A digital twin-based approach and system for the full lifecycle management of power projects

Abstract

This invention belongs to the field of engineering management, specifically relating to a method and system for the full lifecycle management of power projects based on digital twins. It addresses the technical problems of the disconnect between digital twins and construction plans, the lack of real-time data-driven risk assessment, and the absence of a permanent data assetization mechanism. The management method includes the following steps: S1, acquiring the construction plan and digital twin model of the task unit; S2, establishing the association between the task unit and the component; S3, calculating the vector deviation between the on-site physical state and the target state of the digital twin model; calculating the risk coefficient of the component; and forming a dynamic construction plan; S4, structurally encapsulating and hashing the vector deviation, risk coefficient, quality acceptance results, and resource consumption data to generate construction process data bound to the component. This overcomes the shortcomings of the disconnect between plans and models in traditional management, can predict and avoid potential future spatial and resource conflicts, and ensures the integrity and authenticity of process data.

Description

Technical Field

[0001] This invention belongs to the field of engineering management, specifically relating to a method and system for the full life cycle management of power projects based on digital twins. Background Technology

[0002] Power projects are characterized by large investment scale, long construction period, high technical complexity, and numerous participants, placing higher demands on the refinement of full life cycle management. Digital twin or Building Information Modeling (BIM) technology provides a unified foundation for project design, construction, and operation and maintenance by constructing high-precision 3D visualization models. Meanwhile, project management typically relies on network planning techniques, such as Critical Path Method (CPM) and Program Evaluation and Review Technique (PERT), to develop and control construction schedules.

[0003] However, in current practice, digital twin models and construction plans are often two relatively independent management dimensions. The relationship between the two is mostly limited to the macro level. There is a lack of a mechanism to bind specific construction task units with the components in the model in terms of time, space and process. This makes it difficult for the construction plan to reflect the status of the entity, and it is also difficult to provide real-time and automatic feedback on the actual progress and deviations on site to adjust the plan. This affects the collaborative efficiency of project management and the accuracy of decision-making.

[0004] Traditional project management exhibits significant lag and passivity when facing dynamic changes on construction sites: information such as on-site quality inspection and progress tracking is mainly collected through manual inspections and periodic reports, resulting in data transmission delays and susceptibility to distortion. When deviations or quality issues arise during construction, the cascading impact on subsequent processes typically relies on the project manager's experience for qualitative judgment and manual adjustments, lacking a real-time data-driven risk assessment and plan optimization system. This makes it difficult for managers to predict and mitigate potential conflicts arising from localized problems. Furthermore, a large amount of valuable data generated during construction, such as quality acceptance records, resource consumption details, and deviation measurements, is usually archived in a discrete and unstructured format, failing to establish a one-to-one, traceable, and permanent link with the components in the digital twin model. This not only leads to the loss of process data assets but also poses significant challenges to later project operation and maintenance, accountability, and knowledge reuse. Summary of the Invention

[0005] This invention provides a method and system for full life-cycle management of power projects based on digital twins, in order to solve the technical problems in the prior art where digital twins are disconnected from construction plans and lack real-time data-driven risk assessment and permanent data assetization mechanisms.

[0006] In a first aspect, the present invention provides a method for full lifecycle management of power projects based on digital twins, comprising the following steps:

[0007] S1, obtain the construction plan that is decomposed into task units with time and resource requirements, and the digital twin model that is divided into components with spatial location and engineering attributes.

[0008] S2, if there is a time overlap between the planned time window of the task unit and the predetermined construction cycle of the component, and there is a spatial overlap between the working space of the task unit and the spatial position of the component, and the construction object of the task unit is the component in the preset process flow, then the association between the task unit and the component is established.

[0009] S3. Collect on-site physical state information of the component, and calculate the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints; calculate the risk coefficient of the component based on the magnitude and direction of the vector deviation; adjust the planned time window of subsequent task units that have a process flow dependency relationship with the component based on the risk coefficient to form a dynamic construction plan; predict the probability of space and resource conflicts between task units in future cycles based on the dynamic construction plan and the space occupancy model of the task units.

[0010] S4. After the task unit is completed, the vector deviation, risk coefficient, quality acceptance results and resource consumption data are structured, encapsulated and hashed to generate construction process data bound to the component, and the construction process data is stored in the digital twin model as a permanent non-geometric attribute of the component.

[0011] Furthermore, in S2, the association between task units and components is established, including:

[0012] The planned time window of the task unit With the planned construction period of the components Compare, when satisfied At that time, the time condition is met; among which, , These are the start and end times of the planned time window, respectively. , These are the start and end times of the scheduled construction period, respectively.

[0013] The intersection operation is performed between the three-dimensional bounding box of the task unit's work space and the three-dimensional bounding box of the component's spatial position. When the overlapping volume is greater than a preset threshold, the spatial condition is deemed to be met.

[0014] In the pre-defined directed acyclic graph of the process flow, check whether there is a direct directed edge from the node of the task unit to the component node. If it exists, the process flow conditions are deemed to be met.

[0015] When all three conditions are met, it is determined that there is an association between the task unit and the component.

[0016] Furthermore, the on-site physical state information of the components is collected, and based on geometric and physical constraints, the vector deviation between the on-site physical state and the target state of the digital twin model is calculated, including:

[0017] A 3D scanning device is used to scan the actual physical components on site to obtain registered 3D point cloud data. ;

[0018] Geometric model for extracting the target state of components from a digital twin model And pre-determine N key feature points on its surface. , ;

[0019] In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ;

[0020] Calculate the deviation vector from the key feature point to the matching point. ;

[0021] The vector deviation of the component is obtained by averaging all N deviation vectors. .

[0022] Further, based on the magnitude and direction of the vector deviation, the risk coefficient of the component is calculated, including:

[0023] Set the allowable deviation threshold for the component ;

[0024] Calculate vector deviation modulus and the absolute value of its vertical component. ;

[0025] The risk coefficient is calculated using the following formula. : ;

[0026] in, The weighting coefficients for the deviation modulus. This represents the weighting coefficient for the influence of the vertical component.

[0027] Furthermore, adjusting the planned time window for subsequent task units that have a process flow dependency with the component based on the aforementioned risk coefficient includes:

[0028] In the directed acyclic graph of the process flow, identify all subsequent task units that have a direct dependency relationship with the task unit that completes the current component;

[0029] The planned delay time for each subsequent task unit is calculated using the following nonlinear function. : ;in, Based on the time delay factor, For risk sensitivity coefficient, It is the risk factor;

[0030] The original planned start and end times of subsequent task units will be shifted backward. This creates a new planning time window.

[0031] Furthermore, the vector deviation, risk coefficient, quality acceptance results, and resource consumption data are structured, encapsulated, and hashed to generate construction process data bound to the component. This construction process data is then stored as a permanent non-geometric attribute of the component in the digital twin model, including:

[0032] Create a structured encapsulation template containing the following fields: component unique identifier, timestamp, and vector deviation. Risk coefficient The hash values ​​of quality acceptance results, resource consumption data, and data from the previous construction process;

[0033] Serialize the content of the structured encapsulation template into a string;

[0034] The hash value of the string is calculated using a hash algorithm, and the hash value is used as a unique identifier for the current construction process data;

[0035] The structured encapsulation template and its identifier are stored in the component attribute database of the digital twin model.

[0036] Secondly, the present invention provides a power project lifecycle management system based on digital twins, comprising:

[0037] The decomposition module obtains construction plans that are decomposed into task units with time and resource requirements, as well as digital twin models of components that are segmented into components with spatial location and engineering attributes.

[0038] In the association module, if the planned time window of the task unit and the predetermined construction cycle of the component overlap in time, the working space of the task unit and the spatial position of the component overlap in space, and the construction object of the task unit is the component in the preset process flow, then the association relationship between the task unit and the component is established.

[0039] The construction plan generation module collects on-site physical state information of components and calculates the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints. It then calculates the risk coefficient of the component based on the magnitude and direction of the vector deviation. Based on the risk coefficient, it adjusts the planned time windows of subsequent task units that have a process flow dependency relationship with the component to form a dynamic construction plan. Finally, based on the dynamic construction plan and the space occupancy model of the task units, it predicts the probability of space and resource conflicts between task units in future cycles.

[0040] The data storage module, after the task unit is completed, encapsulates and hashes the vector deviation, risk coefficient, quality acceptance results and resource consumption data in a structured manner to generate construction process data bound to the component, and stores the construction process data as a permanent non-geometric attribute of the component in the digital twin model.

[0041] Furthermore, in the association module, the association relationship between task units and components is established, including:

[0042] The planned time window of the task unit With the planned construction period of the components Compare, when satisfied At that time, the time condition is met; among which, , These are the start and end times of the planned time window, respectively. , These are the start and end times of the scheduled construction period, respectively.

[0043] The intersection operation is performed between the three-dimensional bounding box of the task unit's work space and the three-dimensional bounding box of the component's spatial position. When the overlapping volume is greater than a preset threshold, the spatial condition is deemed to be met.

[0044] In the pre-defined directed acyclic graph of the process flow, check whether there is a direct directed edge from the node of the task unit to the component node. If it exists, the process flow conditions are deemed to be met.

[0045] When all three conditions are met, it is determined that there is an association between the task unit and the component.

[0046] Furthermore, the construction plan generation module collects the on-site physical state information of the components and, based on geometric and physical constraints, calculates the vector deviation between the on-site physical state and the target state of the digital twin model, including:

[0047] A 3D scanning device is used to scan the actual physical components on site to obtain registered 3D point cloud data. ;

[0048] Geometric model for extracting the target state of components from a digital twin model And pre-determine N key feature points on its surface. , ;

[0049] In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ;

[0050] Calculate the deviation vector from the key feature point to the matching point. ;

[0051] The vector deviation of the component is obtained by averaging all N deviation vectors. .

[0052] Furthermore, in the construction plan generation module, the risk coefficient of the component is calculated based on the magnitude and direction of the vector deviation, including:

[0053] Set the allowable deviation threshold for the component ;

[0054] Calculate vector deviation modulus and the absolute value of its vertical component. ;

[0055] The risk coefficient is calculated using the following formula. : ;

[0056] in, The weighting coefficients for the deviation modulus. This represents the weighting coefficient for the influence of the vertical component.

[0057] The beneficial effects are as follows: This invention overcomes the shortcomings of traditional management's disconnect between plans and models by establishing a temporal, spatial, and technological correlation between construction plans and digital twin models, thus achieving a mapping between virtual models and on-site construction. Based on the vector deviation between the on-site physical state and the target state of the digital twin model, the risk coefficient of components is calculated. Adjustments are then made to subsequent task plans with technological dependencies based on the risk coefficient, enabling the prediction and avoidance of potential spatial and resource conflicts, thereby improving the predictability and collaborative efficiency of project management. Furthermore, by constructing construction process data permanently bound to components, key information such as vector deviations, risk coefficients, quality acceptance results, and resource consumption data are structured and stored. This not only ensures the integrity and authenticity of process data but also enriches the data in the digital twin model, providing data support for operation and maintenance and accountability throughout the project's entire lifecycle. Attached Figure Description

[0058] Figure 1A flowchart illustrating a digital twin-based approach to the full lifecycle management of power projects;

[0059] Figure 2 This is a schematic diagram illustrating the decomposition of the construction plan and digital twin model.

[0060] Figure 3 A schematic diagram for establishing the relationship between task units and components. Detailed Implementation

[0061] An embodiment of the digital twin-based power project lifecycle management method provided by this invention:

[0062] like Figure 1 As shown, the digital twin-based full lifecycle management method for power projects includes the following steps:

[0063] S1, obtain the construction plan that is decomposed into task units with time and resource requirements, and the digital twin model that is divided into components with spatial location and engineering attributes.

[0064] Specifically, the interface program reads the XML-formatted construction plan file exported by the project management software, parses it to obtain each independent construction task unit, extracts the planned start time, planned completion time, and duration as time attributes, and extracts the required labor, machinery, and material lists as resource requirement attributes, such as... Figure 2 As shown.

[0065] Simultaneously, load the digital twin model in Industrial Foundation Class (IFC) or Revit (RVT) format, traverse the model database to extract each individual engineering component (such as a beam or a foundation slab); read its three-dimensional coordinates and bounding box volume as spatial location attributes, and read its unique component ID, material, design designation, etc., as engineering attributes, such as... Figure 2 As shown.

[0066] S2, if there is a time overlap between the planned time window of the task unit and the predetermined construction cycle of the component, and there is a spatial overlap between the working space of the task unit and the spatial position of the component, and the construction object of the task unit is the component in the preset process flow, then the association between the task unit and the component is established.

[0067] Specifically, the recommendations regarding the relationships between task units and components need to be filtered through a three-tiered logical process using the following steps:

[0068] 1. Time Matching: Traverse all task units and components, and filter out candidate pairs where the planned start and end date range of the task unit overlaps with the planned construction period defined in the component attributes.

[0069] 2. Spatial matching: For the selected candidate pairs, a virtual work space bounding box larger than the component's physical bounding box is generated for the task unit (for example, extending outward by two meters in each of the component's XYZ directions); then it is determined whether the virtual work space bounding box overlaps with the component's own spatial bounding box to ensure the accessibility of the operation.

[0070] 3. Process Logic Matching: The built-in process rule library is queried. For example, the rule library defines that task units with names containing "pouring" should be associated with components of type "concrete foundation." Based on these rules, candidate pairs that meet the spatiotemporal conditions are finally confirmed. For example, a formal data association is established between the task unit named "Pouring the No. 3 Main Transformer Foundation" and the component with ID ZJ-03 in the digital twin model. Figure 3 As shown.

[0071] In an optional embodiment, if the planned time window of the task unit overlaps with the predetermined construction cycle of the component, the working space of the task unit overlaps with the spatial location of the component, and the construction object of the task unit is the component in the preset process flow, then the association between the task unit and the component is established, including:

[0072] The planned time window of the task unit With the planned construction period of the components Compare, when satisfied At that time, the time condition is met; among which, , These are the start and end times of the planned time window, respectively. , These are the start and end times of the scheduled construction period, respectively.

[0073] The intersection operation is performed between the three-dimensional bounding box of the task unit's work space and the three-dimensional bounding box of the component's spatial position. When the overlapping volume is greater than a preset threshold, the spatial condition is deemed to be met.

[0074] In the pre-defined directed acyclic graph of the process flow, check whether there is a direct directed edge from the node of the task unit to the component node. If it exists, the process flow conditions are deemed to be met.

[0075] When all three conditions are met, it is determined that there is an association between the task unit and the component.

[0076] Taking the installation of curtain wall panel M01 as an example, the establishment of the association requires the following three steps:

[0077] 1. Time Matching Judgment: The planned time window for the task unit is from 8:00 to 17:00 on October 10, 2023. The scheduled construction period for curtain wall panel M01 is from October 9 to October 11, 2023. Since the planned time window of the task unit overlaps with the construction period of the component, the time conditions are met.

[0078] 2. Spatial Matching Judgment: The working space of the installation task is a three-dimensional bounding box, covering the range of motion of the hoisting robot and the operating area of ​​the installation personnel, with a volume of 50. The spatial location of curtain wall panel M01 has a three-dimensional bounding box volume of 5. After the intersection operation, the overlap volume of the two bounding boxes is 4.8. If the preset threshold for overlapping volume is 80% of the component volume (i.e., 4...), then... Since 4.8 is greater than 4, the spatial conditions are met.

[0079] 3. Process flow judgment: In the project's preset process flow directed acyclic graph, there is a direct directed edge between the task node of installing curtain wall panel M01 and the component node of curtain wall panel M01, indicating that the task directly acts on this component, so the process flow conditions are met.

[0080] Since the three conditions of time, space, and process are all met, there is a correlation between this installation task and the curtain wall panel M01.

[0081] S3. Collect on-site physical state information of the component, and calculate the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints; calculate the risk coefficient of the component based on the magnitude and direction of the vector deviation; adjust the planned time window of subsequent task units that have a process flow dependency relationship with the component based on the risk coefficient to form a dynamic construction plan; predict the probability of space and resource conflicts between task units in future cycles based on the dynamic construction plan and the space occupancy model of the task units.

[0082] Specifically, taking the management of a concrete foundation after construction as an example, a 3D laser scanner is used to scan the entity, acquiring high-density 3D point cloud data as its on-site entity status information. The system then performs coordinate registration and alignment between this 3D point cloud data and the theoretical design model of the concrete foundation in the digital twin model. Next, at preset key checkpoints in the digital twin model (such as the center point of the top surface and the four corner points of the concrete foundation), the corresponding actual location points in the 3D point cloud data are identified. Finally, the vector deviation from the theoretical position in the digital twin model to the actual position on-site at each checkpoint is calculated. This vector deviation includes deviation components on the X, Y, and Z axes; for example, the vector deviation for a top corner point is 5mm positive on the X-axis, 3mm negative on the Y-axis, and 8mm positive on the Z-axis.

[0083] A dynamic risk coefficient value between 0 and 1 is calculated based on a preset risk assessment model. In this model, the dynamic risk coefficient is equal to the sum of the absolute values ​​of the deviation components in each direction multiplied by the directional weight coefficients. The weight coefficient for the vertical direction (Z-axis) is typically set higher. Subsequently, the system queries the logical relationships in the construction plan to find all subsequent task units that have this concrete foundation construction task as their immediate predecessor. For example, this subsequent task unit might be the equipment installation task on the concrete foundation. Based on the dynamic risk coefficient value, the delay days for subsequent tasks are calculated using a non-linear function (such as a piecewise function), thus updating and forming a dynamic construction plan. Specifically: if the dynamic risk coefficient value is below 0.2, no delay is allowed; if the dynamic risk coefficient value is between 0.2 and 0.6, a 3-day delay is allowed for rectification; if the dynamic risk coefficient value is above 0.6, a 10-day delay is allowed for reinforcement and re-inspection.

[0084] Based on the updated dynamic construction plan, the system predicts the work space occupancy of all task units within the next two weeks. If two different tasks are found to have overlapping work space on the same day, it is marked as a high-probability space conflict, prompting the manager to intervene in advance.

[0085] In an optional embodiment, the on-site physical state information of the component is acquired, and based on geometric and physical constraints, the vector deviation between the on-site physical state and the target state of the digital twin model is calculated, including:

[0086] A 3D scanning device is used to scan the actual physical components on site to obtain registered 3D point cloud data. ;

[0087] Geometric model for extracting the target state of components from a digital twin model And pre-determine N key feature points on its surface. , ;

[0088] In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ;

[0089] Calculate the deviation vector from the key feature point to the matching point. ;

[0090] The vector deviation of the component is obtained by averaging all N deviation vectors. .

[0091] Specifically, taking a steel column that has already been hoisted into place as an example, a terrestrial 3D laser scanner is used to scan the steel column, acquiring 3D point cloud data of millions of points. After registration with the building's overall coordinate system, the 3D point cloud data forms the on-site entity state data. The design model of the steel column is retrieved from the digital twin model, and 20 key feature points are predefined on its surface. These key feature points can be the top and bottom corners, as well as points at other key connection locations. For example, the design coordinates of one of the top corners are (10.000m, 20.000m, 15.000m). For this key feature point, the closest spatially matching point is found in the 3D point cloud data. Assuming the coordinates of this matching point are (10.008m, 20.005m, 14.997m), a deviation vector is calculated. Its components on the three coordinate axes are 0.008m, 0.005m, and -0.003m, respectively. The above calculation is repeated for all 20 key feature points to obtain 20 independent deviation vectors. 20 deviation vectors The corresponding components are added together and the average value is taken to obtain the final vector deviation. It reflects the overall translational and rotational deviation of the entire steel column relative to its design position.

[0092] In an optional embodiment, the risk factor of the component is calculated based on the magnitude and direction of the vector deviation, including:

[0093] Set the allowable deviation threshold for the component ;

[0094] Calculate vector deviation modulus and the absolute value of its vertical component. ;

[0095] The risk coefficient is calculated using the following formula. : ;

[0096] in, The weighting coefficients for the deviation modulus. This represents the weighting coefficient for the influence of the vertical component.

[0097] Taking the aforementioned steel column as an example, its vector deviation calculation results show components of 8mm, 5mm, and -3mm on the X, Y, and Z axes, respectively. According to construction specifications, the allowable deviation threshold for such components is set at 10mm. Weighting coefficients are set based on engineering experience; for example, the overall deviation is considered more important, with a weighting coefficient of 0.6 for the deviation modulus, while the weighting coefficient for the vertical component is 0.4. Calculation process: First, the modulus of the vector deviation is calculated. The modulus is the length of the vector deviation. mm. Simultaneously, obtain the absolute value of the vertical component. mm. Substituting these values ​​into the formula yields the risk coefficient. It is 0.714.

[0098] In an optional embodiment, adjusting the planned time window of subsequent task units that have a process flow dependency with the component based on the risk coefficient includes:

[0099] In the directed acyclic graph of the process flow, identify all subsequent task units that have a direct dependency relationship with the task unit that completes the current component;

[0100] The planned delay time for each subsequent task unit is calculated using the following nonlinear function. : ;in, Based on the time delay factor, For risk sensitivity coefficient, It is the risk factor;

[0101] The original planned start and end times of subsequent task units will be shifted backward. This creates a new planning time window.

[0102] Risk coefficient based on monetary amount Taking a steel column with a length of 0.714 as an example, the following adjustments were made to the subsequent task plan: By consulting the process flow chart, it was identified that the installation of steel beam B2 and the welding of node J5 both depend on the placement of the steel column. Therefore, these two tasks were determined to be subsequent tasks requiring plan adjustments. In this project, the basic time delay factor... The delay time is 2 hours, with a risk sensitivity coefficient of 1.5. The greater the risk, the more exponentially the delay will increase. Calculations show a planned delay of approximately 3.84 hours, rounded to 4 hours. Taking the installation of steel beam B2 as an example, with a 4-hour delay, the original planned start time for steel beam B2 installation was 9:00 AM the following day, and the end time was 11:00 AM. After adjustment, the new planned start time will be postponed by 4 hours to 1:00 PM, and the end time will be correspondingly postponed to 3:00 PM. This is based on a risk factor... The function calculation enables dynamic and intelligent management of construction progress. It ensures that when the deviation risk is small, the plan is less affected; while when the risk increases significantly, it automatically reserves more time to deal with possible retesting, correction or rework, thereby reducing the risk of conflict in subsequent tasks.

[0103] S4. After the task unit is completed, the vector deviation, risk coefficient, quality acceptance results and resource consumption data are structured, encapsulated and hashed to generate construction process data bound to the component, and the construction process data is stored in the digital twin model as a permanent non-geometric attribute of the component.

[0104] Specifically, after the construction and acceptance of the aforementioned concrete foundation are completed, a JSON data record is created. This record encapsulates the calculated vector deviations of all key points, risk coefficient values, quality acceptance results (qualified or unqualified) entered by the on-site quality inspector, and resource consumption data extracted from the construction log (such as actual concrete usage and labor hours). The SHA256 hash value of the JSON data record content is calculated, and this hash value, along with the JSON data record itself, forms a single data unit. If the concrete foundation component has prior construction process data, the hash value of the previous construction process data is also included in the current JSON data record, forming a chain structure. The complete JSON data record string is added as a new custom text attribute to the attribute set of the concrete foundation component in the digital twin model database, achieving a permanent binding between data and the component.

[0105] In an optional embodiment, vector deviation, risk coefficient, quality acceptance results, and resource consumption data are structured, encapsulated, and hashed to generate construction process data bound to the component. This construction process data is then stored as a permanent non-geometric attribute of the component in the digital twin model, including:

[0106] Create a structured encapsulation template containing the following fields: component unique identifier, timestamp, and vector deviation. Risk coefficient The hash values ​​of quality acceptance results, resource consumption data, and data from the previous construction process;

[0107] Serialize the content of the structured encapsulation template into a string;

[0108] The hash value of the string is calculated using a hash algorithm, and the hash value is used as a unique identifier for the current construction process data;

[0109] The structured encapsulation template and its identifier are stored in the component attribute database of the digital twin model.

[0110] After completing the risk assessment and plan adjustments for the steel column, a new construction process data record is generated. This record includes a unique identifier for the steel column (e.g., SZ-08-01) and a timestamp of the current operation (e.g., October 10, 2023, 14:45). The record also includes key quality status data, such as the components of the vector deviation (8mm, 5mm, -3mm) and the calculated dynamic risk coefficient of 0.714. Furthermore, the results of this quality acceptance inspection (marked as pending review) and resource consumption information are entered, such as tower crane A's 0.5 hours of use and installation team B's 2 man-hours. To ensure data traceability, the record also includes a link to the previous construction process data record for this component, such as the hash value of the factory inspection record.

[0111] All the above information (including component identification, timestamp, vector deviation, risk coefficient, quality acceptance results, resource consumption data, and the hash value of the previous construction process data) is merged and serialized into a long string. A hash algorithm such as SHA-256 is used to calculate a unique, fixed-length hash value, which is a string of letters and numbers. This newly generated hash value not only serves as a unique identifier for the current construction process data but also ensures the data's tamper-proof properties. The data record containing all the information and its own hash identifier is stored in the database associated with steel column SZ-08-01 in the digital twin platform, forming an immutable chain of data records.

[0112] An embodiment of the digital twin-based power project lifecycle management system provided by this invention:

[0113] A digital twin-based power project lifecycle management system includes:

[0114] The decomposition module obtains construction plans that are decomposed into task units with time and resource requirements, as well as digital twin models of components that are segmented into components with spatial location and engineering attributes.

[0115] In the association module, if the planned time window of the task unit and the predetermined construction cycle of the component overlap in time, the working space of the task unit and the spatial position of the component overlap in space, and the construction object of the task unit is the component in the preset process flow, then the association relationship between the task unit and the component is established.

[0116] The construction plan generation module collects on-site physical state information of components and calculates the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints. It then calculates the risk coefficient of the component based on the magnitude and direction of the vector deviation. Based on the risk coefficient, it adjusts the planned time windows of subsequent task units that have a process flow dependency relationship with the component to form a dynamic construction plan. Finally, based on the dynamic construction plan and the space occupancy model of the task units, it predicts the probability of space and resource conflicts between task units in future cycles.

[0117] The data storage module, after the task unit is completed, encapsulates and hashes the vector deviation, risk coefficient, quality acceptance results and resource consumption data in a structured manner to generate construction process data bound to the component, and stores the construction process data as a permanent non-geometric attribute of the component in the digital twin model.

[0118] Furthermore, in the association module, the association relationship between task units and components is established, including:

[0119] The planned time window of the task unit With the planned construction period of the components Compare, when satisfied At that time, the time condition is met; among which, , These are the start and end times of the planned time window, respectively. , These are the start and end times of the scheduled construction period, respectively.

[0120] The intersection operation is performed between the three-dimensional bounding box of the task unit's work space and the three-dimensional bounding box of the component's spatial position. When the overlapping volume is greater than a preset threshold, the spatial condition is deemed to be met.

[0121] In the pre-defined directed acyclic graph of the process flow, check whether there is a direct directed edge from the node of the task unit to the component node. If it exists, the process flow conditions are deemed to be met.

[0122] When all three conditions are met, it is determined that there is an association between the task unit and the component.

[0123] Furthermore, the construction plan generation module collects the on-site physical state information of the components and, based on geometric and physical constraints, calculates the vector deviation between the on-site physical state and the target state of the digital twin model, including:

[0124] A 3D scanning device is used to scan the actual physical components on site to obtain registered 3D point cloud data. ;

[0125] Geometric model for extracting the target state of components from a digital twin model And pre-determine N key feature points on its surface. , ;

[0126] In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ;

[0127] Calculate the deviation vector from the key feature point to the matching point. ;

[0128] The vector deviation of the component is obtained by averaging all N deviation vectors. .

[0129] Furthermore, in the construction plan generation module, the risk coefficient of the component is calculated based on the magnitude and direction of the vector deviation, including:

[0130] Set the allowable deviation threshold for the component ;

[0131] Calculate vector deviation modulus and the absolute value of its vertical component. ;

[0132] The risk coefficient is calculated using the following formula. : ;

[0133] in, The weighting coefficients for the deviation modulus. This represents the weighting coefficient for the influence of the vertical component.

Claims

1. A digital-twin-based power project life cycle management method, characterized in that, The method comprises the following steps: S1, obtaining a construction plan which is decomposed into task units with time attribute and resource demand attribute, and a digital twin model which is divided into components with spatial location attribute and engineering attribute; S2, if there is a time intersection between the planned time window of the task unit and the scheduled construction period of the component, there is a spatial overlap between the work space of the task unit and the spatial location of the component, and the construction object of the task unit is the component in the preset process flow, an association relationship between the task unit and the component is established, so that the construction plan and the digital twin model can be associated in time, space and process; S3, Collect the on-site physical state information of the component, and calculate the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints. This includes: scanning the on-site physical entity of the component using a 3D scanning device to obtain registered 3D point cloud data. Geometric model for extracting the target state of components from a digital twin model. And pre-determine N key feature points on its surface. , In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ; Calculate the deviation vector from the key feature point to the matching point. The vector deviation of the component is obtained by averaging all N deviation vectors. Based on the magnitude and direction of the vector deviation, the risk coefficient of the component is calculated, including setting an allowable deviation threshold for the component. ; Calculate vector deviation modulus and the absolute value of its vertical component. The risk coefficient is calculated using the following formula. : ;in, The weighting coefficients for the deviation modulus. The risk coefficient is used to determine the weighting factor of the vertical component; the planned time window of subsequent task units that have a process flow dependency relationship with the component is adjusted based on the risk coefficient to form a dynamic construction plan; based on the dynamic construction plan and the space occupancy model of the task units, the probability of space and resource conflicts between task units in future cycles is predicted. S4, after the task unit is completed, vector deviation, risk coefficient, quality acceptance result and resource consumption data are structured and packaged and hashed, construction process data bound to the component is generated, and the construction process data is stored in the digital twin model as a permanent non-geometric attribute of the component. 2.The digital-twin-based power project whole life cycle management method according to claim 1, characterized in that, In S2, the association relationship between the task unit and the component is established, comprising: planned time window of a task unit with a predetermined construction period of a component are compared, and when the comparison satisfies , it is determined that the time condition is met; wherein, , are respectively a start time and an end time of the planned time window, , are respectively a start time and an end time of the predetermined construction period. Performing intersection operation on the three-dimensional bounding box of the work space of the task unit and the three-dimensional bounding box of the spatial location of the component, and determining that the spatial condition is met when the overlapping volume is greater than a preset threshold; In the preset process flow directed acyclic graph, it is queried whether there is a direct directed edge from the node of the task unit to the node of the component, and if there is, it is determined that the process flow condition is met; When the three conditions are met, it is determined that there is an association relationship between the task unit and the component. 3.The digital-twin-based power project whole life cycle management method according to claim 1, characterized in that, Based on the risk coefficient, the planned time window of the subsequent task unit which has process flow dependence relationship with the component is adjusted, comprising: In the process flow directed acyclic graph, all subsequent task units which have direct dependence relationship with the task unit of the completed component are identified; The planned delay time for each subsequent task element is calculated by the following non-linear function : ; wherein, is the base time delay factor, is the risk sensitivity coefficient, is the risk coefficient; both the scheduled start time and the scheduled end time of the subsequent task unit are shifted backward to form a new scheduled time window.

4. The digital-twin-based power project whole life cycle management method according to any one of claims 1-3, characterized in that, The vector deviation, the risk coefficient, the quality acceptance result and the resource consumption data are structured and packaged and hashed, the construction process data bound to the component is generated, and the construction process data is stored in the digital twin model as a permanent non-geometric attribute of the component, comprising: Create a structured packaging template containing the following fields: component unique identifier, timestamp, vector deviation , risk coefficient , quality acceptance result, resource consumption data, and hash value of the previous construction process data; Serializing the content of the structured packaging template into a string; A hash value of the string is calculated by using a hash algorithm, and the hash value is used as a unique identifier of the current construction process data; The structured packaging template and its identifier are stored in the component attribute database of the digital twin model.

5. A digital-twin-based power project life cycle management system, characterized in that, Comprising: The decomposition module obtains a construction plan which is decomposed into task units with time attribute and resource demand attribute, and a digital twin model which is divided into components with spatial location attribute and engineering attribute; The association module, if there is a time intersection between the planned time window of the task unit and the scheduled construction period of the component, there is a spatial overlap between the work space of the task unit and the spatial location of the component, and the construction object of the task unit is the component in the preset process flow, an association relationship between the task unit and the component is established, so that the construction plan and the digital twin model can be associated in time, space and process; The construction plan generation module collects on-site physical state information of components and calculates the vector deviation between the on-site physical state and the target state of the digital twin model based on geometric and physical constraints. This includes scanning the on-site physical entity of the component using a 3D scanning device to obtain registered 3D point cloud data. Geometric model for extracting the target state of components from a digital twin model. And pre-determine N key feature points on its surface. , In 3D point cloud data Calculate each key feature point The closest matching point in Euclidean distance ; Calculate the deviation vector from the key feature point to the matching point. The vector deviation of the component is obtained by averaging all N deviation vectors. Based on the magnitude and direction of the vector deviation, the risk coefficient of the component is calculated, including setting an allowable deviation threshold for the component. ; Calculate vector deviation modulus and the absolute value of its vertical component. The risk coefficient is calculated using the following formula. : ;in, The weighting coefficients for the deviation modulus. The risk coefficient is used to determine the weighting factor of the vertical component; the planned time window of subsequent task units that have a process flow dependency relationship with the component is adjusted based on the risk coefficient to form a dynamic construction plan; based on the dynamic construction plan and the space occupancy model of the task units, the probability of space and resource conflicts between task units in future cycles is predicted. A data storage module, after the task unit is completed, carries out structured packaging and hash linking on the vector deviation, the risk coefficient, the quality acceptance result and the resource consumption data, generates construction process data bound with the component, and stores the construction process data as permanent non-geometric properties of the component into the digital twin model.

6. The digital-twin-based power project life cycle management system of claim 5, wherein, In the association module, an association relationship between the task unit and the component is established, including: planned time window of a task unit with a predetermined construction period of a component are compared, when the condition is satisfied , the time condition is determined to be met; wherein, , are respectively a start time and an end time of the planned time window, , are respectively a start time and an end time of the predetermined construction period. A space intersection operation is performed on a three-dimensional bounding box of a work space of the task unit and a three-dimensional bounding box of a space position of the component, and when an overlapping volume is greater than a preset threshold, it is determined that the space condition is met; In a preset process flow directed acyclic graph, it is queried whether there is a direct directed edge from a node of the task unit to a node of the component, and if there is, it is determined that the process flow condition is met; When the three conditions are met, it is determined that there is an association relationship between the task unit and the component.

Images