An intelligent information management method and system for the whole life cycle of oil and gas surface engineering

Abstract

This invention discloses an intelligent information management method and system for the entire lifecycle of oil and gas surface engineering, belonging to the field of oil and gas surface engineering technology. This invention constructs a closed-loop management framework with asset status units as the core, sequentially completing asset status unit construction, digital twin modeling and dynamic updating, intelligent decision command generation, and closed-loop feedback correction. The system includes modules for unit construction, twin modeling, intelligent decision-making, and closed-loop correction. It integrates design, construction, and operation lifecycle data through unified asset identification, utilizes digital twins to map project status in real time, and combines intelligent algorithms to conduct progress deviation, remaining service life, and spatial safety risk analysis, continuously optimizing model parameters based on feedback data. This invention solves the problems of data silos, static models, and reliance on manual decision-making in traditional oil and gas engineering management, achieving intelligent, refined, and dynamically self-optimizing full lifecycle management of oil and gas surface engineering.

Description

Technical Field

[0001] This invention belongs to the interdisciplinary field of oil and gas engineering management and intelligent information processing, specifically relating to an intelligent information management method and system for the entire life cycle of oil and gas surface engineering. Background Technology

[0002] Oil and gas surface engineering is a crucial component of oil and gas field development and construction, characterized by large investments, long cycles, involvement of multiple disciplines, and strong systemic nature. Traditional oil and gas surface engineering management models are gradually evolving towards informatization and digitalization. Existing information management systems mainly focus on the following aspects: Geographic Information System (GIS) Applications: Utilizing GIS technology for spatial management and visualization of pipelines and equipment; Engineering project management system: focuses on the management of project schedule, cost, quality, etc.; Professional data management systems: such as SCADA systems for production operation data collection, and ERP systems for asset and financial management; Digital design system: Utilizing BIM / CAD technology for 3D design and digital delivery.

[0003] Although existing technologies have improved the management efficiency of oil and gas surface engineering to some extent, the following significant drawbacks or shortcomings still exist.

[0004] 1. Severe data silos and lack of full lifecycle integration: Data from various stages, including design, construction, and operation, is scattered across different professional systems (such as design software, construction management software, and production operation systems), creating data silos. The existing system lacks a unified, cross-stage data integration platform, making it impossible to achieve seamless data flow throughout the entire lifecycle, from design data to construction progress and operational status. For example, changes to the design model cannot be synchronized to the construction plan and asset ledger in real time.

[0005] 2. Lack of intelligent analysis and prediction capabilities: The main functions of the existing system are data recording, querying, and visualization; For complex engineering management issues, such as early warning of construction progress deviations, prediction of equipment failures, and assessment of pipeline corrosion risks, the system still mainly relies on human experience and lacks intelligent analysis and predictive decision-making capabilities based on big data and artificial intelligence.

[0006] 3. Lack of dynamic digital twin models: Although digital design (BIM / CAD) exists, these models often become static documents after the project is delivered.

[0007] Existing systems cannot dynamically correlate real-time operational data (such as pressure, temperature, and flow rate) with 3D models, and cannot construct a digital twin that reflects the real-time status of the project, resulting in a lack of intuitiveness and predictability in operation and maintenance.

[0008] This invention aims to solve the technical problems of "data silos", "lack of intelligent prediction" and "lack of dynamic digital twins" existing in the prior art. Summary of the Invention

[0009] This invention addresses the problems of data silos, static models, and reliance on manual decision-making in existing oil and gas surface engineering management technologies. It provides an intelligent information management method for the entire lifecycle of oil and gas surface engineering, which achieves standardized integration of data throughout the entire lifecycle of the project, dynamic updates of digital twin models, and enables precise decision-making for construction, operation and maintenance, and safety through intelligent analysis. Furthermore, it continuously optimizes model parameters through closed-loop feedback, thereby improving the intelligence and precision of oil and gas surface engineering management.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following solution: A method for intelligent information management of the entire lifecycle of oil and gas surface engineering includes the following steps: S1. Obtain standardized data for each asset in the design, construction and operation stages of oil and gas surface engineering, assign an asset identifier to each asset, and construct an asset status unit using the standardized data and asset identifiers from each stage. S2. Using asset identifiers as the core, integrate the data of all asset status units to form a full life cycle data lake. Use the full life cycle data lake to build a cross-stage constraint model of design-construction-operation. Use the cross-stage constraint model of design-construction-operation to calculate the cross-stage constraint deviation of the actual state of the asset relative to the design and construction benchmarks. Use the cross-stage constraint deviation combined with the full life cycle data lake to build and dynamically update the digital twin model. S3. Utilize a digital twin model combined with a full lifecycle data lake to perform progress deviation calculations for the construction status evolution of asset status units, predict the remaining useful life of assets evolving to the failure boundary, and assess the spatial topology risk of the construction site. Based on the three calculation results, trigger the corresponding analysis model and generate engineering management decision instructions for construction optimization, fault warning, and safety control. S4. Push the project management decision instruction to the project management execution terminal. The project management execution terminal issues instructions to relevant management personnel. The management personnel submit the instruction execution results to the project management execution terminal. Collect the instruction execution results and asset status change data during the execution process to form feedback data. Use the feedback data to correct the parameters of the asset status unit state evolution equation. At the same time, associate and store the asset status, project management decision instruction and execution results to form a strategy memory. Output the asset status unit with corrected parameters. Use the corrected asset status unit as the new input for the digital twin model building and updating steps in this method. Complete the whole process closed loop and repeatedly execute the above steps to achieve continuous optimization of the asset status unit and digital twin model.

[0011] The core of this invention lies not in simply integrating design, construction, and operation data, nor in introducing artificial intelligence algorithms, but in proposing a full lifecycle management mechanism for oil and gas surface engineering centered on asset state units. Specifically, this invention abstracts engineering assets into asset state units that evolve throughout their lifecycle. By constructing a cross-stage constraint model encompassing design, construction, and operation, it quantifies the deviation between the asset's operational status and historical constraints. Based on the state evolution results of these asset state units, it drives the updating of the digital twin model and engineering management decisions. Unlike existing management systems that primarily rely on data aggregation or status display, this invention achieves proactive calculation of asset state evolution, transforming engineering management from static monitoring to dynamic prediction and proactive intervention based on lifecycle constraint conflicts. This invention aims to provide an intelligent information management method for the entire lifecycle of oil and gas surface engineering, addressing the technical problems of severe data silos, lack of full lifecycle integration, and lack of intelligent analysis and prediction capabilities in existing technologies.

[0012] Preferably, in step S1, when building the asset status unit, the data from the design phase is standardized and organized to form an asset information model, the expression of which is: ; Among them: ID i G is the unique asset identifier for the i-th project asset; i P represents the geometric information of the i-th project asset; i Let i be the set of design attribute parameters for the i-th engineering asset; The relationship expression between construction phase data and the asset information model is as follows: ; Where: D c This refers to the set of construction data collected during the construction phase; T represents the time dimension; and ID represents the set of asset identifiers in the asset information model. The data during the operational phase is organized into time series data, and the expression for this time series data is: ; Where, x i (t k The asset is identified by ID. i The engineering assets at time t k The running state vector.

[0013] Preferably, in step S2, the expression for the integrated full lifecycle data lake is: ; Where: DL represents a full lifecycle data lake; The asset is identified by ID. i The design phase data corresponding to the engineering assets; This indicates the construction phase data corresponding to the project assets; This represents the operational phase data corresponding to the project assets; N represents the total number of assets. The digital twin model that is built and updated, the expression for the digital twin model corresponding to a single project asset is as follows: ; Among them: G i The asset is identified as The geometric model corresponding to the engineering assets; S i (t) represents the operational status parameters of the project asset at time t; P i (t) represents the construction progress parameter of the project asset at time t; R i (t) represents the risk parameter of the project asset at time t; the risk parameter R... i The expression for (t) is: ; That is, given the running state parameter S i Under the condition (t), the asset is identified as ID. i The engineering assets experienced a fault or abnormal event. i The conditional probability.

[0014] Preferably, in step S3, when calculating the progress deviation of the construction status evolution, the expression for the construction progress deviation is: ; Where P actual (t) represents the actual construction completion rate of the project asset at time t, P optimal (t) represents the optimal construction completion rate of the corresponding project asset predicted by the model; when ΔP(t) ≥ the preset schedule deviation threshold θ p At that time, the construction progress optimization is triggered and a construction resource allocation plan is generated; When predicting the remaining useful life of an asset, the remaining useful life is calculated using a Long Short-Term Memory (LSTM) network, expressed as: ; Where h t W represents the hidden state of the project asset at time t. RUL b represents the weight parameters of the fully connected layer. RUL These are bias parameters; and are also based on the hidden state h. t The probability of failure is calculated using the following expression: ; When P fault (t)≥Preset fault warning threshold θ f When the equipment malfunction is detected, a warning is triggered and a predictive maintenance work order is generated.

[0015] Preferably, in step S3, when conducting a spatial topology risk assessment of the construction site, a work risk topology graph G=(V,E) is first constructed, where V is a set of nodes containing the work area, construction equipment, and construction personnel, and E is a set of spatial or work-related edges between nodes; then, node features are aggregated through a graph neural network, and the feature vector expression of node i in the (l+1)th layer is: ; in: The feature vector of node i in the l-th layer includes construction progress, equipment operating status and historical safety risk indicators; N(i) represents the set of neighboring nodes of node i; d i and d j These represent the degrees of node i and node j, respectively, and are used to normalize the aggregation effect; σ represents the activation function, such as ReLU or sigmoid; W (l) It is the learnable weight matrix of the l-th layer of the graph neural network; After feature updates via an L-layer graph neural network, the security risk index expression for node i is: ; Among them, SRI i represents the comprehensive risk evolution index of the asset state unit corresponding to node i under the current spatiotemporal topological constraints; Softmax is the normalized exponential function; Let W represent the node feature vector updated after passing through an L-layer graph neural network; SRI b is the weight parameter; SRI For bias parameters; when SRI i ≥Preset safety threshold θ s When this occurs, a security risk warning is triggered, and the corresponding area or asset is highlighted.

[0016] Preferably, in step S3, the formal expression of the generated engineering management decision instruction is: ; Among them, D k This is the k-th decision instruction; T k The instruction type includes tasks such as construction schedule optimization, equipment maintenance, or safety management; A k For the specific actions to be performed; U k This refers to the corresponding responsible personnel or execution unit.

[0017] Preferably, in step S4, the feedback data expression is formed as follows: ; Among them, D k For the k-th decision instruction, R exec E(t) represents the result of the instruction execution at time t, and E(t) represents the data on the changes in the status of the engineering assets during the execution process at time t.

[0018] Preferably, in step S4, the expression for correcting the parameters of the asset state unit state evolution equation using feedback data is as follows: ; Where, θ (new) For the corrected model parameters, θ (old) Here, η represents the model parameters before correction, L(·) represents the loss function, ΔL(·) is the gradient of the loss function L(θ,F(t)) with respect to the model parameters θ, and F(t) is the feedback data.

[0019] An intelligent information management system for the entire lifecycle of oil and gas surface engineering, used to implement the aforementioned intelligent information management method for the entire lifecycle of oil and gas surface engineering, specifically includes: The unit construction module is used to collect raw data from all stages of the project, standardize and organize the raw data, assign asset identifiers to each project asset, and build and output asset status units. The twin modeling module receives the asset status units output by the unit construction module as input, integrates them to form a full lifecycle data lake, calculates cross-stage constraint deviation, builds and dynamically updates the digital twin model, and then outputs the model. The intelligent decision-making module receives the digital twin model output by the twin modeling module and combines it with the full life cycle data lake as input to perform schedule deviation calculation, remaining useful life prediction, spatial topology risk assessment, and generate and output engineering management decision instructions. The closed-loop correction module receives engineering management decision instructions from the intelligent decision module as input, pushes the instructions to the engineering management execution terminal and completes the issuance with relevant management personnel, collects the instruction execution results and asset status change data to form feedback data, uses the feedback data to correct the asset status unit parameters and form a strategy memory, outputs the asset status unit with corrected parameters and sends it back to the twin modeling module, forming a closed-loop architecture of cyclic optimization.

[0020] Preferably, the closed-loop correction module includes a strategy memory submodule, which is used to associate and store asset status, engineering management decision instructions and execution results. When an asset status similar to the past occurs during engineering operation, the corresponding decision instruction is directly invoked to improve the efficiency of decision output.

[0021] The beneficial effects of this invention are as follows: 1. This invention constructs a standardized data system with asset status units as the core, assigns a unique identifier to each engineering asset, and realizes standardized integration and integrated management of data throughout the entire life cycle of design, construction, and operation. It completely solves the "data silo" problem in the traditional management model and provides a unified and standardized data foundation for subsequent model construction and intelligent analysis. 2. This invention constructs a digital twin model that can be dynamically updated in real time. It uses a full lifecycle data lake as data support and combines cross-stage constraint deviation to achieve real-time synchronization between the model and the actual state of the project. This breaks through the limitations of traditional BIM / CAD static models and realizes "digital mirroring + real-time monitoring" of engineering assets, providing a visual and interactive digital carrier for engineering management. 3. This invention integrates intelligent algorithms such as machine learning and graph neural networks to achieve quantitative analysis and intelligent prediction of construction schedule deviation, remaining asset service life, and spatial topology risks. It replaces traditional manual experience-based decision-making, improves the accuracy and timeliness of engineering management decisions, and effectively reduces safety hazards in engineering construction and operation. 4. This invention constructs a closed-loop management framework for the entire process of "data acquisition - model building - intelligent decision-making - feedback optimization". By correcting the parameters of the asset status unit through feedback data, and sending the corrected parameters back to the digital twin model building step, the model parameters are continuously optimized, making the model prediction results increasingly consistent with the actual state of the project. At the same time, the strategy memory submodule realizes the digital reuse of management experience, further improving the efficiency of project management. 5. The management system of the present invention forms an input-output closed-loop architecture, which is highly matched with the management method. It realizes the automated and intelligent operation of the entire life cycle management of oil and gas surface engineering, greatly improves the level of precision management of oil and gas surface engineering, and has good engineering application value and promotion prospects. Attached Figure Description

[0022] Figure 1 This is a flowchart of an intelligent information management method for the entire life cycle of oil and gas surface engineering according to the present invention.

[0023] Figure 2 This is a diagram of an intelligent information management system for the entire lifecycle of oil and gas surface engineering, as described in this invention. Detailed Implementation

[0024] The present invention will be further described in detail below with reference to the embodiments and accompanying drawings, but the embodiments of the present invention are not limited thereto.

[0025] like Figure 1 As shown, Figure 1 This is a flowchart illustrating an intelligent information management method for the entire lifecycle of oil and gas surface engineering, as described in this invention. The method includes the following steps: S1. Obtain standardized data for each asset in the design, construction and operation stages of oil and gas surface engineering, assign an asset identifier to each asset, and construct an asset status unit using the standardized data and asset identifiers from each stage. S2. Using asset identifiers as the core, integrate the data of all asset status units to form a full life cycle data lake. Use the full life cycle data lake to build a cross-stage constraint model of design-construction-operation. Use the cross-stage constraint model of design-construction-operation to calculate the cross-stage constraint deviation of the actual state of the asset relative to the design and construction benchmarks. Use the cross-stage constraint deviation combined with the full life cycle data lake to build and dynamically update the digital twin model. S3. Utilize a digital twin model combined with a full lifecycle data lake to perform progress deviation calculations for the construction status evolution of asset status units, predict the remaining useful life of assets evolving to the failure boundary, and assess the spatial topology risk of the construction site. Based on the three calculation results, trigger the corresponding analysis model and generate engineering management decision instructions for construction optimization, fault warning, and safety control. S4. Push the project management decision instruction to the project management execution terminal. The project management execution terminal issues instructions to relevant management personnel. The management personnel submit the instruction execution results to the project management execution terminal. Collect the instruction execution results and asset status change data during the execution process to form feedback data. Use the feedback data to correct the parameters of the asset status unit state evolution equation. At the same time, associate and store the asset status, project management decision instruction and execution results to form a strategy memory. Output the asset status unit with corrected parameters. Use the corrected asset status unit as the new input for the digital twin model building and updating steps in this method. Complete the whole process closed loop and repeatedly execute the above steps to achieve continuous optimization of the asset status unit and digital twin model.

[0026] Preferably, in step S1, when building the asset status unit, the data from the design phase is standardized and organized to form an asset information model, the expression of which is: ; Among them: ID i G is the unique asset identifier for the i-th project asset; i P represents the geometric information of the i-th project asset; i Let i be the set of design attribute parameters for the i-th engineering asset; The relationship expression between construction phase data and the asset information model is as follows: ; Where: D c This refers to the set of construction data collected during the construction phase; T represents the time dimension; and ID represents the set of asset identifiers in the asset information model. The data during the operational phase is organized into time series data, and the expression for this time series data is: ; Where, x i (t k The asset is identified by ID. i The engineering assets at time t k The running state vector.

[0027] Preferably, in step S2, the expression for the integrated full lifecycle data lake is: ; Where: DL represents a full lifecycle data lake; The asset is identified by ID. i The design phase data corresponding to the engineering assets; This indicates the construction phase data corresponding to the project assets; This represents the operational phase data corresponding to the project assets; N represents the total number of assets. The digital twin model that is built and updated, the expression for the digital twin model corresponding to a single project asset is as follows: ; Among them: G i The asset is identified as The geometric model corresponding to the engineering assets; S i (t) represents the operational status parameters of the project asset at time t; P i (t) represents the construction progress parameter of the project asset at time t; R i (t) represents the risk parameter of the project asset at time t; the risk parameter R... i The expression for (t) is: ; That is, given the running state parameter S i Under the condition (t), the asset is identified as ID. i The engineering assets experienced a fault or abnormal event. i The conditional probability.

[0028] Preferably, in step S3, when calculating the progress deviation of the construction status evolution, the expression for the construction progress deviation is: ; Where P actual (t) represents the actual construction completion rate of the project asset at time t, P optimal (t) represents the optimal construction completion rate of the corresponding project asset predicted by the model; when ΔP(t) ≥ the preset schedule deviation threshold θ p At that time, the construction progress optimization is triggered and a construction resource allocation plan is generated; When predicting the remaining useful life of an asset, the remaining useful life is calculated using a Long Short-Term Memory (LSTM) network, expressed as: ; Where h t W represents the hidden state of the project asset at time t. RUL b represents the weight parameters of the fully connected layer. RUL These are bias parameters; and are also based on the hidden state h. t The probability of failure is calculated using the following expression: ; When P fault (t)≥Preset fault warning threshold θ f When the equipment malfunction is detected, a warning is triggered and a predictive maintenance work order is generated.

[0029] Preferably, in step S3, when conducting a spatial topology risk assessment of the construction site, a work risk topology graph G=(V,E) is first constructed, where V is a set of nodes containing the work area, construction equipment, and construction personnel, and E is a set of spatial or work-related edges between nodes; then, node features are aggregated through a graph neural network, and the feature vector expression of node i in the (l+1)th layer is: ; in: The feature vector of node i in the l-th layer includes construction progress, equipment operating status and historical safety risk indicators; N(i) represents the set of neighboring nodes of node i; d i and d j These represent the degrees of node i and node j, respectively, and are used to normalize the aggregation effect; σ represents the activation function, such as ReLU or sigmoid; W (l) It is the learnable weight matrix of the l-th layer of the graph neural network; After feature updates via an L-layer graph neural network, the security risk index expression for node i is: ; Among them, SRI i represents the comprehensive risk evolution index of the asset state unit corresponding to node i under the current spatiotemporal topological constraints; Softmax is the normalized exponential function; Let W represent the node feature vector updated after passing through an L-layer graph neural network; SRI b is the weight parameter; SRI For bias parameters; when SRI i ≥Preset safety threshold θ s When this occurs, a security risk warning is triggered, and the corresponding area or asset is highlighted.

[0030] Preferably, in step S3, the formal expression of the generated engineering management decision instruction is: ; Among them, D k This is the k-th decision instruction; T k The instruction type includes tasks such as construction schedule optimization, equipment maintenance, or safety management; A k For the specific actions to be performed; U k This refers to the corresponding responsible personnel or execution unit.

[0031] Preferably, in step S4, the feedback data expression is formed as follows: ; Among them, D k For the k-th decision instruction, R exec E(t) represents the result of the instruction execution at time t, and E(t) represents the data on the changes in the status of the engineering assets during the execution process at time t.

[0032] Preferably, in step S4, the expression for correcting the parameters of the asset state unit state evolution equation using feedback data is as follows: ; Where, θ (new) For the corrected model parameters, θ (old) Here, η represents the model parameters before correction, L(·) represents the loss function, ΔL(·) is the gradient of the loss function L(θ,F(t)) with respect to the model parameters θ, and F(t) is the feedback data.

[0033] like Figure 2 As shown, Figure 2This is a diagram of an intelligent information management system for the entire lifecycle of oil and gas surface engineering, as described in this invention. The system includes: The unit construction module is used to collect raw data from all stages of the project, standardize and organize the raw data, assign asset identifiers to each project asset, and build and output asset status units. The twin modeling module receives the asset status units output by the unit construction module as input, integrates them to form a full lifecycle data lake, calculates cross-stage constraint deviation, builds and dynamically updates the digital twin model, and then outputs the model. The intelligent decision-making module receives the digital twin model output by the twin modeling module and combines it with the full life cycle data lake as input to perform schedule deviation calculation, remaining useful life prediction, spatial topology risk assessment, and generate and output engineering management decision instructions. The closed-loop correction module receives engineering management decision instructions from the intelligent decision module as input, pushes the instructions to the engineering management execution terminal and completes the issuance with relevant management personnel, collects the instruction execution results and asset status change data to form feedback data, uses the feedback data to correct the asset status unit parameters and form a strategy memory, outputs the asset status unit with corrected parameters and sends it back to the twin modeling module, forming a closed-loop architecture of cyclic optimization.

[0034] Preferably, the closed-loop correction module includes a strategy memory submodule, which is used to associate and store asset status, engineering management decision instructions and execution results. When an asset status similar to the past occurs during engineering operation, the corresponding decision instruction is directly invoked to improve the efficiency of decision output.

[0035] To more clearly illustrate the operational logic of this invention, the following uses a core piece of equipment in an oil and gas gathering and transportation station, the high-pressure external transmission compressor (asset identifier: ASU(ID)), as an example. C005 Taking this system as an example, we will explain the complete workflow of this system from data acquisition and evolutionary calculation to closed-loop feedback.

[0036] Phase 1: Construction of Asset Status Units During the construction phase, the system established a unique asset status unit for this compressor and completed the injection of the following "genetic" data: Design baseline injection: Extract the maximum design speed (12,000 RPM), design vibration limit (5.0 mm / s), and theoretical design life (20 years) of the equipment from the BIM model provided by the design institute. Construction Quality Mark: During the installation phase, the system detected a slight deviation in the levelness of the grouting at the equipment's base through construction logs (although it met acceptance standards, it was slightly below the acceptable range). The system will record this as a "Construction Quality Parameter (P)". quality =0.85) "Permanently associated with ASU(ID)" C005), which serves as the initial boundary condition for subsequent state evolution.

[0037] Phase Two: Cross-Phase Constraints and Deviation Perception Three months after the equipment was put into operation, the SCADA system collected a real-time vibration value of 3.5 mm / s. Traditional systems judge that 3.5 mm / s is less than the design limit of 5.0 mm / s, and therefore consider it "normal" with no alarm. The system determines that: by retrieving the design benchmarks and construction records from S1, it performs cross-stage constraint calculations. Model calculations indicate that for a new piece of equipment that has only been in operation for three months, the theoretically expected vibration under the current load should be 1.2 mm / s. Deviation quantification: The system calculated the cross-stage constraint deviation to be +2.3 mm / s. Combined with the "Construction Levelness Deviation" record, the ASU status was marked as "Constraint Conflict Status," indicating an abnormal performance degradation caused by early construction defects.

[0038] Phase Three: State Evolution Deduction and Risk Assessment Based on the aforementioned deviation, the system triggers intelligent evolutionary calculation: failure boundary estimation (RUL calculation): using a time-series evolution model, the "current deviation" and "historical vibration upward trend" are used as input features. The estimation results show that without intervention, ASU(ID) C005 The vibration index will exceed the failure threshold of 5.0 mm / s after 28 days. Based on this, the system outputs: Remaining Time Without Failure (RUL) = 28 days; Spatial topology risk assessment: The system detected through real-time positioning that a maintenance worker was operating within a 2-meter radius of the compressor. The system constructed a spatial risk topology map and calculated that the Spatial Conflict Risk Index (SRI) between the worker node and the high-risk equipment node reached 0.85 (high-risk level), indicating that although the equipment has not yet failed, under this evolutionary trend, the prolonged presence of the worker poses an extremely high safety hazard.

[0039] Phase Four: Closed-Loop Decision Making and Model Evolution Based on the above calculation results, the system automatically performs the following closed-loop operation: Proactive intervention: Push a safety warning to the handheld terminal of on-site maintenance personnel: "Nearby equipment is evolving abnormally, please evacuate quickly"; At the same time, generate a "predictive maintenance work order" to the control center, suggesting that a shutdown and calibration of the base be arranged within 7 days; Verification: After executing the work order, the maintenance team confirmed that the base bolts were indeed slightly loose. After recalibration, the vibration value dropped back to 1.1 mm / s after power-on. Model self-correction: The system stores the evolution path of "construction deviation data + early minor vibration deviation = base loosening fault" in the strategy memory and automatically corrects the weight parameters of the evolution equation. In subsequent operation, once similar characteristics appear, the system will issue an early warning with higher confidence, realizing the self-evolution of management logic.

[0040] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Based on the technical essence of the present invention, any simple modifications, equivalent substitutions, and improvements made to the above embodiments within the spirit and principles of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for intelligent information management of the entire lifecycle of oil and gas surface engineering, characterized in that, Includes the following steps: S1. Obtain standardized data for each asset in the design, construction and operation stages of oil and gas surface engineering, assign an asset identifier to each asset, and construct an asset status unit using the standardized data and asset identifiers from each stage. S2. Using asset identifiers as the core, integrate the data of all asset status units to form a full life cycle data lake. Use the full life cycle data lake to build a cross-stage constraint model of design-construction-operation. Use the cross-stage constraint model of design-construction-operation to calculate the cross-stage constraint deviation of the actual state of the asset relative to the design and construction benchmarks. Use the cross-stage constraint deviation combined with the full life cycle data lake to build and dynamically update the digital twin model. S3. Utilize a digital twin model combined with a full lifecycle data lake to perform progress deviation calculations for the construction status evolution of asset status units, predict the remaining useful life of assets evolving to the failure boundary, and assess the spatial topology risk of the construction site. Based on the three calculation results, trigger the corresponding analysis model and generate engineering management decision instructions for construction optimization, fault warning, and safety control. S4. Push the project management decision instruction to the project management execution terminal. The project management execution terminal issues instructions to relevant management personnel. The management personnel submit the instruction execution results to the project management execution terminal. Collect the instruction execution results and asset status change data during the execution process to form feedback data. Use the feedback data to correct the parameters of the asset status unit state evolution equation. At the same time, associate and store the asset status, project management decision instruction and execution results to form a strategy memory. Output the asset status unit with corrected parameters. Use the corrected asset status unit as the new input for the digital twin model building and updating steps in this method. Complete the whole process closed loop and repeatedly execute the above steps to achieve continuous optimization of the asset status unit and digital twin model.

2. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S1, when building the asset status unit, the data from the design phase is standardized and organized to form an asset information model. The expression of this model is: ； Among them: ID i G is the unique asset identifier for the i-th project asset; i P represents the geometric information of the i-th project asset; i Let i be the set of design attribute parameters for the i-th engineering asset; The relationship expression between construction phase data and the asset information model is as follows: ； Where: D c This refers to the set of construction data collected during the construction phase; T represents the time dimension; and ID represents the set of asset identifiers in the asset information model. The data during the operational phase is organized into time series data, and the expression for this time series data is: ； Where, x i (t k The asset is identified by ID. i The engineering assets at time t k The running state vector.

3. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S2, the expression for the integrated full lifecycle data lake is: ； Where: DL represents a full lifecycle data lake; The asset is identified by ID. i The design phase data corresponding to the engineering assets; This indicates the construction phase data corresponding to the project assets; This represents the operational phase data corresponding to the project assets; N represents the total number of assets. The digital twin model that is built and updated, the expression for the digital twin model corresponding to a single project asset is as follows: ； Among them: G i The asset is identified as The geometric model corresponding to the engineering assets; S i (t) represents the operational status parameters of the project asset at time t; P i (t) represents the construction progress parameter of the project asset at time t; R i (t) represents the risk parameter of the project asset at time t; the risk parameter R... i The expression for (t) is: ； That is, given the running state parameter S i Under the condition (t), the asset is identified as ID. i The engineering assets experienced a fault or abnormal event. i The conditional probability.

4. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S3, when calculating the schedule deviation for the construction status evolution, the expression for the construction schedule deviation is: ； Where P actual (t) represents the actual construction completion rate of the project asset at time t, P optimal (t) represents the optimal construction completion rate of the corresponding project asset predicted by the model; when ΔP(t) ≥ the preset schedule deviation threshold θ p At that time, the construction progress optimization is triggered and a construction resource allocation plan is generated; When predicting the remaining useful life of an asset, the remaining useful life is calculated using a Long Short-Term Memory (LSTM) network, expressed as: ； Where h t W represents the hidden state of the project asset at time t. RUL b represents the weight parameters of the fully connected layer. RUL These are bias parameters; and are also based on the hidden state h. t The probability of failure is calculated using the following expression: ； When P fault (t)≥Preset fault warning threshold θ f When the equipment malfunction is detected, a warning is triggered and a predictive maintenance work order is generated.

5. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S3, when conducting a spatial topology risk assessment of the construction site, the first step is to construct an operational risk topology graph G=(V, E), where V is a set of nodes containing the operational area, construction equipment, and construction personnel, and E is a set of spatial or operational connection edges between nodes. Then, node features are aggregated using a graph neural network. The feature vector expression for node i at layer l+1 is: ； in: The feature vector of node i in the l-th layer includes construction progress, equipment operating status and historical safety risk indicators; N(i) represents the set of neighboring nodes of node i; d i and d j These represent the degrees of node i and node j, respectively, and are used to normalize the aggregation effect; σ represents the activation function; W (l) It is the learnable weight matrix of the l-th layer of the graph neural network; After feature updates via an L-layer graph neural network, the security risk index expression for node i is: ； Among them, SRI i represents the comprehensive risk evolution index of the asset state unit corresponding to node i under the current spatiotemporal topological constraints; Softmax is the normalized exponential function; Let W represent the node feature vector updated after passing through an L-layer graph neural network; SRI b is the weight parameter; SRI For bias parameters; when SRI i ≥Preset safety threshold θ s When this occurs, a security risk warning is triggered, and the corresponding area or asset is highlighted.

6. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S3, the formal expression of the generated engineering management decision instruction is: ； Among them, D k This is the k-th decision instruction; T k The instruction type includes tasks such as construction schedule optimization, equipment maintenance, or safety management; A k For the specific actions to be performed; U k This refers to the corresponding responsible personnel or execution unit.

7. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S4, the feedback data expression is formed as follows: ； Among them, D k For the k-th decision instruction, R exec E(t) represents the result of the instruction execution at time t, and E(t) represents the data on the changes in the status of the engineering assets during the execution process at time t.

8. The intelligent information management method for the entire life cycle of oil and gas surface engineering according to claim 1, characterized in that, In step S4, the expression for correcting the parameters of the asset state unit state evolution equation using feedback data is as follows: ； Where, θ (new) For the corrected model parameters, θ (old) Here, η represents the model parameters before correction, L(·) represents the loss function, ΔL(·) is the gradient of the loss function L(θ,F(t)) with respect to the model parameters θ, and F(t) is the feedback data.

9. A smart information management system for the entire lifecycle of oil and gas surface engineering, characterized in that, The method for implementing the intelligent information management of the entire life cycle of oil and gas surface engineering as described in any one of claims 1-8 specifically includes: The unit construction module is used to collect raw data from all stages of the project, standardize and organize the raw data, assign asset identifiers to each project asset, and build and output asset status units. The twin modeling module receives the asset status units output by the unit construction module as input, integrates them to form a full lifecycle data lake, calculates cross-stage constraint deviation, builds and dynamically updates the digital twin model, and then outputs the model. The intelligent decision-making module receives the digital twin model output by the twin modeling module and combines it with the full life cycle data lake as input to perform schedule deviation calculation, remaining useful life prediction, spatial topology risk assessment, and generate and output engineering management decision instructions. The closed-loop correction module receives engineering management decision instructions from the intelligent decision module as input, pushes the instructions to the engineering management execution terminal and completes the issuance with relevant management personnel, collects the instruction execution results and asset status change data to form feedback data, uses the feedback data to correct the asset status unit parameters and form a strategy memory, outputs the asset status unit with corrected parameters and sends it back to the twin modeling module, forming a closed-loop architecture of cyclic optimization.

10. The intelligent information management system for the entire lifecycle of oil and gas surface engineering according to claim 9, characterized in that, The closed-loop correction module includes a strategy memory submodule, which is used to associate and store asset status, engineering management decision instructions and execution results. When an asset status similar to the past occurs during engineering operation, the corresponding decision instructions are directly invoked to improve the efficiency of decision output.

Images