A method and system for full-lifecycle monitoring of engineering projects based on state evolution constraints
By constructing state evolution constraint chains and dynamic constraint envelopes, the problem of difficulty in determining the source of state changes in the full-cycle supervision of transportation infrastructure is solved, and accurate tracking and supervision of state changes are achieved.
Patent Information
- Application Number
- CN202611275846.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-21
- Publication Date
- 2026-11-17
AI Technical Summary
Existing engineering supervision methods are insufficient to effectively track the sources of changes in the condition of transportation infrastructure during the construction, acceptance, operation and maintenance stages. This leads to the easy confusion of construction deviations, cumulative changes during service and rebound changes after treatment, making it difficult to determine the reasonable range of the current condition changes.
By using a state evolution constraint-based method, state fragments of engineering objects, set of stage state anchor points, state baseline transition events, and disposal rewriting events are generated. A state evolution constraint chain is constructed, and the source is decoupled to generate a dynamic constraint envelope, clarifying the source of state changes and the reasonable range of changes.
It enables accurate tracking and monitoring of changes in the status of transportation infrastructure throughout its entire lifecycle, reduces confusion among sources of status changes, and can determine whether current changes exceed reasonable limits and their main sources, thereby improving the accuracy and efficiency of monitoring.
Smart Images

Figure CN123117526A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of data processing technology for the supervision of transportation infrastructure projects, specifically a method and system for full-cycle supervision of projects based on state evolution constraints. Background Technology
[0002] During the design, construction, acceptance, operation, maintenance, and expansion of transportation infrastructure projects such as roads, bridges, tunnels, slopes, and utility tunnels, design parameters, construction deviation records, acceptance and testing records, operation monitoring records, inspection and defect records, maintenance and treatment records, and environmental load records are continuously generated. These data correspond to the formation status, delivery status, service status, and repair status of the project object at different stages.
[0003] The full-cycle state of an engineering object is not a series of isolated snapshots, but a continuous process of change that is inherited, shifted, rewritten, and accumulated as stages transition, service loads, environmental effects, and maintenance are applied. Existing engineering supervision methods typically manage projects, sections, components, inspection items, defect types, or maintenance tasks. They record inspection results, monitoring indicators, inspection defects, and maintenance loops, and can provide alerts for abnormal data, pending issues, and rectification results according to preset thresholds, standard limits, or task milestones. This method can meet the needs of routine data collection, stage acceptance, inspection reporting, and maintenance task management.
[0004] The limitations of existing technologies include at least the following problems: When an engineering object is transferred between the construction, acceptance, operation and maintenance stages, construction deviations may be transformed into delivery benchmarks at acceptance nodes, and maintenance and disposal may partially rewrite the existing change trajectory. Although conventional supervision methods that record by stage ledger, by indicator threshold and by disease category can complete single-stage anomaly alerts and closed-loop archiving, when continuously tracking the same engineering object, it is easy to classify the existing deviations after benchmark migration, cumulative changes in service, rebound changes after disposal and new expansion changes into the same type of state change. This makes it difficult to classify the current state change into the corresponding evolution source, and it is also difficult to determine whether it exceeds the reasonable range of change jointly defined by the previous state, stage benchmark and disposal records. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a method and system for full-cycle monitoring of engineering projects based on state evolution constraints. This method addresses the problem that construction deviations, cumulative changes during service, and rebound changes after treatment are easily confused in cross-stage monitoring of transportation infrastructure projects, making it difficult to attribute the source of current state changes and determine the reasonable range of changes.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a method for full-cycle engineering supervision based on state evolution constraints, comprising the following steps: acquiring full-cycle state data of a target transportation infrastructure project; performing object aggregation, stage merging, and spatiotemporal alignment based on project object identifiers, stage node identifiers, time identifiers, spatial location identifiers, and state indicators to generate project object state fragments; extracting state indicators under each stage node based on the project object state fragments, and generating a set of stage state anchor points according to the stage node order; generating state benchmark migration events based on the benchmark differences between adjacent stage state anchor points in the stage state anchor point set, and processing events based on the pre- and post-processing nodes in the project object state fragments. Changes in state indicators generate disposal rewriting events; the set of stage state anchor points, state baseline migration events, and disposal rewriting events are written into the state link of the same engineering object to generate a state evolution constraint chain; the current state change segment is extracted from the state evolution constraint chain, and residual recursive decoupling is performed on the current state change segment based on the state baseline migration events, disposal rewriting events, and the state evolution constraint chain to generate source decoupling results; the dynamic constraint envelope corresponding to the current state change segment is generated based on the source decoupling results and the state evolution constraint chain, and the dynamic constraint envelope is used as the reasonable change range of the current state change segment; the current state change segment is mapped to the dynamic constraint envelope to generate the full-cycle monitoring results of the project.
[0007] Furthermore, the specific steps for generating the state fragment of the engineering object are as follows: extract the engineering object identifier, stage node identifier, time identifier, spatial location identifier, and state index from the full-cycle state data of the engineering project; generate an object spatial index based on the engineering object identifier and spatial location identifier, and aggregate the full-cycle state data of the engineering project into the object spatial index; divide the full-cycle state data of the engineering project under the object spatial index into the corresponding stage nodes based on the stage node identifier; sort the state indicators under the same object spatial index and the same stage node based on the time identifier to generate the state sequence within the stage; generate the state fragment of the engineering object based on the object spatial index, stage node, and state sequence within the stage.
[0008] Furthermore, the specific steps for generating the stage state anchor point set are as follows: extract the stage node sequence and the state indicators corresponding to each stage node from the state fragment of the engineering object; determine the anchoring window corresponding to each stage node based on the stage node sequence; merge the state indicators within the anchoring window to generate the anchor point state vector; bind the anchor point state vector to the engineering object identifier, spatial location identifier, stage node identifier, and time identifier to generate the stage state anchor point; arrange the stage state anchor points according to the stage node sequence to generate the stage state anchor point set.
[0009] Furthermore, the specific steps for generating the state reference transition event are as follows: extract the previous stage anchor point state vector and the subsequent stage anchor point state vector corresponding to the adjacent stage state anchor points from the stage state anchor point set; calculate the cross-stage reference difference vector based on the previous stage anchor point state vector and the subsequent stage anchor point state vector; generate the reference continuity weight based on the direction matching degree and magnitude continuity between the cross-stage reference difference vector and the subsequent stage anchor point state vector; extract the reference transition component from the cross-stage reference difference vector based on the reference continuity weight to generate the reference transition amount; and generate the state reference transition event based on the reference transition amount.
[0010] Furthermore, the specific steps for generating the disposal rewrite event are as follows: extract the disposal node identifier from the state fragment of the engineering object, and determine the pre-disposal state window and the post-disposal state window based on the disposal node identifier; extract the pre-disposal state index from the pre-disposal state window, and extract the post-disposal state index from the post-disposal state window; calculate the disposal rewrite amount based on the pre-disposal state index and the post-disposal state index; extract the rebound state change fragment based on the state sequence in the stage after the post-disposal state window, and calculate the rebound direction value and the rebound amplitude; generate the disposal rewrite event based on the disposal rewrite amount, the rebound direction value, and the rebound amplitude.
[0011] Furthermore, the specific steps for generating the state evolution constraint chain are as follows: The stage state anchor points in the stage state anchor point set are used as state link nodes; a baseline migration constraint edge is generated based on the state baseline migration event; service accumulation constraint edges are generated based on the time difference and state difference between adjacent state indicators in the state segment of the engineering object; disposal rewriting constraint edges are generated based on the disposal rewriting event; adjacent spatial units corresponding to the state segment of the engineering object are determined based on spatial location identifiers, and spatial expansion constraint edges are generated based on the direction of change of state indicators of adjacent spatial units; the state link nodes, baseline migration constraint edges, service accumulation constraint edges, disposal rewriting constraint edges, and spatial expansion constraint edges are connected according to time identifiers to generate the state evolution constraint chain.
[0012] Further, the specific steps for generating the source decoupling result are as follows: Extract the state change vector corresponding to the current state change segment from the state evolution constraint chain; perform projection matching between the state change vector and the baseline migration amount in the state baseline migration event to generate the stock deviation inheritance value; subtract the stock deviation inheritance value from the state change vector to generate the first residual vector; match the first residual vector with the service accumulation constraint edge to generate the service accumulation value, and subtract the service accumulation value from the first residual vector to generate the second residual vector; match the second residual vector with the disposal rewrite event to generate the disposal rebound value, and subtract the disposal rebound value from the second residual vector to generate the third residual vector; match the third residual vector with the space expansion constraint edge to generate the new expansion value, and generate the source decoupling result based on the stock deviation inheritance value, service accumulation value, disposal rebound value, and new expansion value.
[0013] Furthermore, the specific steps for generating the dynamic constraint envelope are as follows: Based on the source decoupling results, determine the proportions of the inherited stock deviation value, service accumulation value, disposal rebound value, and newly added extended value in the current state change segment; generate the baseline migration constraint boundary based on the inherited stock deviation value and the state baseline migration event; generate the service accumulation constraint boundary based on the service accumulation value and the service accumulation constraint edge; generate the disposal rebound constraint boundary based on the disposal rebound value and the disposal rewrite event; generate the newly added extended constraint boundary based on the newly added extended value and the spatial extension constraint edge; and generate the dynamic constraint envelope by weighting and combining the baseline migration constraint boundary, service accumulation constraint boundary, disposal rebound constraint boundary, and newly added extended constraint boundary based on each proportion.
[0014] Furthermore, the specific steps for generating the full-cycle monitoring results of the project are as follows: mapping the current state change segment to a dynamic constraint envelope and calculating the envelope deviation; determining the dominant change source corresponding to the current state change segment based on the envelope deviation and the source decoupling result; generating a state tracking record based on the current state change segment, the dominant change source, and the state evolution constraint chain; marking the state tracking record with a monitoring level based on the envelope deviation; and generating the full-cycle monitoring results of the project based on the state tracking record and the monitoring level mark.
[0015] A full-cycle monitoring system for engineering projects based on state evolution constraints includes: a state fragment generation module, used to acquire full-cycle state data of the target transportation infrastructure project, and perform object aggregation, stage merging, and spatiotemporal alignment to generate state fragments of the project objects; a state anchor point generation module, used to extract state indicators under each stage node based on the state fragments of the project objects, and generate a set of stage state anchor points according to the stage node order; an evolution event generation module, used to generate state baseline transition events based on the set of stage state anchor points, and generate disposal rewriting events based on the state fragments of the project objects; and a constraint chain generation module, used to connect the set of stage state anchor points... The state baseline migration event and the disposal rewriting event are written into the state link of the same engineering object to generate a state evolution constraint chain; the source decoupling module is used to extract the current state change segment from the state evolution constraint chain, and perform residual recursive decoupling on the current state change segment based on the state baseline migration event, the disposal rewriting event and the state evolution constraint chain to generate the source decoupling result; the constraint envelope generation module is used to generate the dynamic constraint envelope corresponding to the current state change segment based on the source decoupling result and the state evolution constraint chain; the supervision result generation module is used to map the current state change segment to the dynamic constraint envelope to generate the full-cycle supervision result of the project.
[0016] The present invention has the following beneficial effects: (1) The engineering full-cycle supervision method based on state evolution constraints collects, merges, and aligns the engineering full-cycle state data through engineering object identification, stage node identification, time identification, spatial location identification and state indicators. It organizes the scattered detection, monitoring, inspection and disposal data formed in the design, construction, acceptance, operation, maintenance and expansion process into engineering object state fragments, and further generates a set of stage state anchor points. This enables the state indicators of the same engineering object at different stages to participate in subsequent processing in the same object, same location and same stage order, reducing the problem of state record fragmentation caused by different data sources, different component numbers and inconsistent mileage intervals, so that cross-stage supervision is no longer limited to the level of ledger summary.
[0017] (2) The engineering full-cycle supervision method based on state evolution constraints generates state benchmark migration events through the benchmark differences between state anchor points of adjacent stages, and generates disposal rewriting events through the changes in state indicators before and after disposal nodes. The process of construction forming deviation being accepted by the acceptance stage to form delivery benchmarks and the process of maintenance rewriting existing state trajectories are extracted separately and written into the state evolution constraint chain of the same engineering object. This makes state changes no longer simply archived according to stage, indicator or disease name, but can put benchmark acceptance, service accumulation, disposal rewriting and spatial expansion into the same state link for tracking, reducing the problem of mutual confusion between existing deviations, new changes and post-disposal rebound.
[0018] (3) The engineering full-cycle supervision method based on state evolution constraints extracts the current state change segment from the state evolution constraint chain and matches the state change vector with the baseline migration amount, service accumulation constraint edge, disposal rewriting event and spatial expansion constraint edge in sequence to generate the stock deviation inheritance value, service accumulation value, disposal rebound value and new expansion value. This enables the current state change to be decomposed according to the source, and then generates a dynamic constraint envelope according to the proportion of each source component. The dynamic constraint envelope is used as the reasonable change range of the current state change segment, thereby avoiding the treatment of changes from different sources under the same scale when simply using a fixed threshold judgment. This allows the supervision result to reflect whether the current change exceeds the reasonable range and which evolution source the excess range mainly comes from.
[0019] (4) The engineering full-cycle monitoring system based on state evolution constraints, through the sequential cooperation of the state fragment generation module, state anchor point generation module, evolution event generation module, constraint chain generation module, source decoupling module, constraint envelope generation module and monitoring result generation module, enables the engineering full-cycle state data to form a complete processing link from data collection, stage anchoring, event extraction, constraint chain construction, source decoupling, envelope generation to monitoring result output. The modules maintain the data correspondence through the same engineering object identifier and spatial location identifier, reducing the workload of manual repeated verification between data at different stages, different inspection records and different maintenance and treatment records, and also enabling the engineering full-cycle monitoring results to directly correspond to the target object, spatial location, dominant change source and monitoring level mark.
[0020] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description
[0021] Figure 1 This is a timing diagram of a full-cycle engineering monitoring method based on state evolution constraints according to the present invention.
[0022] Figure 2 This is an architecture diagram of an engineering full-cycle monitoring system based on state evolution constraints according to the present invention. Detailed Implementation
[0023] Please see Figure 1 This invention provides a technical solution: a method for full-cycle monitoring of engineering projects based on state evolution constraints, comprising the following steps: acquiring full-cycle state data of a target transportation infrastructure project; performing object aggregation, stage merging, and spatiotemporal alignment based on project object identifiers, stage node identifiers, time identifiers, spatial location identifiers, and state indicators to generate project object state fragments; extracting state indicators under each stage node based on the project object state fragments, and generating a set of stage state anchor points according to the stage node order; generating state benchmark migration events based on the benchmark differences between adjacent stage state anchor points in the stage state anchor point set, and based on the state indicators before and after the disposal node in the project object state fragments. The changes between them generate disposal rewriting events; the set of stage state anchor points, state baseline migration events, and disposal rewriting events are written into the state link of the same engineering object to generate a state evolution constraint chain; the current state change segment is extracted from the state evolution constraint chain, and residual recursive decoupling is performed on the current state change segment based on the state baseline migration events, disposal rewriting events, and state evolution constraint chain to generate source decoupling results; the dynamic constraint envelope corresponding to the current state change segment is generated based on the source decoupling results and the state evolution constraint chain, and the dynamic constraint envelope is used as the reasonable change range of the current state change segment; the current state change segment is mapped to the dynamic constraint envelope to generate the full-cycle monitoring results of the project.
[0024] Among them, the full-cycle status data of the project is a set of status data generated during the design, construction, acceptance, operation, maintenance and reconstruction and expansion of the target transportation infrastructure project; The project object identifier is used to point to the target object in the monitored road section, bridge component, tunnel section, slope unit, and utility tunnel section; Stage node identifiers are used to indicate the stage that the target object is currently in; The time stamp is used to indicate the moment when data collection, detection, inspection, processing, and retesting occur; Spatial location markers are used to indicate mileage, coordinates, component locations, and section boundaries; Status indicators are used to represent the structural response, appearance defects, geometric offset, material degradation, and post-treatment changes of the target object at the corresponding stage node.
[0025] Specifically, the steps for generating a state fragment of a project object are as follows: The project object identifier, stage node identifier, time identifier, spatial location identifier, and status index are extracted from the project's full-cycle status data, specifically as follows: Read the record header and record content from the full life cycle status data of the project, extract the project object identifier, stage node identifier, time identifier and spatial location identifier from the record header, extract the status index from the record content, and encapsulate the project object identifier, stage node identifier, time identifier, spatial location identifier and status index into status element record.
[0026] An object spatial index is generated based on the project object identifier and spatial location identifier. The entire life cycle status data of the project is then collected into the object spatial index. Specifically: Using the project object identifier as the primary key and the spatial location identifier as the location key, an object spatial index is generated by concatenating them. Status element records with the same project object identifier and spatial location identifier are grouped under the same object spatial index. Object space indexes are generated according to the following formula: ; in, Represents the object space index. Represents a hash mapping function. Indicates the identification of the project object. Indicates spatial location; In one implementation, the object space index is shown in Table 1: Table 1. Example of object space index Based on the stage node identifier, the project's full-cycle status data under the object space index is divided into corresponding stage nodes, specifically as follows: Read the stage node identifier of each status element record under the object space index, write status element records with the same stage node identifier into the same stage node data group, and arrange the stage node data group in the order of design, construction, acceptance, operation, maintenance and renovation / expansion.
[0027] Based on time identifiers, state indicators under the same object spatial index and the same stage node are sorted to generate a state sequence within the stage, specifically as follows: Read the status indicators under the same object spatial index and the same stage node, sort the status indicators in ascending order according to the time identifier, and combine the sorted status indicators, time identifier and spatial location identifier into a state sequence within the stage.
[0028] The process generates project object state fragments based on object space indexes, stage nodes, and intra-stage state sequences, specifically as follows: The project object state fragment is generated by using the object space index as the fragment belonging field, the stage node as the fragment stage field, and the state sequence within the stage as the fragment content field.
[0029] In this implementation plan, the engineering object state fragment merges the cross-stage state indicators of the same engineering object at the same spatial location into the same data fragment, so that the stage state anchor point, state reference migration event and disposal rewriting event all revolve around the same engineering object, reducing the break in the state chain caused by the mismatch between component number, mileage range and detection location.
[0030] Specifically, the steps for generating the set of stage state anchor points are as follows: Extracting the stage node sequence and the corresponding state indicators for each stage node from the state fragment of the project object, specifically: Read the stage node data group from the status fragment of the project object, extract the stage node identifier and the status index within the stage node data group, and obtain the stage node order according to the project sequence relationship corresponding to the stage node identifier.
[0031] The anchoring window for each stage node is determined based on the stage node sequence, specifically as follows: Using the start and end times of the stage nodes as boundaries, state indicators close to the stage transition time are selected as anchor candidate data, and the time range of the anchor candidate data is used as the anchor window. In one implementation, the anchoring window for the acceptance phase is the status indicators within three days before and after the completion of the acceptance test, and the anchoring window for the maintenance phase is the status indicators between the first retest and the second retest after the treatment is completed.
[0032] Within the anchoring window, state indices are merged to generate anchor point state vectors, specifically as follows: Perform dimensional conversion on the status indicators within the anchor window so that the same status indicator participates in the merging under the same numerical caliber. The time weights are set according to the distance of the state indicators from the center of the anchor window. The state indicators are then weighted and merged according to the time weights to generate the anchor point state vector. The anchor point state vector is generated according to the following formula: ; in, Indicates the first Anchor point state vectors under each stage node Indicates the first Within the anchoring window of the first stage node A vector of state indicators, Indicates the first Within the anchoring window of the first stage node The time weights corresponding to each state indicator vector Indicates the first The number of state indicator vectors within the anchor window of each stage node; Time weight According to the The time distance between the acquisition time of each state indicator vector and the center time of the anchor window is determined. The smaller the time distance, the greater the time weight.
[0033] The anchor point state vector is bound to the project object identifier, spatial location identifier, stage node identifier, and time identifier to generate the stage state anchor point, specifically as follows: The anchor point state vector is bound to the engineering object identifier, spatial location identifier, stage node identifier, and anchor window center time identifier to generate stage state anchor points with object affiliation, spatial affiliation, stage affiliation, and time affiliation.
[0034] Arrange the stage state anchors in the order of stage nodes to generate a set of stage state anchors.
[0035] In this implementation plan, the stage state anchor point consolidates multiple discrete state indicators within a stage into an anchor point state vector, enabling the benchmark difference calculation between the construction stage, acceptance stage, operation stage and maintenance stage according to the same vector caliber, avoiding the confusion of cross-stage comparison caliber caused by directly using scattered test values.
[0036] Specifically, the steps for generating state baseline transition events are as follows: Extracting the previous stage anchor state vector and the next stage anchor state vector corresponding to adjacent stage state anchor points from the set of stage state anchor points, specifically: Traverse the set of stage state anchors in the order of stage nodes, select adjacent stage state anchors, take the stage state anchor that is earlier in time as the previous stage state anchor, and take the stage state anchor that is later in time as the next stage state anchor. Read the previous stage anchor state vector corresponding to the previous stage state anchor, and read the next stage anchor state vector corresponding to the next stage state anchor.
[0037] The cross-stage benchmark difference vector is calculated based on the anchor point state vector of the previous stage and the anchor point state vector of the subsequent stage, specifically as follows: Subtracting the anchor state vector of the previous stage from the anchor state vector of the later stage yields the cross-stage benchmark difference vector. The cross-stage benchmark difference vector is calculated using the following formula: ; in, Indicates the first The stage node to the first Cross-stage baseline difference vector between stage nodes Indicates the first The state vector of the anchor point of the next stage under each stage node. Indicates the first The state vector of the previous stage anchor point under each stage node.
[0038] Based on the direction matching degree and magnitude continuity between the cross-stage benchmark difference vector and the subsequent stage anchor point state vector, a benchmark continuity weight is generated, specifically as follows: Calculate the cosine of the direction angle between the cross-stage baseline difference vector and the subsequent stage anchor point state vector, and convert the cosine of the direction angle into the direction matching degree. Calculate the degree of similarity between the cross-stage benchmark difference vector magnitude and the subsequent stage anchor point state vector magnitude to obtain the magnitude continuity. Input the direction matching degree and the magnitude acceptance degree into the acceptance weight function to generate the benchmark acceptance weight; The benchmark acceptance weight is generated according to the following formula: ; ; ; in, Indicates the first The stage node to the first The baseline weighting between nodes in each stage. Indicates the first The stage node to the first The degree of directional matching between nodes in each stage. Indicates the first The stage node to the first The amplitude continuity between each stage node This represents the weighting coefficient for directional matching. This represents the amplitude continuity weighting coefficient, and , This represents a constant to prevent the denominator from being zero. Direction matching degree Used to characterize the directional similarity between the cross-stage benchmark difference vector and the subsequent stage anchor point state vector; Amplitude tolerance It is used to characterize the degree of similarity between the cross-stage benchmark difference vector magnitude and the subsequent stage anchor point state vector magnitude.
[0039] Based on the benchmark transfer weight, the benchmark migration component is extracted from the cross-stage benchmark difference vector to generate the benchmark migration amount, which is as follows: Multiply the cross-stage benchmark difference vector with the benchmark acceptance weight to extract the benchmark migration component accepted by the state anchor point of the later stage, and use the benchmark migration component as the benchmark migration amount. The baseline migration is generated according to the following formula: ; in, Indicates the first The stage node to the first The baseline migration amount between each stage node.
[0040] A state baseline migration event is generated based on the baseline migration amount, specifically as follows: The baseline migration amount, the previous stage state anchor point, the next stage state anchor point, the engineering object identifier, the spatial location identifier, and the time identifier are combined to generate a state baseline migration event.
[0041] In this implementation plan, the direction matching degree is used to determine whether the cross-stage benchmark difference is accepted along the direction of the subsequent stage state, and the amplitude acceptance degree is used to determine whether the cross-stage benchmark difference is accepted by the subsequent stage state in terms of amplitude. By extracting the benchmark migration amount through the benchmark acceptance weight, the delivery benchmark formed after acceptance can be separated from ordinary fluctuations, reducing the situation of repeatedly judging the accepted existing deviations as new changes.
[0042] Specifically, the steps for handling the rewrite event are as follows: Extract the disposal node identifier from the state fragment of the engineering object, and determine the pre-disposal state window and post-disposal state window based on the disposal node identifier. Specifically: Read the disposal node identifier in the status fragment of the project object, take the status indicator before the time corresponding to the disposal node identifier and the same spatial location as the disposal as the pre-disposal status window, and take the re-measured status indicator after the time corresponding to the disposal node identifier and the same spatial location as the disposal as the post-disposal status window.
[0043] Extract pre-disposal status indicators from the pre-disposal status window and post-disposal status indicators from the post-disposal status window, specifically as follows: Extract status indicators with the same name, spatial location, and detection caliber from the pre-treatment status window and post-treatment status window, and pair them according to the time identifier.
[0044] The amount of rewriting data is calculated based on the pre-disposal and post-disposal status indicators, specifically as follows: Subtracting the pre-treatment state index from the post-treatment state index yields the change in the state trajectory caused by the treatment node, and this change is used as the treatment rewrite amount.
[0045] Based on the state sequence within the stage following the post-treatment state window, the rebound state change segment is extracted, and the rebound direction value and rebound amplitude are calculated, specifically as follows: In the state sequence of the stage following the post-treatment state window, extract the state index that is continuously connected to the post-treatment state index, and subtract the post-treatment state index from the state index to obtain the rebound state change segment. The unit direction vector of the rebound state change segment is used as the rebound direction value, and the vector norm of the rebound state change segment is used as the rebound amplitude value.
[0046] The disposal rewrite event is generated based on the disposal rewrite amount, the rebound direction value, and the rebound amplitude. Specifically, the disposal rewrite amount, the rebound direction value, the rebound amplitude, the disposal node identifier, the engineering object identifier, the spatial location identifier, and the time identifier are encapsulated into a disposal rewrite event.
[0047] In this implementation plan, the immediate rewriting caused by the treatment and the rebound changes after the treatment are distinguished by the pre-treatment status window, the post-treatment status window and the rebound status change segment. This allows the status changes after the treatment node to be decoupled from the source separately, reducing the situation where the treatment rebound is mixed into the service cumulative changes.
[0048] Specifically, the steps for generating the state evolution constraint chain are as follows: The stage state anchor points in the stage state anchor point set are taken as state link nodes, specifically as follows: Read the set of stage state anchor points in the order of stage nodes, convert each stage state anchor point into a state link node, and use the anchor point state vector as the node value of the state link node.
[0049] The baseline transition constraint edges are generated based on the state baseline transition events, specifically as follows: Write the state baseline migration event between the corresponding adjacent state link nodes, and use the baseline migration amount as the edge value to generate baseline migration constraint edges.
[0050] Service accumulation constraint edges are generated based on the time difference and state difference between adjacent state indices in the state segment of the engineering object. Specifically: Within the same engineering object state segment, calculate the time difference and state difference between adjacent state indicators, use the ratio of state difference to time difference as the service evolution rate, write the service evolution rate between the corresponding state link nodes, and generate service cumulative constraint edges.
[0051] Based on the disposal rewriting event, disposal rewriting constraint edges are generated, specifically as follows: Write the rewrite event between the state link nodes before and after the rewrite node, and use the rewrite amount, bounce direction value, and bounce amplitude value as edge values to generate rewrite constraint edges.
[0052] Based on spatial location identifiers, adjacent spatial units corresponding to the state segments of engineering objects are determined, and spatial expansion constraint edges are generated based on the direction of change of state indicators of adjacent spatial units. Specifically: Based on the spatial location identifier, identify the adjacent spatial units that are continuously connected to the state segment of the engineering object in terms of mileage, coordinates, component connection relationship and structural force transmission relationship. Extract the change direction of the state index of the adjacent spatial units and write the change direction of the state index of the adjacent spatial units into the spatial expansion constraint edge. In one implementation, the decoupling relationship between the constraint edge and its source is shown in Table 2: Table 2 Examples of Decoupling Relationships Between Constraint Edges and Sources A state evolution constraint chain is generated by connecting state link nodes, baseline migration constraint edges, service accumulation constraint edges, disposal rewriting constraint edges, and space expansion constraint edges according to time identifiers. Specifically: The state link nodes are connected sequentially according to the time identifier, and the baseline migration constraint edge, service accumulation constraint edge, disposal rewriting constraint edge and space expansion constraint edge are written between the corresponding nodes to generate a state evolution constraint chain.
[0053] In this implementation scheme, the state evolution constraint chain incorporates baseline migration, service accumulation, disposal rewriting, and spatial expansion into the state link of the same engineering object, enabling the current state change segment to trace its source along both the time and space directions, thereby reducing source confusion caused by single-point threshold judgment.
[0054] Specifically, the steps for generating the source decoupling result are as follows: Extracting the state change vector corresponding to the current state change segment from the state evolution constraint chain is as follows: Read the current state index and the node value of the previous state link node corresponding to the current state change segment, subtract the node value of the previous state link node from the current state index, and generate a state change vector.
[0055] The state change vector is projected and matched with the baseline migration amount in the state baseline migration event to generate the stock deviation inheritance value, which is as follows: Project the state change vector onto the direction of the reference migration quantity to obtain a component in the same direction as the reference migration quantity, and use this component as the inherited value of the stock deviation. The inherited value of the stock deviation is calculated according to the following formula: ; in, Indicates the current time The corresponding inherited value of the stock deviation, Indicates the current time The corresponding state change vector, Indicates the first The stage node to the first The baseline migration amount between each stage node This represents a constant used to prevent the denominator from being zero.
[0056] The first residual vector is generated by subtracting the inherited value of the stock deviation from the state change vector. The first residual vector is matched with the service accumulation constraint edge to generate the service accumulation value. The service accumulation value is then subtracted from the first residual vector to generate the second residual vector, which is as follows: Read the service evolution rate in the service accumulation constraint edge, calculate the service accumulation candidate value according to the time interval corresponding to the current state change segment, project the first residual vector onto the service accumulation candidate value direction to obtain the service accumulation value, and subtract the service accumulation value from the first residual vector to generate the second residual vector.
[0057] The second residual vector is matched with the disposal rewrite event to generate the disposal bounce value. The disposal bounce value is then subtracted from the second residual vector to generate the third residual vector, which is as follows: Read the rebound direction value and rebound amplitude value in the rewrite event, project the second residual vector onto the direction corresponding to the rebound direction value, and use the rebound amplitude value to limit the projection length to obtain the rewrite rebound value. Subtract the rewrite rebound value from the second residual vector to generate the third residual vector.
[0058] The third residual vector is matched with the spatial expansion constraint edge to generate a new expansion value. Based on the inherited stock deviation value, service accumulation value, disposal rebound value, and the new expansion value, a source decoupling result is generated, specifically as follows: Read the direction of change of state index of adjacent spatial unit in the spatial extension constraint edge, project the third residual vector onto the direction of change of state index of adjacent spatial unit to obtain the new extension value, and encapsulate the existing deviation inheritance value, service accumulation value, disposal rebound value and the new extension value as the source decoupling result.
[0059] In this implementation plan, the residual recursive decoupling is used to deduct the inherited value of the stock deviation, the accumulated value of service and the rebound value of disposal in sequence, so that the remaining changes enter the new extension judgment, reduce the mutual obscuring between different evolution sources, and make the dominant change source of the current state change segment clearer.
[0060] Specifically, the steps for generating the dynamic constraint envelope are as follows: Based on the source decoupling results, the proportions of inherited deviation values, service accumulation values, disposal rebound values, and newly added extended values in the current state change segment are determined as follows: Calculate the magnitudes of the inherited value of the stock deviation, the cumulative value of service, the rebound value of disposal, and the newly added extended value, and divide each magnitude by the sum of all magnitudes to obtain the corresponding percentage.
[0061] The baseline migration constraint boundary is generated based on the inherited value of the existing deviation and the state baseline migration event, specifically as follows: The baseline migration constraint boundary is generated by taking the baseline migration amount in the state baseline migration event as the center and the amplitude fluctuation of the stock deviation inheritance value as the boundary width. The amplitude fluctuation is taken as the root mean square value of the residual of the same state index within the preceding state window corresponding to the state baseline migration event.
[0062] The service accumulation constraint boundary is generated based on the service accumulation value and the service accumulation constraint edge, specifically as follows: The service accumulation constraint boundary is generated by taking the service evolution rate in the service accumulation constraint edge as the boundary slope and the amplitude fluctuation of the service accumulation value as the boundary width; the amplitude fluctuation is taken as the root mean square value of the residual of the same state index in the corresponding time period of the service accumulation constraint edge.
[0063] The disposal bounce constraint boundary is generated based on the disposal bounce value and the disposal rewrite event, specifically as follows: The handling bounce constraint boundary is generated by taking the bounce direction value in the handling rewrite event as the boundary direction and the difference between the bounce amplitude and the handling bounce value as the boundary width.
[0064] The newly extended constraint boundary is generated based on the newly added extended value and the spatial extended constraint edge, specifically as follows: The new extended constraint boundary is generated by taking the direction of change of the state index of the adjacent spatial unit in the spatial extension constraint edge as the boundary direction and the magnitude of the newly added extended value and the magnitude of the change of the state index of the adjacent spatial unit as the boundary width.
[0065] Based on each proportion, a weighted combination of the baseline migration constraint boundary, service accumulation constraint boundary, springback disposal constraint boundary, and newly added extended constraint boundary is generated to produce a dynamic constraint envelope, which is as follows: Based on the proportions of inherited values of existing deviations, accumulated values of service, rebound values of disposal, and newly added extended values, the corresponding constraint boundaries are weighted and combined to obtain the dynamic constraint envelope corresponding to the current state change segment; The dynamic constraint envelope is generated according to the following formula: ; ; in, Indicates the current time Next The proportion of each source component, Indicates the current time Next One source component, Indicates the index of the summation of the source components. Indicates the current time The corresponding dynamic constraint envelope, Indicates the first Each source component corresponds to the lower boundary of the constraint boundary. Indicates the first Each source component corresponds to the upper boundary of the constraint boundary. This represents a constant to prevent the denominator from being zero. in, The corresponding inventory deviation inheritance value. The corresponding cumulative service value at that time The rebound value should be handled accordingly. The corresponding new extended value will be added at that time.
[0066] In this implementation plan, a dynamic constraint envelope is generated by the proportion of source components, so that the reasonable range of variation can change with the proportion of existing deviation, service accumulation, disposal rebound and new expansion, reducing the situation where a fixed threshold is used to apply the same judgment scale to different evolution sources.
[0067] Specifically, the steps for generating the full-cycle monitoring results of the project are as follows: Map the current state change segment to the dynamic constraint envelope and calculate the envelope deviation, specifically as follows: The state change vector corresponding to the current state change segment is compared with the upper and lower boundaries of the dynamic constraint envelope. When the state change vector is within the dynamic constraint envelope, the envelope deviation is zero. When the state change vector exceeds the dynamic constraint envelope, the distance of the state change vector beyond the boundary is calculated as the envelope deviation.
[0068] Based on the envelope deviation and source decoupling results, the dominant change source corresponding to the current state change segment is determined, specifically as follows: By comparing the proportions of inherited values of existing deviations, accumulated values of service, rebound values of disposal, and newly added extended values in the source decoupling results, the source component with the highest proportion is taken as the dominant change source corresponding to the current state change segment.
[0069] A state tracking record is generated based on the current state change segment, the dominant source of change, and the state evolution constraint chain, specifically as follows: Write the current state change segment, the dominant change source, the envelope deviation, the corresponding state link node, and the corresponding constraint edge into the same state tracking record.
[0070] The status tracking records are labeled with regulatory levels based on envelope deviation, specifically as follows: The envelope deviation is compared with a preset level threshold, and the status tracking record is marked with a regulatory level based on the comparison result.
[0071] The project's full-cycle supervision results are generated based on status tracking records and regulatory level markings, specifically as follows: By combining status tracking records, regulatory level markers, engineering object identifiers, spatial location identifiers, and time identifiers corresponding to current status change segments, a full-cycle engineering supervision result is generated.
[0072] In this implementation plan, the regulatory results are generated by combining the envelope deviation and the dominant source of change. The regulatory results indicate whether the current state change exceeds the reasonable range of change, and provide the evolutionary source corresponding to the excess range. This facilitates the differentiation of regulatory differences caused by baseline migration, service accumulation, disposal rebound, and new expansion.
[0073] Please see Figure 2This invention provides a technical solution: a full-cycle engineering monitoring system based on state evolution constraints, comprising: a state fragment generation module, used to acquire full-cycle state data of a target transportation infrastructure project, and perform object aggregation, stage merging, and spatiotemporal alignment to generate state fragments of the project object; a state anchor point generation module, used to extract state indicators under each stage node based on the state fragments of the project object, and generate a set of stage state anchor points according to the stage node order; an evolution event generation module, used to generate state baseline migration events based on the set of stage state anchor points, and generate disposal rewriting events based on the state fragments of the project object; and a constraint chain generation module, used to... The process involves several modules: a stage state anchor set, state baseline migration events, and disposal rewriting events are written into the state chain of the same engineering object to generate a state evolution constraint chain; a source decoupling module extracts the current state change fragment from the state evolution constraint chain and performs residual recursive decoupling on the current state change fragment based on the state baseline migration events, disposal rewriting events, and the state evolution constraint chain to generate a source decoupling result; a constraint envelope generation module generates a dynamic constraint envelope corresponding to the current state change fragment based on the source decoupling result and the state evolution constraint chain; and a monitoring result generation module maps the current state change fragment to the dynamic constraint envelope to generate a full-cycle monitoring result for the engineering project.
[0074] Among them, the state fragment generation module, state anchor point generation module, evolution event generation module, constraint chain generation module, source decoupling module, constraint envelope generation module, and supervision result generation module are connected in the order of transmission of engineering object state fragments, stage state anchor point sets, state benchmark migration events, disposal rewriting events, state evolution constraint chains, source decoupling results, dynamic constraint envelopes, and engineering full-cycle supervision results. Each module maintains the data correspondence through the same engineering object identifier and spatial location identifier.
[0075] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.
[0076] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.
Claims
1. A method for full-cycle monitoring of engineering projects based on state evolution constraints, characterized in that, Includes the following steps: Acquire the full-cycle status data of the target transportation infrastructure project, and perform object aggregation, stage merging and spatiotemporal alignment based on project object identifiers, stage node identifiers, time identifiers, spatial location identifiers and status indicators to generate project object status fragments; Based on the state fragments of engineering objects, extract the state indicators under each stage node, and generate a set of stage state anchor points according to the stage node order. A state baseline migration event is generated based on the baseline difference between adjacent stage state anchor points in the stage state anchor point set, and a disposal rewrite event is generated based on the change in state indicators before and after the disposal node in the state segment of the engineering object. Write the set of stage state anchor points, state baseline transition events, and disposal rewriting events into the state link of the same engineering object to generate a state evolution constraint chain; Extract the current state change segment from the state evolution constraint chain, and perform residual recursive decoupling on the current state change segment based on the state baseline transition event, disposal rewriting event and state evolution constraint chain to generate source decoupling result; Based on the source decoupling results and the state evolution constraint chain, a dynamic constraint envelope corresponding to the current state change segment is generated, and the dynamic constraint envelope is used as the reasonable change range of the current state change segment. Map the current state change segments to the dynamic constraint envelope to generate full-cycle monitoring results for the project.
2. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating a project object state fragment are as follows: Extract project object identifiers, stage node identifiers, time identifiers, spatial location identifiers, and status indicators from the project's full lifecycle status data; An object spatial index is generated based on the engineering object identifier and spatial location identifier, and the engineering life cycle status data is collected into the object spatial index. Based on the stage node identifier, the project's full-cycle status data under the object space index is divided into the corresponding stage nodes; Based on time identifiers, the state indicators under the same object space index and the same stage node are sorted to generate a state sequence within the stage; Project object state fragments are generated based on object space indexes, stage nodes, and state sequences within stages.
3. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating the stage state anchor point set are as follows: Extract the sequence of stage nodes and the corresponding state indicators of each stage node from the state fragment of the engineering object; The anchoring window corresponding to each stage node is determined based on the stage node sequence. Within the anchoring window, state indices are merged to generate anchor point state vectors; Bind the anchor point state vector to the project object identifier, spatial location identifier, stage node identifier, and time identifier to generate stage state anchor points; Arrange the stage state anchors in the order of stage nodes to generate a set of stage state anchors.
4. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating state baseline transition events are as follows: Extract the previous stage anchor state vector and the next stage anchor state vector corresponding to the adjacent stage state anchor points from the set of stage state anchor points; Calculate the cross-stage benchmark difference vector based on the anchor point state vector of the previous stage and the anchor point state vector of the next stage; Based on the directional matching degree and magnitude continuity between the cross-stage benchmark difference vector and the subsequent stage anchor point state vector, a benchmark continuity weight is generated. Based on the benchmark transfer weight, the benchmark migration component is extracted from the cross-stage benchmark difference vector to generate the benchmark migration amount; State baseline migration events are generated based on baseline migration amounts.
5. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for handling rewrite events are as follows: Extract disposal node identifiers from the state fragments of engineering objects, and determine the pre-disposal state window and post-disposal state window based on the disposal node identifiers; Extract the pre-disposal status indicators from the pre-disposal status window, and extract the post-disposal status indicators from the post-disposal status window. The amount of rewritten data is calculated based on the pre- and post-treatment status indicators. Extract the rebound state change segment based on the state sequence within the stage after the post-treatment state window, and calculate the rebound direction value and rebound amplitude. The disposal rewrite event is generated based on the disposal rewrite amount, the bounce direction value, and the bounce amplitude.
6. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating state evolution constraint chains are as follows: Use the stage state anchor points in the stage state anchor point set as state link nodes; Generate baseline migration constraint edges based on state baseline migration events; Service cumulative constraint edges are generated based on the time difference and state difference between adjacent state indices in the state segment of the engineering object. Generate disposal rewriting constraint edges based on disposal rewriting events; Based on spatial location identifiers, adjacent spatial units corresponding to the state segments of engineering objects are determined, and spatial expansion constraint edges are generated based on the change direction of the state indicators of adjacent spatial units. A state evolution constraint chain is generated by connecting state link nodes, baseline migration constraint edges, service accumulation constraint edges, disposal rewriting constraint edges, and space expansion constraint edges according to time identifiers.
7. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating the source decoupling results are as follows: Extract the state change vector corresponding to the current state change segment from the state evolution constraint chain; Project the state change vector with the baseline migration amount in the state baseline migration event to generate the stock deviation inheritance value. The first residual vector is generated by subtracting the inherited value of the stock deviation from the state change vector. The first residual vector is matched with the service accumulation constraint edge to generate the service accumulation value, and the service accumulation value is subtracted from the first residual vector to generate the second residual vector; The second residual vector is matched with the disposal rewrite event to generate the disposal bounce value, and the disposal bounce value is subtracted from the second residual vector to generate the third residual vector; The third residual vector is matched with the spatial expansion constraint edge to generate new expansion values, and the source decoupling result is generated based on the existing deviation inheritance value, service accumulation value, disposal rebound value and new expansion value.
8. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating dynamic constraint envelopes are as follows: Based on the source decoupling results, determine the proportions of the inherited value of existing deviation, the accumulated value of service, the rebound value of disposal, and the newly added extended value in the current state change segment; Generate baseline migration constraint boundaries based on inherited values of existing deviations and state baseline migration events; Generate service accumulation constraint boundaries based on service accumulation values and service accumulation constraint edges; Generate disposal bounce constraint boundaries based on disposal bounce values and disposal rewrite events; Generate new extended constraint boundaries based on the newly added extended values and spatial extended constraint edges; Based on each proportion, a weighted combination of the baseline migration constraint boundary, service accumulation constraint boundary, disposal rebound constraint boundary, and newly added extended constraint boundary is used to generate a dynamic constraint envelope.
9. The engineering lifecycle monitoring method based on state evolution constraints according to claim 1, characterized in that, The specific steps for generating the full-cycle monitoring results of the project are as follows: Map the current state change segment to the dynamic constraint envelope and calculate the envelope deviation. The dominant change source corresponding to the current state change segment is determined based on the envelope deviation and source decoupling results. A state tracking record is generated based on the current state change fragment, the dominant source of change, and the state evolution constraint chain; The status tracking records are marked with a regulatory level based on the envelope deviation. The project's full-cycle regulatory results are generated based on status tracking records and regulatory level markings.
10. A lifecycle monitoring system for engineering projects based on state evolution constraints, employing the lifecycle monitoring method for engineering projects based on state evolution constraints as described in any one of claims 1-9, characterized in that, include: The state fragment generation module is used to acquire the full life cycle state data of the target transportation infrastructure project, and perform object aggregation, stage merging and spatiotemporal alignment to generate state fragments of the project objects. The state anchor point generation module is used to extract state indicators under each stage node based on the state fragment of the engineering object, and generate a set of stage state anchor points according to the stage node order. The evolution event generation module is used to generate state baseline transition events based on the set of stage state anchor points, and to generate disposal rewriting events based on the state fragments of engineering objects. The constraint chain generation module is used to write the set of stage state anchor points, state baseline transition events, and disposal rewriting events into the state link of the same engineering object to generate a state evolution constraint chain. The source decoupling module is used to extract the current state change segment from the state evolution constraint chain, and perform residual recursive decoupling on the current state change segment based on the state baseline transition event, disposal rewriting event and state evolution constraint chain to generate the source decoupling result; The constraint envelope generation module is used to generate the dynamic constraint envelope corresponding to the current state change segment based on the source decoupling result and the state evolution constraint chain. The monitoring result generation module is used to map the current state change segments to the dynamic constraint envelope and generate full-cycle monitoring results for the project.
