A power production-based full-chain digitalization construction collaborative management method and system

Abstract

The present application relates to the technical field of electric power construction management, in particular to a full-chain intelligent construction collaborative management method and system based on electric power production; the method comprises the following steps: constructing a construction actual measurement parameter reverse mapping path, reversely synchronizing the construction actual measurement data set to a design platform; structuring the change information in the debugging event log into measurement point mapping change data, and constructing a debugging change information reverse synchronization path based on a point structure loop three-way binding structure; constructing a dynamic transformer area operation image, introducing a transformer area deviation attribution recognition model to compare the dynamic transformer area operation image with static archive data to generate a transformer area operation deviation matrix, and dynamically adjusting the reactive power compensation control strategy and the load prediction model parameters; and uniformly archiving the construction actual measurement data set, the measurement point mapping change data and the dynamic transformer area operation image in real time. The present application realizes a reverse closed-loop collaborative management method for the whole life cycle of electric power production and construction.

Description

Technical Field

[0001] This invention relates to the field of power construction management technology, specifically to a method and system for collaborative management of power production based on the full-chain digital intelligence of construction. Background Technology

[0002] As the power industry accelerates its development towards "digitalization, automation, and intelligence," distribution substations, as a crucial component of the power grid's end, have a significant impact on power supply quality, energy efficiency control, and intelligent operation and maintenance. Against the backdrop of the national "dual-carbon" strategy and the in-depth advancement of power grid intelligent transformation and distribution automation, the power construction sector is gradually transitioning from traditional static filing and manual management models to a data-driven, end-to-end collaborative construction model, covering the entire lifecycle of substation planning, engineering design, construction and commissioning, operation and acceptance, and operation and maintenance. However, in actual power construction, problems such as data fragmentation, model disconnect, and lack of feedback mechanisms exist between different construction stages, leading to significant discrepancies between the substation's operational status and the recorded information. Therefore, there is an urgent need to construct a reversible, traceable, and learnable digital and intelligent collaborative construction method.

[0003] During the full-link construction of power distribution areas, there are three structural problems that cannot be reversed in management: First, during the construction and commissioning phases, if the change information of the measurement point configuration cannot be reverse-synchronized to the dispatch control layer, it is very easy to cause a serious disconnect between the actual operating equipment and the ledger parameters, which will lead to risks such as misidentification and misoperation in the later operation monitoring and control strategy deployment. Second, during the operation phase, if a dynamic comparison mechanism cannot be established between the operating status of the distribution area and the archive data, it will be difficult to trace and attribute the load anomalies, phase sequence disorders or topology misconnection problems that occur in the distribution area, and ultimately can only rely on manual experience to handle them, which is not only time-consuming but also cannot be systematically archived. Third, traditional reactive power compensation strategies and load forecasting models rely on static archive data and cannot respond to the dynamic changes in the operating status of the distribution area, resulting in control strategy failure and a decrease in forecast accuracy. Summary of the Invention

[0004] The purpose of this invention is to provide a collaborative management method and system for the full-chain digital and intelligent construction of power production, so as to solve the three structural problems of irreversible management mentioned in the background art.

[0005] To achieve the above objectives, the technical solution of the present invention is: a collaborative management method for the full-chain digital and intelligent construction of power production, comprising:

[0006] S1. During the construction phase, collect actual construction data and construct an actual construction dataset. Map and spatially register the actual construction dataset with the design structural reference parameters, construct a reverse mapping path for the actual construction parameters, and reverse synchronize the actual construction dataset to the design platform for parameter verification and configuration evaluation.

[0007] S2. During the debugging phase, monitor the on-site debugging commands to generate debugging event logs, structure the change information in the debugging event logs into measurement point mapping change data, and construct a reverse synchronization path for debugging change information based on the three-way binding structure of point-structured loops. Reverse synchronize the measurement point mapping change data to the target operating system and perform consistency verification of the point table mapping relationship.

[0008] S3. During the operation and maintenance phase, load measurement data and phase current waveform data of the transformer area are collected in real time. Based on phase difference analysis and load distribution law, the actual operating phase sequence and phase load distribution mode of the transformer area are identified, and a dynamic transformer area operation profile is constructed. The transformer area deviation attribution identification model is introduced to compare the dynamic transformer area operation profile with the static archive data to generate the transformer area operation deviation matrix. The transformer area operation deviation matrix is ​​then updated in reverse to the transformer area archive data, and the reactive power compensation control strategy and load prediction model parameters are dynamically adjusted.

[0009] S4. Real-time unified archiving of construction measurement datasets, measurement point mapping change data, and dynamic transformer area operation profiles.

[0010] Preferably, in step S1, the construction measurement data is elevation measurement data collected from the tunnel face of the altered rock tunnel. The elevation measurement data includes the coordinate data of multiple measuring points numbered sequentially according to the tunnel axis and their corresponding tunnel section numbering information, which constitutes the construction measurement dataset. The basic unit of the construction measurement dataset is the measuring point.

[0011] The design structural reference parameters are the design coordinate data and cross-sectional structural parameters exported from the engineering structural model before tunnel construction.

[0012] The construction measurement dataset is mapped and spatially registered with the design structural reference parameters to obtain a standardized construction measurement dataset.

[0013] Preferably, in S1, the reverse mapping path of construction measured parameters refers to mapping the registered construction measured data to the corresponding design structural reference parameter number using the structural component number as an index, and realizing the back transmission path of the construction measured data set to the design platform through data field comparison and timestamp association.

[0014] The method for constructing the reverse mapping path of the measured construction parameters includes:

[0015] A one-to-one correspondence is established between the standardized construction measurement dataset and the design structural reference parameters according to the axial direction of the structural components; a measurement point index table is established based on the standardized construction measurement dataset, marking the measurement point number, measurement point coordinate data and corresponding design component information; data comparison rules are set, and the standardized construction measurement dataset that meets the data comparison rules is back-transmitted to the design platform through the standard data interface.

[0016] Preferably, in S2, the debug event log refers to the set of raw log data generated by the debugger or the automatic recording mechanism during the equipment installation, system debugging and linkage testing phases.

[0017] The measurement point mapping change data refers to the parameterized record data extracted from the debugging event log, which is used to characterize the change in the spatial structure position of the measurement point number.

[0018] The process of structuring the change information in the debug event log into measurement point mapping change data specifically includes:

[0019] Extract entries related to measurement point number changes from the debugging event log, identify the measurement point number change field and coordinate offset field contained therein; construct measurement point change record units, sort them according to log timestamps, and extract valid continuous change events; associate and map the measurement point identifiers before and after the change with their corresponding structural component information, and generate structured measurement point mapping change data.

[0020] Preferably, in S2, the point-structured loop three-way binding structure refers to a three-element association structure established between the construction measurement point number, the structural component identification number, and the control loop node number;

[0021] The construction methods of the point-structured loop three-way binding structure include:

[0022] Extract the measurement point numbers and their corresponding structural component numbers from the standardized construction measurement dataset; based on the electrical control logic in the design model, parse the control loop node number corresponding to each structural component; using the component number as an intermediary, construct a point-to-loop three-way binding structure based on the measurement point number, structural component number, and control loop node number.

[0023] Preferably, in S2, the reverse synchronization path for debugging change information refers to a logical data feedback path constructed based on the three-way binding structure of the point-structured loop, used to synchronize the measurement point mapping change data to the target operating system via the design platform; the reverse synchronization path for debugging change information includes a dual verification mechanism of the control logic mapping table and the measurement point index table.

[0024] Preferably, in S3, the actual operating phase sequence of the transformer substation refers to the actual physical connection relationship of the three-phase conductors in the actual operating state of the transformer substation; the phase load distribution mode of the transformer substation refers to the load distribution pattern and phase relationship based on the actual measured load; the dynamic operating profile of the transformer substation refers to a structured description set that reflects the operating state of the transformer substation under different time periods and different load conditions, which is constructed by integrating load measurement data, phase current waveform data and phase difference analysis results, including a phase sequence structure diagram, a phase load dynamic distribution diagram and a time-series operating label sequence;

[0025] The construction of the dynamic station operation profile includes the following steps:

[0026] Based on the transformer substation operation monitoring data, the load measurement data is statistically normalized according to time windows to extract the load distribution pattern within each time window; phase difference analysis is performed using phase current waveform data to identify the three-phase current phase sequence structure within the corresponding time window; the load distribution pattern and phase sequence structure are integrated to label and construct the operation tag information corresponding to each time window; and the operation status tags of multiple time windows are constructed into a dynamic transformer substation operation profile according to the time sequence.

[0027] Preferably, in S3, the transformer area deviation attribution identification model is a transformer area operation deviation analysis and cause localization model constructed based on artificial intelligence algorithms. It is constructed based on dynamic transformer area operation profile, historical operation data curves, transformer area archive dataset and transformer area operation event log, and is used to compare the dynamic transformer area operation profile with the static transformer area archive data to generate a transformer area operation deviation matrix.

[0028] The method for constructing the transformer area deviation attribution identification model includes:

[0029] The dynamic transformer substation operation profile is decomposed into a time series to extract phase sequence evolution trends, load fluctuation patterns, and operation label sequences. A transformer substation topology map is constructed based on a graph neural network, with measurement point nodes, component nodes, and control nodes as graph structure inputs. The residual feature extraction method is used to identify the difference regions between the dynamic transformer substation operation profile and static archive parameters, forming an initial deviation vector set. An attribution classifier is introduced to classify the initial deviation vector set, output deviation cause labels, and calculate the influence weight of each deviation cause label through a weight backpropagation mechanism to generate a transformer substation operation deviation matrix.

[0030] Preferably, in step S3, updating the transformer area operation deviation matrix in reverse to the transformer area archive data specifically includes:

[0031] A field mapping table is established between the transformer area operation deviation matrix and static archive parameters, and a response binding relationship is established between the deviation cause label and the corresponding archive field. Based on the field mapping table, the archive fields with structural deviations are updated with weights and confidence corrections, and the transformer area archive dataset is updated. Based on the updated transformer area archive dataset, the reactive power compensation control strategy is dynamically reconstructed, including adjusting the compensation capacity threshold, compensation switching strategy, and compensation triggering logic. The load forecasting model parameters are then jointly corrected by combining the transformer area load change trend and abnormal operation mode.

[0032] On the other hand, the present invention provides a collaborative management system for the full-chain digital and intelligent construction of power production, including a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the aforementioned collaborative management method for the full-chain digital and intelligent construction of power production.

[0033] Compared with the prior art, the above-mentioned technical solution of the present invention has the following beneficial technical effects:

[0034] 1. In this invention, by establishing a reverse management mechanism at different construction stages such as construction, commissioning, and operation, on-site measured data, commissioning change information, and operational deviation results can be transmitted back to the design archives and control model in real time. This avoids the system inconsistency problem caused by the one-way flow of data at each stage in the prior art, which cannot be back-flowed for verification, and realizes closed-loop data management throughout the construction process.

[0035] 2. In this invention, by reversing the update of the operating deviation matrix of the transformer area to the archive data and dynamically adjusting the reactive power compensation strategy and load forecasting model parameters, the entire process of power construction is transformed from unidirectional drive to bidirectional feedback, which can continuously optimize strategy configuration and forecast accuracy at different stages. Attached Figure Description

[0036] Figure 1 This is a flowchart of one embodiment of the present invention. Detailed Implementation

[0037] Example 1, as Figure 1 As shown, the present invention proposes a collaborative management method for the entire chain of digital and intelligent construction in power production, the specific implementation steps of which are as follows:

[0038] S1. During the construction phase, collect actual construction data and construct an actual construction dataset. Map and spatially register the actual construction dataset with the design structural reference parameters, construct a reverse mapping path for the actual construction parameters, and reverse synchronize the actual construction dataset to the design platform for parameter verification and configuration evaluation.

[0039] S2. During the debugging phase, monitor the on-site debugging commands to generate debugging event logs, structure the change information in the debugging event logs into measurement point mapping change data, and construct a reverse synchronization path for debugging change information based on the three-way binding structure of point-structured loops. Reverse synchronize the measurement point mapping change data to the target operating system and perform consistency verification of the point table mapping relationship.

[0040] S3. During the operation and maintenance phase, load measurement data and phase current waveform data of the transformer area are collected in real time. Based on phase difference analysis and load distribution law, the actual operating phase sequence and phase load distribution mode of the transformer area are identified, and a dynamic transformer area operation profile is constructed. The transformer area deviation attribution identification model is introduced to compare the dynamic transformer area operation profile with the static archive data to generate the transformer area operation deviation matrix. The transformer area operation deviation matrix is ​​then updated in reverse to the transformer area archive data, and the reactive power compensation control strategy and load prediction model parameters are dynamically adjusted.

[0041] S4. Real-time unified archiving of construction measurement datasets, measurement point mapping change data, and dynamic transformer area operation profiles.

[0042] In this embodiment S1, the construction measurement data is the elevation measurement data collected based on the tunnel face of the altered rock tunnel. The elevation measurement data includes the coordinate data of multiple measuring points numbered sequentially according to the tunnel axis and their corresponding tunnel section numbering information, which constitutes the construction measurement dataset; the basic unit of the construction measurement dataset is the measuring point.

[0043] The design structural reference parameters are the design coordinate data and cross-sectional structural parameters exported from the engineering structural model before tunnel construction; wherein, the design coordinate data is the structural positioning information of preset measuring points in three-dimensional space, and the cross-sectional structural parameters include the tunnel excavation profile curvature, cross-sectional inclination angle and structural alignment control nodes.

[0044] The construction measurement dataset is mapped and spatially registered with the design structural reference parameters to obtain a standardized construction measurement dataset.

[0045] In this embodiment, each measuring point includes the following data fields: measuring point number, measuring point coordinate data, measurement timestamp, component number, and status identifier.

[0046] In this embodiment, the construction measurement dataset is mapped and spatially registered with the design structural reference parameters to obtain a standardized construction measurement dataset, as detailed below:

[0047] The construction measurement dataset and the design structural reference parameters are numbered one-to-one according to the tunnel axis sequence; the rotation matrix and translation vector between the construction measurement dataset and the design structural reference parameters are calculated using a three-dimensional rigid body transformation algorithm based on the minimum distance matching principle; the registration error is evaluated using a spatial registration residual function, and the spatial registration process is completed when the residual is less than a set threshold, and the standardized construction measurement dataset of the registered construction measurement points is output.

[0048] In this embodiment, the construction measurement dataset includes elevation point data, cross-sectional profile data, and deformation trend data collected by measurement equipment such as tunnel ground-penetrating radar, high-precision rangefinders, and automated inclinometers; the design structural reference parameters are the design cross-sectional line, main axis direction, and reference point coordinate information determined in the early stages of construction; during the data correspondence process, the construction measurement data point set and the design reference parameter point set are first numbered according to the tunnel axial sequence. The tunnel axial sequence refers to the linear sequence formed by arranging the measurement points or structural nodes along the main axis direction in space, based on the tunnel's main axis direction, to support the consistency of subsequent point-to-point registration logic; subsequently, based on the minimum... A three-dimensional rigid body transformation algorithm based on the distance matching principle is used to transform the coordinates of the construction measurement dataset. This algorithm calculates the rotation matrix and translation vector describing the spatial transformation relationship between the construction measurement points and the design points by minimizing the Euclidean distance residual between them, and completes the initial registration operation accordingly. To verify the registration accuracy, a spatial registration residual function is further introduced for error evaluation. This residual function uses the Euclidean distance between the registered point pairs as the evaluation index. When the residual is less than a set threshold (e.g., 10cm), the spatial registration is considered successful, and a standardized construction measurement dataset is output. The standardized construction measurement dataset is used for subsequent state parameter alignment processing with the altered rock prediction model.

[0049] In this embodiment S1, the reverse mapping path of construction measured parameters refers to mapping the registered construction measured data to the corresponding design structural reference parameter number using the structural component number as an index, and realizing the back transmission path of the construction measured data set to the design platform through data field comparison and timestamp association.

[0050] The method for constructing the reverse mapping path of the measured construction parameters includes:

[0051] A one-to-one correspondence is established between the standardized construction measurement dataset and the design structural reference parameters according to the axial direction of the structural components; a measurement point index table is established based on the standardized construction measurement dataset, marking the measurement point number, measurement point coordinate data and corresponding design component information; data comparison rules are set, and the standardized construction measurement dataset that meets the data comparison rules is back-transmitted to the design platform through the standard data interface.

[0052] In this embodiment, the reverse mapping path of construction measurement parameters is used to reverse synchronize the standardized construction measurement dataset to the design platform to achieve secondary verification and configuration optimization of the design structural reference parameters. Specifically, firstly, a one-to-one correspondence is established between the standardized construction measurement dataset and the design structural reference parameters according to the axial direction of the tunnel structural components, and a structural component-measuring point mapping index table is generated using the structural component number as the index field. Then, a measuring point index table is constructed based on the measuring point information in the standardized construction measurement dataset. The index table includes fields such as measuring point number, three-dimensional coordinate value, collection timestamp, and corresponding design component information. Data comparison rules are set, and a field-level comparison operation is performed using the structural component number as the primary key field to identify the differences between the measured and designed structural component parameters (such as embedment depth, cross-sectional geometry, and centerline offset). The standardized construction measurement dataset that meets the data comparison rules is packaged and sent back to the design platform in a standard data interface format as input data for structural parameter optimization verification, and configuration rationality analysis and design tolerance range verification are performed in the design platform.

[0053] In this embodiment, the structural component number refers to the coding information used to uniquely identify each tunnel component in the design structural reference parameters. These numbers are sequentially assigned according to the tunnel's axial direction, with each number corresponding to a specific design section, component location, and structural attributes. During the registration phase, the measured construction data establishes a mapping relationship with this number through coordinate space location, enabling precise data field alignment. Section geometric parameters refer to the spatial dimension parameters related to the profiles of each section contained in the design structural reference parameters, including but not limited to the maximum section width, bottom slab elevation, arch crown elevation, and sidewall thickness. These are all important indicators for structural stability assessment and deviation determination. Parameter extraction is based on the analysis of the registered three-dimensional coordinate point cloud. Timestamp association involves comparing the collection time field in each measured construction data record with the construction stage time interval in the design plan to determine whether the data was collected within the actual construction window period of the component. The time field includes the data collection start and end times, the planned construction stage time nodes, and the allowable offset. These three elements constitute the basis for the comparison rules used to determine time validity.

[0054] In this embodiment, the construction measurement dataset is back transmitted to the design platform via the construction measurement parameter reverse mapping path for parameter consistency verification. When a deviation between the construction parameters and the original design is detected, a re-evaluation process is triggered to automatically verify and adjust the design parameters of the affected structural components.

[0055] In this embodiment S2, the debugging event log refers to the original log data set generated by the debugging operator or the automatic recording mechanism during the equipment installation, system debugging and linkage testing phases. It includes component start and stop time, interface communication status, module configuration parameters, signal on / off information and abnormal record entries, which are used to perform structured tracking of the power-on and communication status of each component.

[0056] The measured point mapping change data refers to the parameterized record data extracted from the debugging event log, which is used to characterize the changes in the measured point number and the spatial structural position of the measured point number. It includes the measured point number before and after the change, coordinate offset, adjustment information of the mapped component number, and change timestamp, which is used for subsequent correction of the measured point structural mapping relationship and model update. Among them, the measured point is the basic unit of the construction measured dataset, which refers to the discrete data point with a unique number and spatial coordinate information collected by high-precision measuring equipment during the construction of the tunnel structure. This type of measured point constitutes the basic unit of the construction measured dataset.

[0057] The process of structuring the change information in the debug event log into measurement point mapping change data specifically includes:

[0058] Extract entries related to measurement point number changes from the debugging event log, identify the measurement point number change field and coordinate offset field contained therein; construct measurement point change record units, sort them according to log timestamps, and extract valid continuous change events; associate and map the measurement point identifiers before and after the change with their corresponding structural component information, and generate structured measurement point mapping change data containing number mapping, position offset, and mapping component revision information.

[0059] In this embodiment, the measuring point is the smallest unit of the construction measurement dataset. It refers to discrete point data with unique numbers and three-dimensional spatial coordinates collected by precision measuring tools such as total stations, laser rangefinders, or 3D scanning equipment during tunnel structure construction. Measuring points are usually attached to key nodes or monitoring and control positions of structural components to reflect the actual structural state and spatial positioning. Measuring point mapping change data refers to a type of structured data generated by extracting change records related to the measuring point number based on component configuration change information recorded in the commissioning event log. It is used to characterize changes such as number replacement, spatial position offset, or mapping component number adjustment that occur at a measuring point during the commissioning phase. The generation process includes: filtering change event entries containing measurement point number fields or spatial coordinate fields from the debugging event log; constructing change record units based on the measurement point number field and coordinate offset field, and cleaning and sorting the event chain according to the timestamp order to remove duplicate or invalid events; cross-comparing the measurement point numbers before and after the change with their corresponding structural component numbers to identify the adjustment of the mapping relationship; and finally generating a standardized measurement point mapping change dataset containing measurement point number change mapping, coordinate position offset vector, mapping component number revision and its corresponding timestamp, which is used for subsequent measurement point backtracking analysis, structural mapping relationship correction and design model update.

[0060] In this embodiment, component start-up and stop time refers to the specific timestamp record of each electrical component or control module from startup (power-on) to shutdown (power-off) during the debugging process, which is often used to judge the continuity of the component debugging status and the correctness of the timing logic; interface communication status refers to the data communication stability information after the connection between each component is established through communication channels such as bus, serial port or Ethernet, including whether the communication is successfully established, transmission delay, packet loss rate and abnormal interruption flag; module configuration parameters refer to the set of parameter values ​​set for the control module or subsystem during the debugging phase, including baud rate, device address, logic port number, limit threshold, etc., which are highly related to the final system operation configuration; signal on / off information refers to the real-time on or off status of input and output signals during the debugging process, used to determine the correctness of control signals and the consistency of response behavior; abnormal record entries refer to unexpected events automatically detected by the debugging system or manually recorded, including communication failure, hardware response anomalies, etc. The log includes fields for events such as common errors, parameter loading errors, and system self-test failures, along with an event number, affected area, and trigger time. The "Measurement Point Number Change" field is one of the structured fields in the debugging event log, typically used to record operations such as replacing, renaming, or deregistering measurement point numbers. The "Coordinate Offset" field refers to the three-dimensional vector parameter indicating the spatial position shift of a measurement point due to equipment replacement, construction deviations, or data corrections during debugging. It is used to assess the range of difference between the actual spatial position and the original measurement point position. The "Mapped Component Number" is the logical number that uniquely identifies a structural component in the construction or design model. During changes, this number may change due to component partitioning optimization, identification rule adjustments, or model revisions. Measurement points need to be re-bound to the new numbered component to maintain structural mapping consistency. The "Structural Mapping Relationship" refers to the one-to-one logical connection between the measurement point number and its bound structural component, used to support subsequent structural analysis, model reconstruction, and deviation assessment.

[0061] In this embodiment S2, the point-structured loop three-way binding structure refers to the three-element association structure established between the construction measurement point number, the structural component identification number and the control loop node number, which is used to realize the collaborative mapping and linkage verification of construction measurement data, design structural model and electrical loop logic;

[0062] The construction methods of the point-structured loop three-way binding structure include:

[0063] Extract the measurement point numbers and their corresponding structural component numbers from the standardized construction measurement dataset; based on the electrical control logic in the design model, parse the control loop node number corresponding to each structural component; using the component number as an intermediary, construct a point-to-loop three-way binding structure based on the measurement point number, structural component number, and control loop node number.

[0064] In this embodiment S2, the control loop node number refers to a unique logical identifier assigned to a controllable device, sensor, or actuator configured in a structural component within the control loop topology of the power system or electrical automation engineering. This control loop node number is used to accurately identify key points in the control link during system configuration, loop debugging, and data tracing, ensuring logical consistency and path traceability between the electrical control logic structure and physical components.

[0065] In this embodiment S2, the point-structured loop three-way binding structure supports the automatic identification of the control loop affiliation and logic response path of the measurement point in the debugging event log. When the measurement point location or component number changes, the system can automatically update its control loop node number through the binding relationship.

[0066] In this embodiment S2, the reverse synchronization path for debugging change information refers to a logical data feedback path constructed based on the three-way binding structure of the point-structured loop, which is used to synchronize the measurement point mapping change data to the target operating system via the design platform; the reverse synchronization path for debugging change information includes a dual verification mechanism of the control logic mapping table and the measurement point index table.

[0067] In this embodiment, the control logic mapping table is used to record the logical association information between each measurement point number in the operating system and its corresponding control loop node number. The control logic mapping table includes fields such as measurement point number, control loop node number, logical binding path, and control direction identifier, which facilitates logical consistency tracking after structural adjustments or number changes. The measurement point index table is used to maintain information such as the measurement point number, corresponding component number, spatial coordinates, and acquisition timestamp generated during the construction phase, serving as the basic index structure for the construction measurement dataset. The control logic mapping table and the measurement point index table constitute a dual verification mechanism, ensuring the consistency of the returned data in spatial positioning on the one hand, and guaranteeing its matching in logical path and control interface on the other hand.

[0068] In this embodiment, the measurement point mapping change data is reverse-synchronized to the target operating system to perform a consistency check of the point table mapping relationship, as follows:

[0069] The system extracts a table showing the correspondence between currently used measurement point numbers and structural components from the target operating system; it analyzes the measurement point mapping change data generated during the commissioning phase, including change records of measurement point numbers, corresponding coordinate position information, and adjustment information of the components to which they belong; it compares the commissioning change data with the original mapping data in the operating system to determine whether there are any conflicts in measurement point numbers, inconsistencies in the matching relationship between measurement points and components, or coordinate position offsets exceeding the allowable error range; if the consistency requirements are met, the updated measurement point numbers and corresponding component information are synchronously written into the database of the target operating system, and the synchronization results are recorded; if there are any inconsistencies, prompt information or verification records are automatically generated for subsequent manual verification or supplementary entry; the above verification process ensures that the measurement data at the construction site remains consistent with the point settings in the operating system after the update; the target operating system includes a SCADA system, a BIM system, and a loop diagram model.

[0070] In this embodiment S3, the actual operating phase sequence of the transformer substation refers to the actual physical connection relationship of the three-phase conductors in the actual operating state of the transformer substation; the phase load distribution mode of the transformer substation refers to the load distribution pattern and phase relationship based on the actual measured load distribution pattern; the dynamic operating profile of the transformer substation refers to the structured description set that reflects the operating state of the transformer substation under different time periods and different load conditions, which is constructed by integrating load measurement data, phase current waveform data and phase difference analysis results, including phase sequence structure diagram, phase load dynamic distribution diagram and time sequence operating label sequence;

[0071] The construction of the dynamic station operation profile includes the following steps:

[0072] Based on the transformer substation operation monitoring data, the load measurement data is statistically normalized according to time windows to extract the load distribution pattern within each time window; phase difference analysis is performed using phase current waveform data to identify the three-phase current phase sequence structure within the corresponding time window; the load distribution pattern and phase sequence structure are integrated to label and construct the operation tag information corresponding to each time window; and the operation status tags of multiple time windows are constructed into a dynamic transformer substation operation profile according to the time sequence.

[0073] In this embodiment S3, the load measurement data refers to the numerical data reflecting the power demand of each line or equipment, which is collected in real time by power monitoring equipment deployed at each key node in the distribution area. Specifically, it includes indicators such as active power, reactive power, voltage, current, and power factor. The phase current waveform data refers to the three-phase current time-domain waveform data obtained in a high sampling rate manner, which is used to reflect the amplitude, phase, and frequency characteristics of each phase current during operation and to analyze unbalanced load conditions and abnormal disturbance characteristics. The phase difference analysis refers to calculating the phase difference change trend between any two phases based on the phase angle difference between voltage and current extracted from the three-phase current waveform, so as to identify the phase sequence change caused by the change of load access mode in the power system. The load distribution law refers to the time-series statistics and cluster mining based on the load measurement data to identify the distribution pattern, load offset characteristics, and power flow direction trend of each node load in the three-phase system within a typical time period.

[0074] In this embodiment S3, the load measurement data is statistically normalized according to time windows, including calculating statistical indicators such as minimum, maximum, mean, and standard deviation, and then normalizing them to eliminate the offset caused by the difference in absolute values ​​between different time windows; phase difference analysis is performed using phase current waveform data to identify the three-phase current phase sequence structure within the corresponding time window; the three-phase current phase sequence structure is used to determine whether there are electrical abnormalities such as incorrect wiring phase sequence or unbalanced three-phase load during operation; the operation tag information includes phase sequence category, load concentration level, load imbalance degree index, etc., to classify the current operating status; the dynamic transformer area operation profile is structured with time as the horizontal axis and operation status tags as the vertical axis, forming a time-series evolution model that can be used to identify trend changes, abnormal fluctuations, and periodic patterns.

[0075] In this embodiment S3, the transformer area deviation attribution identification model is a transformer area operation deviation analysis and cause localization model constructed based on artificial intelligence algorithms. It is constructed based on dynamic transformer area operation profile, historical operation data curve, transformer area archive dataset and transformer area operation event log. It is used to compare the dynamic transformer area operation profile with the static transformer area archive data, generate a transformer area operation deviation matrix, and classify the sources of deviation and assess the degree of impact.

[0076] The method for constructing the transformer area deviation attribution identification model includes:

[0077] The dynamic transformer substation operation profile is decomposed into a time series to extract phase sequence evolution trends, load fluctuation patterns, and operation label sequences. A transformer substation topology map is constructed based on a graph neural network, with measurement point nodes, component nodes, and control nodes as graph structure inputs. The residual feature extraction method is used to identify the difference regions between the dynamic transformer substation operation profile and static archive parameters, forming an initial deviation vector set. An attribution classifier is introduced to classify the initial deviation vector set, output deviation cause labels, and calculate the influence weight of each deviation cause label through a weight backpropagation mechanism to generate a transformer substation operation deviation matrix.

[0078] In this embodiment, the transformer substation operation dataset consists of four types of heterogeneous data: dynamic transformer substation operation profile, historical operation data curve, transformer substation archive dataset, and transformer substation operation event log. The historical operation data curve is a time series composed of voltage, current, and load data from previous periods. The transformer substation archive dataset contains static structured information such as design electrical parameters, user structure, wiring method, and transformer model. The operation event log includes event records such as transformer substation tripping, load switching, and maintenance. The dynamic transformer substation operation profile and the static archive dataset are compared using a key field alignment strategy, including phase sequence structure comparison, load statistical index comparison, and equipment configuration consistency comparison. The comparison results are input to the residual extraction module to generate an initial set of deviation vectors.

[0079] In this embodiment, the dynamic transformer area operation profile is sliced ​​according to a fixed time window or based on an event-triggered mechanism. Wavelet transform is used to identify time-domain change features, and three core time-series features are extracted: phase sequence evolution trend, load fluctuation pattern, and operation tag sequence. The phase sequence evolution trend represents the trajectory of the phase difference between the three-phase currents changing over time. The load fluctuation pattern is used to characterize the rise, fall, and sudden changes in load values ​​within a selected time window. The operation tag sequence is an operation status identifier obtained by discretizing and encoding based on equipment status and load level, which is used to train a graph neural network structure to identify abnormal behavior patterns.

[0080] In this embodiment, a graph neural network is used to perform topological modeling of the transformer substation. The measurement point nodes (such as intelligent acquisition terminals), component nodes (such as cables and transformers), and control nodes (such as switches and circuit breakers) in the substation are abstracted as nodes in the graph. An adjacency matrix is ​​established based on the electrical connection relationship. The node attributes include equipment type, voltage level, phase, and current state. The edge attributes include connection resistance, cable length, and connection topology type. This graph structure is input into the graph neural network model for high-order relationship modeling and node feature aggregation.

[0081] In this embodiment, a residual feature extraction module performs a difference analysis on the dynamic transformer area operation profile and the transformer area archive dataset. A convolutional residual structure or differential feature channel is used to extract the difference regions between the two, identifying key areas that deviate from the archive parameters in terms of topology, phase sequence, or load pattern. These difference regions are encoded as a set of initial deviation vectors, each vector containing a node number, deviation type, deviation direction, and residual strength. Deviation cause labels include incorrect wiring phase sequence, uneven load distribution, equipment rewiring not recorded, and missing measurement data. An attribution classifier maps the initial deviation vector set to specific deviation cause categories, using a decision tree model, support vector machine, or lightweight neural network. The input is the multidimensional features of the initial deviation vector, and the output is the deviation cause label. The attribution classifier has an internal weight backpropagation mechanism, assigning influence weights to each deviation cause based on the confidence level of the classifier output, ultimately generating a transformer area operation deviation matrix.

[0082] In this embodiment S3, the operation deviation matrix of the transformer area is updated in reverse to the transformer area archive data, specifically including:

[0083] A field mapping table is established between the transformer area operation deviation matrix and static archive parameters, and a response binding relationship is established between the deviation cause label and the corresponding archive field. Based on the field mapping table, the archive fields with structural deviations are updated with weights and confidence corrections, and the transformer area archive dataset is updated. Based on the updated transformer area archive dataset, the reactive power compensation control strategy is dynamically reconstructed, including adjusting the compensation capacity threshold, compensation switching strategy, and compensation triggering logic. The load forecasting model parameters are then jointly corrected by combining the transformer area load change trend and abnormal operation mode.

[0084] In this embodiment, the reactive power compensation control strategy refers to the strategy of automatically or manually adjusting the compensation device under different operating conditions by using compensation methods such as capacitor banks and SVG devices to address the problem of voltage fluctuation and reactive power imbalance in the power distribution area. The load forecasting model parameters are a set of key control variables used to support the power load forecasting model in modeling, training, and predicting the load change trends of each phase over a future period. Among them, the power load forecasting model is a basic model commonly used in power systems to predict electrical load.

[0085] In this embodiment, the field mapping table is used to establish a one-to-one or one-to-many binding relationship between the deviation cause labels in the substation operation deviation matrix and the static field information in the substation archive dataset; the field mapping table defines rules through pre-configured deviation fields; the field weight update is used to adjust the weight contribution of the archive field in subsequent strategy generation and model prediction; the confidence correction process is used to requantify the confidence level of the field to guide the subsequent model fusion and parameter transfer strategies.

[0086] In this embodiment, the reactive power compensation control strategy includes adjusting the compensation capacity threshold, the compensation switching strategy, and the compensation triggering logic, as detailed below:

[0087] The compensation capacity threshold adjustment dynamically adjusts the minimum or maximum switching capacity threshold of each capacitor bank based on the power factor deviation and load fluctuation revealed by the operating deviation. The compensation switching strategy update reconstructs the switching logic of the compensation equipment (such as the start / stop voltage threshold of the parallel capacitors on the low-voltage side of the transformer) based on the phase shift, harmonic interference or sudden load phenomena revealed in the deviation assessment results, and optimizes voltage stability and power quality. The compensation trigger logic optimization introduces a predictive triggering mechanism based on load trends, which changes the reactive power compensation behavior from passive response to predictive triggering, such as activating the corresponding compensation equipment in advance when a significant load increase trend is detected.

[0088] Example 2: The present invention proposes a collaborative management system for the entire chain of digital and intelligent construction based on power production, which is applied to the collaborative management method for the entire chain of digital and intelligent construction based on power production proposed in Example 1. It includes a memory, a processor, and a computer program stored in the memory and executable on the processor. The processor executes the computer program to implement the collaborative management method for the entire chain of digital and intelligent construction based on power production in Example 1.

[0089] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited thereto. Various changes can be made within the scope of knowledge possessed by those skilled in the art without departing from the spirit of the present invention.

Claims

1. A collaborative management method for the entire chain of digital and intelligent construction in power production, characterized in that, Includes the following steps: S1. During the construction phase, collect actual construction data and construct an actual construction dataset. Map and spatially register the actual construction dataset with the design structural reference parameters, construct a reverse mapping path for the actual construction parameters, and reverse synchronize the actual construction dataset to the design platform for parameter verification and configuration evaluation. In S1, the reverse mapping path of construction measured parameters refers to mapping the registered construction measured data to the corresponding design structural reference parameter number using the structural component number as an index, and realizing the back transmission path of the construction measured data set to the design platform through data field comparison and timestamp association. The method for constructing the reverse mapping path of the measured construction parameters includes: A one-to-one correspondence is established between the standardized construction measurement dataset and the design structural reference parameters according to the axial direction of the structural components; a measurement point index table is established based on the standardized construction measurement dataset, marking the measurement point number, measurement point coordinate data and corresponding design component information; data comparison rules are set, and the standardized construction measurement dataset that meets the data comparison rules is transmitted back to the design platform through the standard data interface; S2. During the debugging phase, monitor the on-site debugging commands to generate debugging event logs, structure the change information in the debugging event logs into measurement point mapping change data, and construct a reverse synchronization path for debugging change information based on the three-way binding structure of point-structured loops. Reverse synchronize the measurement point mapping change data to the target operating system and perform consistency verification of the point table mapping relationship. In S2, the point-structure loop three-way binding structure refers to a three-element association structure established between the construction measurement point number, the structural component identification number, and the control loop node number; the debugging change information reverse synchronization path refers to a logical data feedback path built based on the point-structure loop three-way binding structure, used to synchronize the measurement point mapping change data to the target operating system via the design platform; the debugging change information reverse synchronization path includes a dual verification mechanism of the control logic mapping table and the measurement point index table; S3. During the operation and maintenance phase, load measurement data and phase current waveform data of the transformer area are collected in real time. Based on phase difference analysis and load distribution law, the actual operating phase sequence and phase load distribution mode of the transformer area are identified, and a dynamic transformer area operation profile is constructed. The transformer area deviation attribution identification model is introduced to compare the dynamic transformer area operation profile with the static archive data to generate the transformer area operation deviation matrix. The transformer area operation deviation matrix is ​​then updated in reverse to the transformer area archive data, and the reactive power compensation control strategy and load prediction model parameters are dynamically adjusted. In S3, the actual operating phase sequence of the transformer substation refers to the actual physical connection relationship of the three-phase conductors in the actual operating state of the substation; the phase load distribution mode of the transformer substation refers to the load distribution pattern and phase relationship based on the actual measured load; the dynamic operating profile of the transformer substation refers to a structured description set reflecting the operating state of the transformer substation under different time periods and different load conditions, which is constructed by integrating load measurement data, phase current waveform data and phase difference analysis results, including phase sequence structure diagram, phase load dynamic distribution diagram and time-series operating label sequence; the transformer substation deviation attribution identification model is a transformer substation operation deviation analysis and cause localization model constructed based on artificial intelligence algorithm, which is constructed based on dynamic operating profile of the transformer substation, historical operating data curve, transformer substation archive dataset and transformer substation operation event log, and is used to compare dynamic operating profile of the transformer substation with static archive data of the transformer substation to generate transformer substation operation deviation matrix; S4. Real-time unified archiving of construction measurement datasets, measurement point mapping change data, and dynamic transformer area operation profiles.

2. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 1, characterized in that: In S1, the construction measurement data is the elevation measurement data collected based on the tunnel face of the altered rock tunnel. The elevation measurement data includes the coordinate data of multiple measuring points numbered sequentially according to the tunnel axis and their corresponding tunnel section numbering information, which constitute the construction measurement dataset. The basic unit of the construction measurement dataset is the measuring point. The design structural reference parameters are the design coordinate data and cross-sectional structural parameters exported from the engineering structural model before tunnel construction. The construction measurement dataset is mapped and spatially registered with the design structural reference parameters to obtain a standardized construction measurement dataset.

3. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 2, characterized in that: In S2, the debug event log refers to the set of raw log data generated by debug operators or automatic recording mechanisms during the equipment installation, system debugging and linkage testing phases. The measurement point mapping change data refers to the parameterized record data extracted from the debugging event log, which is used to characterize the change in the spatial structure position of the measurement point number. The process of structuring the change information in the debug event log into measurement point mapping change data specifically includes: Extract entries related to the change of measurement point number from the debugging event log, and identify the measurement point number change field and coordinate offset field contained therein; Construct a measurement point change record unit, sort it according to the log timestamp, and extract valid continuous change events; associate and map the measurement point identifiers before and after the change with their corresponding structural component information, and generate structured measurement point mapping change data.

4. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 3, characterized in that: The construction methods of the point-structured loop three-way binding structure include: Extract the measurement point numbers and their corresponding structural component numbers from the standardized construction measurement dataset; based on the electrical control logic in the design model, parse the control loop node number corresponding to each structural component; using the component number as an intermediary, construct a point-to-loop three-way binding structure based on the measurement point number, structural component number, and control loop node number.

5. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 4, characterized in that: The construction of the dynamic station operation profile includes the following steps: Based on the transformer substation operation monitoring data, the load measurement data is statistically normalized according to time windows to extract the load distribution pattern within each time window; phase difference analysis is performed using phase current waveform data to identify the three-phase current phase sequence structure within the corresponding time window; the load distribution pattern and phase sequence structure are integrated to label and construct the operation tag information corresponding to each time window; and the operation status tags of multiple time windows are constructed into a dynamic transformer substation operation profile according to the time sequence.

6. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 5, characterized in that: The method for constructing the transformer area deviation attribution identification model includes: The dynamic transformer substation operation profile is decomposed into a time series to extract phase sequence evolution trends, load fluctuation patterns, and operation label sequences. A transformer substation topology map is constructed based on a graph neural network, with measurement point nodes, component nodes, and control nodes as graph structure inputs. The residual feature extraction method is used to identify the difference regions between the dynamic transformer substation operation profile and static archive parameters, forming an initial deviation vector set. An attribution classifier is introduced to classify the initial deviation vector set, output deviation cause labels, and calculate the influence weight of each deviation cause label through a weight backpropagation mechanism to generate a transformer substation operation deviation matrix.

7. The collaborative management method for the full-chain digital and intelligent construction of power production according to claim 6, characterized in that: In step S3, the operation deviation matrix of the transformer area is updated in reverse to the transformer area archive data, specifically including: A field mapping table is established between the transformer area operation deviation matrix and static archive parameters, and a response binding relationship is established between the deviation cause label and the corresponding archive field. Based on the field mapping table, the archive fields with structural deviations are updated with weights and confidence corrections, and the transformer area archive dataset is updated. Based on the updated transformer area archive dataset, the reactive power compensation control strategy is dynamically reconstructed, including adjusting the compensation capacity threshold, compensation switching strategy, and compensation triggering logic. The load forecasting model parameters are jointly corrected by combining the transformer area load change trend and operation anomaly mode. Among them, the field mapping relationship table is used to establish a one-to-one or one-to-many binding relationship between the deviation cause labels in the substation operation deviation matrix and the static field information in the substation archive dataset; the weight update is used to adjust the weight contribution of the archive field in subsequent strategy generation and model prediction; and the confidence correction is used to requantify the confidence level of the field.

8. A collaborative management system for the entire chain of digital and intelligent construction based on power production, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: The processor executes a computer program to implement the collaborative management method for the full-chain digital and intelligent construction of power production as described in any one of claims 1-7.

Images