplitude field. The instruction version number field references the control constraint package version number and appends the instruction version number of this step.

[0157] The instruction issuance and scheduling unit schedules the issuance order based on the time-sharing scheduling field, communication resource status, and valve group equipment response time entry, and writes the scheduling result into the issuance plan field. When the current plan field shows that there is link congestion or equipment busy status in the same valve group coordination group, the instruction issuance and orchestration unit writes the delay flag field and slightly tightens the target opening field, while also writing the delay reason field.

[0158] After the instruction is issued, the execution receipt acquisition unit collects the valve group execution receipt. The execution receipt includes at least the valve group identification field, the receipt timestamp field, the actual opening degree field, the execution status field, and the fault code field. The receipt is then summarized into a receipt summary field and written into the additional area of ​​the valve group opening instruction set. The receipt summary field serves as one of the inputs for the subsequent abnormal diagnosis entries in S430. At the same time, it can be traced back to the opening status sampling field in the monitoring data frame set of S310 in the cross-step link, thus forming a closed-loop record on the execution side.

[0159] Upon completion, the output of this step is a valve group opening instruction set. This valve group opening instruction set is called as the input of the "valve group opening instruction set" in the subsequent S430. At the same time, the constraint reference field and rule reference field in the valve group opening instruction set are used to bind the control decision with the control constraint package version anchor field, which facilitates the closed-loop update package's audit record of version evolution.

[0160] S430. Perform abnormal diagnosis and degradation processing on the valve group opening instruction set to generate a closed-loop update package.

[0161] Specifically, this step receives the valve group opening instruction set output by S420 and connects it to the control constraint package output by S410. At the same time, it backtracks and references the monitoring data frame set output by S310, the state recognition input slot set output by S320, and the relevant summary fields of the state vector package output by S330. The valve group opening instruction set includes at least a time window index field, a valve group identifier field, a target opening field, an instruction sequence number field, an instruction version number field, a constraint reference field, a rule reference field, and a receipt summary field. The control constraint package includes at least an interlocking rule area, a gating veto flag field, a conflict resolution record field, and a version anchoring field.

[0162] This step is performed by the anomaly diagnosis and degradation handling module, which consists of an anomaly diagnosis item executor, evidence aggregator, root cause classifier, degradation handling item executor, handling writeback, closed-loop recorder, and version rollback orchestrator. The anomaly diagnosis item executor is used to perform discrimination according to the diagnosis item library; the evidence aggregator is used to aggregate evidence from the monitoring side, status side, and execution side; the root cause classifier is used to output the anomaly type and attribution summary; the degradation handling item executor is used to perform handling actions when an anomaly exists; the handling writeback is used to write the handling actions back to the instruction link or configuration link; the closed-loop recorder is used to generate closed-loop update packages; and the version rollback orchestrator is used to organize rollback or freeze actions under version anomaly conditions.

[0163] Furthermore, the evidence aggregator performs alignment and aggregation of multi-source evidence according to the time window index field. Monitoring-side evidence comes from the seepage pressure sampling field, flow rate sampling field, opening status sampling field, quality marker field, missing measurement marker field, and abnormal acquisition event field summary in the monitoring data frame set. Status-side evidence comes from the consistency check summary field, marker set field, and high-concern window marker field summary in the status identification input slot set. Execution-side evidence comes from the receipt summary field, fault code field, and delay marker field summary in the valve group opening command set. Rule-side evidence comes from the interlocking rule area and gating veto marker field of the control constraint package.

[0164] The evidence aggregator assembles the aforementioned evidence into a diagnostic evidence package. The diagnostic evidence package includes a time window index field, a valve group identification field, a pressure control unit index field, an evidence summary field, an evidence source field, and a version chain field. The evidence source field records the category of the evidence, whether it comes from the monitoring side, the status side, the execution side, or the rule side. The version chain field records the instruction version number, the control constraint package version number, the status model version number, and the pressure budget table version number, which facilitates the formation of a complete version chain for subsequent closed-loop update packages.

[0165] Understandably, the abnormal diagnostic entry executor executes abnormal diagnostic entries on the diagnostic evidence package. An abnormal diagnostic entry consists of several discrimination rules, and each discrimination rule includes at least a trigger condition field, an evidence matching field, a discrimination output field, and a treatment suggestion field.

[0166] The trigger condition field can reference the high-concern window flag field, the gated rejection flag field, or the fault code field. The evidence matching field is used to match the missing test distribution description field summary, action mismatch flag field summary, timing anomaly flag field summary, or unknown node flag field summary in the evidence summary field. For example, when the receipt summary field shows that the actual opening field has not followed the target opening field for a long time and the fault code field exists, and at the same time the status recognition input slot set flag set field shows the action mismatch flag field summary, the abnormal diagnosis item actuator discrimination output field writes the actuator abnormality category. When the monitoring data frame set quality marker field shows that the abnormal acquisition event field summary appears frequently and the time-series abnormal marker field summary is significant, the abnormal diagnosis item executor determines the output field and writes the acquisition link abnormality category; When the gating veto flag field exists and the conflict resolution record field shows frequent switching of constraint priorities, the exception diagnosis entry executor determines the exception category in the output field of the rule caliber.

[0167] The root cause classifier performs classification and summarization on the discrimination output fields. The classification output is the anomaly classification result field, which includes at least the anomaly type field, the anomaly object field, the attribution evidence field summary, and the confidence summary field. The confidence summary field references the state credibility field and the evidence source field, which facilitates the executor of the downgraded disposal item to select the disposal path.

[0168] Furthermore, the downgraded handling item executor executes the downgraded handling item after the anomaly classification result field is generated. The downgraded handling item includes at least a handling trigger field, a handling action field, a handling scope field, a handling persistence condition field, and a recovery condition field. The handling trigger field references the anomaly type field and the confidence summary field. When the anomaly type field is the actuator anomaly category and the confidence summary field exceeds the threshold, the handling action field adds the valve group identifier field to the isolation list field and triggers the conservative action of the interlock rule diagram rollback node field. At the same time, the handling scope field expands the valve group grouping mark field to the same valve group. The persistence condition field references the continued existence of the fault code field, and the recovery condition field references the receipt summary field showing that the execution status field has returned to normal. If the anomaly type field is a data acquisition link anomaly category, the handling action field marks the slot entry corresponding to the channel number or the measurement point type set as a reduced-weight input, switches the status update strategy entry to the alternative estimation path, and simultaneously writes a conservative amplitude flag on the instruction issuance orchestration unit side, tightening the change amplitude of the target opening field in the subsequent time window index field. If the anomaly type field is a rule caliber anomaly category, the handling action field triggers the version rollback orchestrator to freeze the control constraint package version anchoring field to the previous stable version number, and writes the conflict resolution record field to the freeze reason field. The handling writeback device implements the handling action field to the executable object, including issuing isolation actions to the valve group controller, writing the freeze flag field to the configuration management system, and writing the re-orchestration trigger field to the monitoring operation orchestrator. All of the above writebacks form a writeback record field in the closed-loop recorder, which includes a writeback object field, a writeback timestamp field, a writeback action summary field, and a writeback version chain field.

[0169] Furthermore, at the end of this step, the closed-loop recorder generates a closed-loop update package. The closed-loop update package is defined as a closed-loop record structure containing a time window index field, a valve group opening instruction set summary field, an execution receipt summary field, an anomaly classification result field, a handling action summary field, a write-back record field, a version chain field, and a traceability index field. The valve group opening instruction set summary field references the instruction sequence number field and the valve group identification field set summary; the execution receipt summary field references the actual opening field and the execution status field summary; the anomaly classification result field references the anomaly type field and the anomaly object field; the handling action summary field references the isolation list field and the freeze reason field summary; and the traceability index field records the frame timestamp field range that can be traced back to the monitoring data frame set, the time window index field range that can be traced back to the status identification input slot set, and records the control constraint package version anchor field, thereby supporting the rapid location of the evidence chain during subsequent audit playback.

[0170] When the version rollback orchestrator detects a break in the version chain field or an inconsistency between the version anchor field and the on-site configuration version, it generates a rollback orchestration record field and writes it into the closed-loop update package. At the same time, it writes a reassembly trigger field to the control constraint assembly and interlocking rule assembly module, so that the next time window index field enters the reassembly process. The trigger link matches the recording structure of the S410 version anchor and audit logger, thereby maintaining the traceability of the closed-loop link.

[0171] Upon completion, the output of this step is a closed-loop update package. In the cross-step chain, the closed-loop update package is used to drive the monitoring and arrangement adjustment of the subsequent time window index field, the switching of the status update strategy, the freezing or rollback of the control constraint version, and forms a continuous evidence chain and version chain with the monitoring data frame set, the status identification input slot set, the status vector package, the control constraint package, and the valve group opening instruction set.

[0172] Specifically, in S410, firstly, this step receives the state vector packet output by S330 and connects it to the pressure budget table output by S130; The state vector package includes a time window index field, a pressure control unit index field, a permeability parameter zoning state field, a groundwater recharge boundary state field, a pressure relief branch blockage state field, a state reliability field, an update source summary field, and a version reference field. The pressure budget table includes a pressure control tunnel section number, a pressure control sector number, a pressure control unit index field, lining safety constraint entries, surrounding rock stability constraint entries, external water pressure change rate constraint entries, longitudinal pressure gradient constraint entries, and constraint entry caliber fields.

[0173] This step is executed by the control constraint assembly and interlocking rule assembly module. First, the constraint expression structure is assembled: the constraint assembler reads the pressure budget table and performs a consistency check on the constraint item caliber field. It checks the consistency of the naming caliber of the pressure control unit index field, the granularity caliber of the pressure control hole segment number, and the granularity caliber of the pressure control sector number with the index caliber of the pressure control unit mesh generated by S120. When an inconsistency is found, the caliber conflict event field is written and a backtracking action is triggered. Subsequently, the constraint assembler assembles the lining safety constraint item and the surrounding rock stability constraint item into a static constraint area, and assembles the external water pressure change rate constraint item and the longitudinal pressure gradient constraint item into a dynamic constraint area. The static constraint area contains constraint object field, constraint boundary field, constraint priority field and constraint source field, and the dynamic constraint area contains time window index field reference, state reference slot field, change rate threshold field, gradient threshold field, trigger condition field and release condition field.

[0174] The state mapper accesses the state vector package and maps the permeability parameter partition state field, the groundwater recharge boundary state field, and the pressure relief branch blockage state field to the state reference slot field of the dynamic constraint area. It also maps the state credibility field to the credibility gating field of the gating decision-maker, establishes a one-to-one mapping relationship according to the pressure control unit index field, and establishes a one-to-many mapping relationship according to the branch number or valve group identifier field. It generates a state snapshot summary field and writes it into the state reference summary area of ​​the control constraint package.

[0175] The interlocking rule orchestrator assembles interlocking rule entries and parses them into an interlocking rule diagram, which includes rule node fields, condition node fields, action node fields, and backoff node fields. The condition node fields reference the trigger condition fields of the dynamic constraint area and the status reference slot fields. The action node fields reference the valve group identifier fields and define the allowed opening change direction and upper limit of the change range. The backoff node fields define the degradation action.

[0176] The gating decision maker performs admission control on the control constraint package. The input includes the state confidence field, the caliber conflict event field, the state reference summary area, and the constraint item caliber field. When the state confidence field is lower than the gating threshold, the gating rejection flag field is written and the fallback node field is triggered.

[0177] The conflict resolver handles constraint conflicts, performs priority decisions based on the constraint priority field and rule version number, and writes the results to the conflict resolution record field. The version anchorer generates the control constraint package version number and registers it in the version anchor field.

[0178] In this process, to transform the continuous state fields in the state vector package into discrete indices that can be used for constraint assembly, Formula ① is introduced to calculate the weighted fusion value of state reliability. Formula ① is defined as: in, This indicates the total number of index fields for the pressure control unit; The traversal index for the pressure control unit; Indicates the first The status reliability field value corresponding to each pressure control unit index; Indicates the weighting coefficient; The output metric represents the weighted value of the credibility of the merged state.

[0179] Data source mapping: Extract the state credibility field from the state vector package and denot it as... The evidence source category is extracted from the updated source summary field and converted into... Together, they form the formula ① .

[0180] Based on the output of Formula ①, in order to further transform static and dynamic constraints into executable mathematical forms, Formula ② is introduced to construct a standardized expression for the constraint boundary.

[0181] Formula ② is defined as: in, Indicates the index of the constraint entry; Indicates the first The original constraint values ​​corresponding to each constraint entry; and They represent the first Minimum and maximum allowed values ​​for each constraint entry; The output function represents the standardized constraint values.

[0182] Data source mapping: Extract the constraint boundary field from the aforementioned pressure budget table and denot it as... Extract the constraint entry caliber field and convert it to and Together, they form the formula in ②. Formula ② directly uses the formula from formula ①. As a gating condition: when hour, Used for normal constraint assembly; when When using conservative values This is to indicate a demotion.

[0183] Formula ② As input to the constraint expression structure, it is used for subsequent collaborative control decisions in S420. The output of this step is a control constraint package, which includes a static constraint area, a dynamic constraint area, an interlocking rule area, a gating veto flag field, a conflict resolution record field, and a version anchoring field, and is called by the "control constraint package" input in subsequent S420.

[0184] Furthermore, the pressure control unit index and valve group mapping matrix are extracted from the control constraint package to make valve group collaborative control decisions and generate a valve group opening instruction set.

[0185] Specifically, this step receives the control constraint package output by S410 and references the valve group mapping matrix generated in the preceding S220. The control constraint package includes a static constraint area, a dynamic constraint area, an interlocking rule area, an interlocking rule index field, a gate veto flag field, a conflict resolution record field, and a version anchoring field. The valve group mapping matrix includes a valve group identifier field, a branch number field, a pressure control unit index field, a grouping flag field, and a mapping relationship registration field. This step is executed by the valve group collaborative control decision module. The control target assembly unit reads the set of dynamic constraint area entries corresponding to the current time window index field from the control constraint package and extracts the pressure control unit index field to form a control object set. At the same time, it reads the mapping relationship registration field from the valve group mapping matrix and maps the control object set to the valve group identifier field set and the branch number field set to generate a valve group control context package.

[0186] The candidate action generation unit generates a set of candidate actions based on the valve group control context package, including action object field, action direction field, action amplitude field, action timestamp field, and action basis field. The constraint filtering unit performs a feasibility determination on the candidate action set. The input includes static constraint area, dynamic constraint area, gating veto flag field, interlocking rule diagram, and conflict resolution record field. The determination of the static constraint area mainly targets the boundary constraints formed by the lining safety constraint item and the surrounding rock stability constraint item. The determination of the dynamic constraint area mainly targets the time-varying constraints formed by the external water pressure change rate constraint item and the longitudinal pressure gradient constraint item.

[0187] The collaborative solution unit performs collaborative control decisions on the subset of actionable actions, constructs valve group collaborative groups according to the valve group grouping label field, and constructs a candidate opening combination space for each valve group collaborative group. It performs constraint propagation processing to project the longitudinal pressure gradient constraint entries to the valve group collaborative group boundary of the adjacent pressure control unit index field, projects the external water pressure change rate constraint entries to the action timestamp field range within the time window index field, and parses the mutual exclusion relationship of the interlocking rule graph.

[0188] The instruction framing unit converts the valve group opening combination result into a valve group opening instruction set. The instruction issuance and arrangement unit arranges the issuance order according to the time-division arrangement field, communication resource status, and valve group equipment response time entries. The execution receipt acquisition unit collects the valve group execution receipts and generates a receipt summary field. In this process, in order to transform constraints into optimizable objectives, formula ③ is introduced to define the objective function of valve group coordinated control.

[0189] Formula ③ is defined as: in, Indicates the number of valve group coordination groups; For the traversal index of the valve group coordination group; Indicates the first The current opening status value of each valve group coordination group; Indicates the first Reference opening value for each valve group in coordination; Indicates the rate of change of opening degree; and These are the weighting coefficients; The output objective function value represents the weighted sum of control deviation and variation penalty, used for minimizing optimization of the collaborative solution unit.

[0190] Data source mapping: Extract the opening status sampling field from the monitoring data frame set and record it as... The state reference slot field is extracted from the control constraint package and converted into... Calculated from the action timestamp field Determined by the constraint priority field and Together they form the formula ③ Formula ③ directly uses the formula from formula ②. As a constraint: In optimization, it is required that... To ensure that constraints are satisfied.

[0191] Based on the objective function of formula ③, formula ④ is introduced to solve the optimization problem of the optimal opening combination. Formula ④ is defined as: in, Let be the decision variable, representing the th The target opening value of each valve group in coordination; and These represent the minimum and maximum allowable values ​​for the opening, respectively; This represents the maximum permissible rate of change in opening degree; the optimization problem aims to minimize... At the same time, the constraints are satisfied.

[0192] Formula ④ uses formula ③ As the objective function, and directly referencing formula ② Integration constraints: when At that time, tightening constraints such as The solution method employs a quadratic programming algorithm, outputting the optimal value. value.

[0193] Data source mapping: The constraint boundary fields are extracted from the control constraint package and denoted as... and Extract the rate of change threshold field and denot it as , and formula ③ Together, they form the optimization problem of formula ④.

[0194] Output of Formula ④ The target opening field of the valve group opening instruction set is used for subsequent S430 anomaly diagnosis. The output of this step is the valve group opening instruction set, which includes the time window index field, valve group identifier field, target opening field, instruction sequence number field, instruction version number field, constraint reference field, rule reference field, and receipt summary field, and is called by the "valve group opening instruction set" input of the subsequent S430.

[0195] Furthermore, the valve group opening instruction set is processed by executing abnormal diagnosis entries and downgrade handling entries to generate a closed-loop update package.

[0196] Specifically, this step receives the valve group opening instruction set output by S420 and connects it to the control constraint package output by S410. At the same time, it backtracks and references the monitoring data frame set output by S310, the status identification input slot set output by S320, and the relevant summary fields of the status vector package output by S330. The valve group opening instruction set includes a time window index field, a valve group identifier field, a target opening field, an instruction sequence number field, an instruction version number field, a constraint reference field, a rule reference field, and a receipt summary field. The control constraint package includes an interlocking rule area, a gating veto flag field, a conflict resolution record field, and a version anchoring field.

[0197] This step is performed by the anomaly diagnosis and degradation handling module. The evidence aggregator performs alignment and aggregation of multi-source evidence according to the time window index field. The monitoring side evidence comes from the seepage pressure sampling field, flow sampling field, opening status sampling field, quality mark field, missing measurement mark field and anomaly acquisition event field summary of the monitoring data frame set. The status side evidence comes from the consistency check summary field, mark set field and high concern window mark field summary of the status identification input slot set. The execution side evidence comes from the receipt summary field, fault code field and delay mark field summary of the valve group opening command set. The rule side evidence comes from the interlocking rule area and gate veto mark field of the control constraint package. They are assembled into a diagnostic evidence package.

[0198] The abnormal diagnosis item executor executes abnormal diagnosis items on the diagnostic evidence package. The discrimination rules include trigger condition fields, evidence matching fields, discrimination output fields, and treatment suggestion fields.

[0199] The root cause classifier outputs anomaly classification result fields. The demotion action executor executes demotion action entries after the anomaly classification result fields are generated. The action action fields include the isolation list field, conservative action trigger, and demotion input flags.

[0200] The write-back unit writes the action back to the instruction or configuration link. The closed-loop recorder generates a closed-loop update packet. During this process, to quantify the degree of anomaly, formula ⑤ is introduced to calculate the anomaly confidence index.

[0201] Formula ⑤ is defined as: in, Indicates the number of evidence types; The traversal index for the evidence type has a range of values. ; Indicates the first Abnormal scoring of similar evidence; Indicates the weight of evidence; The output metric represents the anomaly confidence level.

[0202] Data source mapping: Extract the evidence summary field from the diagnostic evidence package and denot it as... Extract the evidence source field and convert it to Together, they form the formula in ⑤. Formula ⑤ directly uses the output of Formula ④. For reference: when When there is a large deviation between the actual opening degree field and the reply summary field, adjust... Increase. Formula ⑤ As an input for anomaly diagnosis, it is used for subsequent decision-making in Formula ⑥.

[0203] Based on the anomaly confidence level in Formula ⑤, Formula ⑥ is introduced to determine the mathematical expression for the downgrade action.

[0204] Formula ⑥ is defined as: in, Indicates a sign for handling the action. This indicates that a downgraded procedure (such as that for an isolation valve assembly) is being performed. This indicates normal operation; Indicates a dynamic threshold; Basic threshold; For adjustment coefficients; The output from formula ① represents the weighted value of state credibility; formula ⑥ aims to... The action is triggered at that time.

[0205] Data source mapping: output of formula ⑤ With formula ① Joint calculation And thus decide Formula ⑥ directly uses the formula from formula ⑤. As input, and referring to formula ① Achieve dynamic adjustment. Output of Formula ⑥ As a summary field of the action to be processed in the closed-loop update package, it is used to drive the adjustment of the subsequent time window index field.

Claims

1. A method for dynamic control of external water pressure and design of drainage system for deeply buried water conveyance tunnels, characterized in that, include: S100: Obtain the partition information field set, structural parameter field set, and water resource constraint field set; perform multi-source boundary merging and priority adjudication processing; and generate a pressure budget table. S200. Based on the pressure budget table, perform valve grouping and mapping relationship registration processing, and perform measurement point type assembly and field caliber binding processing to generate a monitoring arrangement package; S300. Based on the monitoring and orchestration package, perform data quality inspection, slot assembly, and coupling state update processes to generate a state vector package. S400. Based on the state vector package, perform control constraint assembly and interlocking rule assembly processing, and execute valve group collaborative control decision processing to generate a closed-loop update package.

2. The method according to claim 1, characterized in that, The process of multi-source boundary merging and priority adjudication includes: The multi-source boundary merging and priority adjudication process includes extracting hydrogeological zoning boundaries from the zoning information field set, extracting lining joint and structural change boundaries from the structural parameter field set, extracting sensitive section boundaries that require limiting drainage intensity from the water resource constraint field set, and setting priority adjudication rules to merge conflicting boundaries.

3. The method according to claim 1, characterized in that, The process of registering valve grouping and mapping relationships includes: The valve group grouping and mapping relationship registration process includes generating valve group identifiers according to grouping rules and assigning branch numbers to the corresponding valve group identifiers. The grouping rules include constraints such as the inability to cross the boundary of the pressure control tunnel, the independent addressability of the pressure control sector, and the operation and maintenance reachability constraints. The process also includes performing intra-group consistency checks and cross-tunnel boundary checks to generate a valve group mapping matrix containing valve group identifiers, branch numbers, and pressure control unit indexes.

4. The method according to claim 1, characterized in that, The process of assembling measurement point types and binding field specifications includes: The process of assembling measurement point types and binding field calibers includes assembling pressure measurement points, flow measurement points, and opening status measurement point types. It maps measurement point installation location entries to branch pre-buried location entries and pressure control sector numbers, plans acquisition channels, binds field calibers, registers a unified time reference, and generates a monitoring arrangement package containing a set of measurement point types, measurement point installation location entries, field caliber binding record entries, unified time reference entries, and data quality rule entries.

5. The method according to claim 1, characterized in that, The process of performing data quality inspection and slot assembly includes: The monitoring data frame set is divided into time windows, and the pressure segment, flow segment and aperture segment are extracted. The consistency tests of caliber, time series, topology and physical constraints are performed. The consistency inspection includes consistency of unit diameter, consistency of measurement range, and consistency of missing measurement placeholders. The temporal consistency check includes sampling time offset detection, duplicate sampling detection, and timestamp reversal detection; The topology consistency check includes whether the branch number exists, whether the pressure control unit index exists, and whether the valve group identifier and the branch number are consistent. The physical constraint consistency test includes consistency tests of the rate of change of seepage pressure with the direction of opening action, consistency tests of the change of flow rate with the change of opening, and consistency tests of the change of seepage pressure difference between adjacent pressure control tunnel sections with the longitudinal pressure gradient constraint entries. It generates a state recognition input slot set containing time window index field, seepage pressure segment field, flow rate segment field, opening segment field and tag set field.

6. The method according to claim 1, characterized in that, The process of updating the coupling state includes: The coupled state update process includes loading prior configuration and performing quality gating and action association gating. The quality gating is based on the timing anomaly marker field and the caliber conflict marker field in the consistency inspection summary field; The action association gating is based on the action mismatch flag field; Integrating execution constraint consistency with credibility; The constraint unification reads the state constraint entries and performs boundary clipping and rate of change clipping on the three types of state fields; The credibility fusion input includes the fragment quality summary field of the slot entry, the missing measurement distribution description field, the tag set field, and the frame skeleton version tag in the reference version reference field, and generates a state vector package containing the groundwater recharge boundary state field, the permeability parameter zoning state field, the pressure relief branch blockage state field, and the state credibility field.

7. The method according to claim 1, characterized in that, The process of performing controlled and constrained assembly includes: Assemble the constraint entries in the stress budget table into a constraint expression structure that includes a static constraint area and a dynamic constraint area. The static constraint area includes constraint object field, constraint boundary field, and constraint priority field, while the dynamic constraint area includes time window index field reference, state reference slot field, rate of change threshold field, gradient threshold field, trigger condition field, and release condition field. Map the state fields in the state vector package to the state reference slots in the dynamic constraint area, establish a one-to-one mapping relationship based on the pressure control unit index field, and establish a one-to-many mapping relationship based on the branch number or valve group identifier field.

8. The method according to claim 1, characterized in that, The process of assembling interlocking rules includes: Arrange interlocking rule entries to generate an interlocking rule diagram containing rule node fields, condition node fields, action node fields, and rollback node fields; Execute access control; when the state credibility field is lower than the gate control threshold, write the gate rejection flag field and trigger the fallback node field. Perform conflict resolution by prioritizing multiple constraint conflicts based on the constraint priority field and rule version number.

9. The method according to claim 1, characterized in that, The process of collaborative control decision processing for valve groups includes: The valve group collaborative control decision processing includes a set of assembly control objects; Read the set of dynamic constraint entries corresponding to the index field of this time window from the control constraint package; Extract the index field of the pressure control unit to form a set of control objects; Read the mapping relationship from the valve group mapping matrix, map the set of controlled objects to the set of valve group identifier fields and the set of branch number fields, and generate the valve group control context package.

10. The method according to claim 1, characterized in that, The closed-loop update package includes: The closed-loop update package includes an anomaly classification result field, a handling action summary field, and a version chain field.

Images