A method and system for optimizing visual maintenance decision of a flexible driven subsea production system

Abstract

The application discloses a kind of underwater production system maintenance decision optimization method and system driven by toughness, belong to underwater production system maintenance decision field, this method includes: predicting system remaining useful life, based on Weibull distribution modeling component natural degradation performance, combined with Poisson-normal impact model Quantification random impact degradation, superimposed calculation system total degradation performance and predict remaining useful life;Predict oil and gas production, adopt I generalized mathematical model to fit historical production data;Real-time toughness evaluation, fusion technical index and economic index, through performance retention function and yield stability function to calculate comprehensive toughness value;Maintenance decision optimization, with real-time toughness value as input variable, establish dual-objective decision model of maximum system toughness and minimum maintenance cost, adopt multi-objective optimization algorithm to solve optimal maintenance trigger threshold, the application is driven by toughness index Maintenance decision, effectively improve the reliability and economy of underwater production system.

Description

Technical Field

[0001] This invention relates to the field of underwater production system maintenance technology, and proposes a resilience-driven method and system for optimizing condition-based maintenance decisions for underwater production systems. Background Technology

[0002] Subsea production systems, including subsea wellheads, subsea manifolds, and subsea control systems, operate under harsh environments such as high pressure, low temperature, and corrosion, making them prone to degradation and failure. Current maintenance strategies primarily include scheduled maintenance, fault-based maintenance, and condition-based maintenance. Condition-based maintenance makes decisions based on the real-time status of the equipment, enabling the rational planning of maintenance time, reducing unnecessary repairs, lowering costs, and improving system availability.

[0003] CN113378480B discloses a condition-based maintenance method and system for subsea production trees based on remaining useful life prediction. This method predicts the remaining useful life of components by establishing component degradation impact models, incomplete maintenance models, and remaining useful life prediction models, considering degradation characteristics and the effects of incomplete maintenance. CN112883549B proposes a method for establishing a condition-based maintenance model that considers the impact of random impacts. Based on the Wiener process, it constructs a staged degradation model under imperfect maintenance, incorporating the influence of maintenance and random impacts on the degradation amount and rate, and obtains the analytical solution of the remaining useful life probability density function based on the first-arrival time concept. CN118096121A discloses an intelligent maintenance decision-making method and system for subsea production control systems. The system employs steps such as nonlinear Wiener process model fixed parameter estimation, Kalman filter drift parameter update, and remaining service life prediction, and optimizes maintenance decisions through multi-agent deep reinforcement learning to adapt to complex operating conditions. CN118536433B discloses an intelligent operation and maintenance and production optimization method and system for marine oil subsea production systems, which calculates static and dynamic health indices through sensor data and formulates the optimal maintenance strategy with the goals of maximizing resilience and minimizing maintenance costs. CN119919124A proposes a resilience-driven condition-based maintenance decision optimization method, which includes job safety analysis, maintenance operation risk assessment, initial optimization under known risk conditions, and resilience-driven secondary optimization, implementing secondary decision-making from the perspective of risk and stability.

[0004] As can be seen from the above patent literature, traditional condition-based maintenance methods are based on single technical indicators such as reliability, risk, and remaining service life, combined with maintenance costs for multi-objective optimization. However, these technical and economic indicators are independent of each other, failing to consider their correlation and ignoring the intrinsic link between system performance degradation and production changes. This makes it impossible to comprehensively assess the combined impact of maintenance decisions on system performance and economic benefits. Furthermore, existing degradation modeling only considers either natural degradation or random shocks, making it difficult to accurately describe the actual degradation patterns in complex marine environments. Simultaneously, the evaluation of maintenance effectiveness is insufficient, making it difficult to quantify the impact of incomplete maintenance on remaining service life, thus affecting the accuracy and effectiveness of decision-making. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a resilience-driven condition-based maintenance decision optimization method and system for underwater production systems. This invention solves the following technical problems: In traditional condition-based maintenance methods for underwater production systems, technical and economic indicators are independent of each other, and the correlation between the two is not considered. This invention aims to achieve synergistic optimization of the shock resistance and recovery capabilities of underwater production systems, extend the service life of the system, and save on the total life-cycle maintenance costs.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a resilience-driven condition-based maintenance decision optimization method for underwater production systems, comprising: S1, predicting the remaining service life of the system: modeling the natural degradation performance of components based on the Weibull distribution, quantifying the random impact degradation amount by combining the Poisson-normal impact model, superimposing and calculating the total degradation performance of the system and predicting the remaining service life; S2, predicting oil and gas production: fitting historical production data with a type I generalized mathematical model to establish a production prediction function; S3, real-time resilience assessment: integrating technical and economic indicators, calculating the comprehensive resilience value through the performance retention function and the production stability function; S4, maintenance decision optimization: using the real-time resilience value as an input variable, establishing a dual-objective decision model that maximizes system resilience and minimizes maintenance costs, and using a multi-objective optimization algorithm to solve for the optimal maintenance trigger threshold.

[0007] Furthermore, step S1 includes: calculating the component's natural degradation performance using the Weibull cumulative distribution function. In the formula, F j ( t ) indicates the first j The performance of a component at time t is generally the component's probability of failure, where t represents the time variable. η j Let be the scale parameter of the j-th component. β j Let be the shape parameter of the j-th component; the system's natural degradation performance is the weighted sum of the degradation performances of each component: In the formula, DN ( t ) represents the system's natural degradation performance. W j ( t ) indicates the first j The weight coefficients of each component, where N represents the total number of components; the weight coefficients are determined by normalizing the component degradation sensitivity.

[0008] Furthermore, in the Poisson-normal impact model: the number of impact events follows a homogeneous Poisson process. In the formula, N ( t ) indicates the time interval [0, t The total number of events that occurred within the period. c Represents random variables N ( t Possible values, λ The intensity parameter remains constant throughout the entire time interval; the degradation amount of a single impact follows a normal distribution.

[0009] Furthermore, the type I generalized mathematical model in step S2 is: In the formula, Indicates annual oil and gas production. a, b, c These are model constants. m The classification factors for the model are defined as follows: parameters a, b, c, and m are determined by fitting historical production data using the least squares method.

[0010] Furthermore, the performance retention function in step S3 is: In the formula, RUL ( t )express t The real-time remaining service life of an underwater production system is the remaining service life under the combined effects of natural degradation and external impacts. rul ( t )express t The expected remaining service life of the underwater production system at any given time, i.e., the remaining service life under conditions of only natural degradation. t 0 is the starting point of the performance resilience assessment cycle, which is generally 0; the production stability function is: In the formula, R pro ( t )express t Production stability at any given time Indicates annual oil and gas production. This is the actual output after adjustment for performance impact factors.

[0011] Furthermore, step S4 uses a service life regression factor to describe the effect of incomplete maintenance: pm = 1 - (cm / CR)(T-RUL(t)) / MRL In the formula, pm This represents the system's maintenance life reduction factor. cm This indicates the planned maintenance costs, depending on the situation. CR This indicates the replacement cost. T This indicates the total lifespan of the underwater production system. MRL This represents the system's average remaining lifespan; the remaining lifespan after maintenance is updated as: RUL(t+1) = T - (T - RUL(t)) * pm; where, RUL(t +1 ) express t +1 indicates the real-time remaining service life of the underwater production system after the completion of any maintenance activity.

[0012] Furthermore, the repair costs include both incomplete repair costs and complete repair costs: In the formula, C ALL The total cost of condition-based maintenance and replacement activities for underwater production systems. M j For components j The number of repairs depending on the situation. R j For components j Number of replacements t mj For components j The time required to perform a condition-based maintenance. t rj For components j The time required to carry out a change activity. C mj For components j The unit time cost of performing condition-based maintenance. C rj For components j The unit time cost of replacement; among which Mj , Rj These represent the number of incomplete repairs and the number of replacements for component j, respectively.

[0013] Furthermore, the average degradation performance of a component over its entire lifecycle is expressed as... In the formula, This represents the average degradation performance of a component over its entire lifecycle. T This indicates the total lifespan of the underwater production system. F j ( t ) indicates the first j Performance of component j at time t; calculate degradation sensitivity of component j. In the formula, NThis indicates the number of components in an underwater production system.

[0014] Furthermore, the actual output Corrected by performance impact factor In the formula, rul ( t ) indicates that the system at time t t The remaining useful life, Φ(t) is the performance impact function, which reflects the proportion of the actual degradation state on the production capacity.

[0015] Furthermore, after the same component has undergone three incomplete repairs, a full repair is performed.

[0016] This invention also includes a resilience-driven condition-based maintenance decision optimization system for underwater production systems, comprising: a remaining useful life prediction module: including a system natural degradation performance prediction unit and a system performance prediction unit under external shock influence; the system natural degradation performance prediction unit includes a component natural degradation performance prediction subunit and a component weight evaluation subunit, used to calculate weighted natural degradation performance; the system performance prediction unit under external shock influence includes an external shock arrival time prediction subunit and an external shock intensity prediction subunit, used to establish a Poisson-normal shock model; and an oil and gas production prediction module: connected to the remaining useful life prediction module via a cable, based on... The system employs a generalized mathematical model to predict production; a real-time resilience assessment module, comprising a technical resilience assessment unit and an economic resilience assessment unit, connected via cables to the remaining useful life prediction module and the oil and gas production prediction module, respectively, to calculate a comprehensive resilience value by integrating performance retention and production stability; a condition-based maintenance decision module, connected via cables to the real-time resilience assessment module, to establish a dual-objective decision model; and a maintenance decision output module, connected via cables to the condition-based maintenance decision module, to output the optimal maintenance trigger threshold to the main control station. Each module transmits data via cables, and the system's natural degradation performance prediction unit uses a Weibull distribution to model component degradation.

[0017] Furthermore, the condition-based maintenance decision module includes: a service life rollback factor calculation unit for calculating the remaining useful life update value after incomplete maintenance; and a maintenance cost calculation unit for calculating the total cost including both incomplete and complete maintenance costs; wherein the service life rollback factor calculation unit executes the formula: pm=1-(cm / CR) (T-RUL(t)) / MRL In the formula, RUL(t) represents the real-time remaining service life of the underwater production system after the conditional maintenance activity ends at time t; the maintenance cost calculation unit executes the formula. In the formula, C ALL The total cost of condition-based maintenance and replacement activities for underwater production systems, where Mj is the number of condition-based maintenance operations for component j; R j t represents the number of times component j is replaced; mjThe time required to perform a condition-based maintenance on component j; t rj The time required to perform a replacement activity for component j; C mj C represents the unit time cost of performing condition-based maintenance on component j. rj The unit time cost for replacing component j.

[0018] The beneficial effects of this invention are as follows: By inherently coupling performance retention and production stability, the technical and economic indicators in the resilience evaluation process are dynamically correlated, improving the accuracy of resilience assessment. Using real-time resilience as the independent variable for condition-based maintenance decisions, the shock resistance and recovery capabilities of underwater production systems are synergistically optimized. This extends the service life of underwater production systems while saving maintenance costs throughout their entire service life. Compared with traditional condition-based maintenance methods based on a single technical indicator, this invention establishes a correlation between technical and economic indicators, making maintenance decisions more comprehensive and accurate. It can ensure system reliability while also considering economic benefits, providing a more scientific decision-making basis for the maintenance and management of underwater production systems. Attached Figure Description

[0019] Figure 1 This is a flowchart of a resilience-driven approach to optimizing condition-based maintenance decisions for underwater production systems.

[0020] Figure 2 This is a schematic diagram illustrating the resilience assessment of integrated technology and economic indicators;

[0021] Figure 3 This is a schematic diagram of an underwater production system;

[0022] Figure 4 This is a schematic diagram of a resilience-driven underwater production system for condition-based maintenance decision optimization.

[0023] In the diagram, 101 is the hydraulic power unit, 102 is the main control station, 103 is the power unit, 104 is the surface umbilical cable terminal assembly, 105 is the umbilical cable, 106 is the underwater umbilical cable terminal assembly, 107 is the underwater control module, and 108 is the Christmas tree; 201 is the remaining service life prediction module, 202 is the system natural degradation performance prediction unit, 203 is the component natural degradation performance prediction subunit, 204 is the component weight evaluation subunit, 205 is the system performance prediction unit under external impact, 206 is the external impact arrival time prediction subunit, 207 is the external impact intensity prediction subunit, 208 is the oil and gas production prediction module, 209 is the real-time toughness assessment module, 210 is the technical toughness assessment unit, 211 is the economic toughness assessment unit, 212 is the service life regression factor calculation unit, 213 is the maintenance cost calculation unit, 214 is the condition-based maintenance decision module, and 215 is the maintenance decision output module. Detailed Implementation

[0024] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0025] Example 1

[0026] like Figure 1 As shown, a resilience-driven condition-based maintenance decision optimization method for subsea production systems includes four steps: prediction of the remaining service life of the subsea production system, prediction of oil and gas production of the subsea production system, resilience assessment integrating technical and economic indicators, and the resilience-driven condition-based maintenance decision optimization method.

[0027] S1: Prediction of Remaining Service Life of Underwater Production Systems

[0028] The Weibull distribution has strong adaptability and clear physical interpretation, making it suitable for describing physical degradation phenomena such as wear, fatigue, and corrosion. It is widely used in reliability analysis and life modeling of engineering equipment.

[0029] The Weibull distribution is used to model the performance degradation process of underwater production systems under natural degradation conditions. The probability density function of the Weibull distribution can be expressed as: ;in, t Indicates time, η Let be the scale parameter of the Weibull distribution. β Let be the shape parameter of the Weibull distribution.

[0030] Based on the probability density function of the Weibull distribution, the corresponding cumulative distribution function can be obtained, and the component's performance at a certain time can be calculated. t The probability of failure occurring.

[0031] The performance of a component at its current moment is represented by its failure probability. In an underwater production system, the [missing information - likely a specific component or parameter] is... j The performance degradation of each component can be calculated as follows: The scale and shape parameters of the components can be obtained from historical data using the maximum likelihood method.

[0032] A subsea production tree system can be viewed as a series system consisting of several independent components, meaning that the failure of any component will lead to the failure of the entire subsea production system.

[0033] To calculate the cumulative degradation of the system, it is necessary to determine the weight coefficients of each component in the system.

[0034] Calculate the sensitivity of each component to the degradation of system performance, normalize the sensitivity, and determine the weight ratio of each component.

[0035] No. j The average degradation performance of a component over its entire lifecycle can be expressed as: ;No. j Sensitivity of individual components to system degradation performance S j for ;in, N This indicates the number of components in an underwater production system.

[0036] No. j The weighting of each component in relation to system degradation performance W j for Based on the degradation performance and weighting coefficient of each component, in t The cumulative degradation performance of the system due to natural degradation at time step (i.e., time step) can be calculated as follows: In addition to natural performance degradation under normal conditions, underwater production systems can also be affected by random shocks such as seabed geological activity.

[0037] External shocks occur randomly, and their impact on the system is uncertain.

[0038] Homogeneous Poisson processes can be used to simulate the frequency of external shock events and to simulate the magnitude of the shock using a normal distribution. A Poisson-normal shock model can be established, and the homogeneous Poisson process can be expressed as follows: ;in, N ( t ) indicates the time interval [0, t The total number of events that occurred within the period. c Represents random variables N ( t Possible values, λ The intensity parameter remains constant throughout the entire time interval.

[0039] It is approximated that the impact amount caused by each impact follows a normal distribution, and the cumulative degradation performance of the system caused by the external impact event is: ;in, Y k For the first k The system degradation performance caused by the secondary impact event follows a mean of [value missing]. μ Y The standard deviation is σ Y It follows a normal distribution.

[0040] The total degradation performance of the system is the sum of the cumulative degradation performance caused by natural degradation and the cumulative degradation performance caused by external shock events.

[0041] When the system's performance degrades to 0, the system reaches a failure state. The remaining service life of the system is represented by the time interval at which the system's performance first reaches 0, which can be calculated as: RUL=inf{t:D(t)≤0|D(0)>0}.

[0042] S2: Oil and Gas Production Forecast for Subsea Production Systems

[0043] A type I generalized mathematical model is used to predict oil and gas production, describing the entire process of oil and gas production growth, peak, and decline.

[0044] Class I generalized mathematical models take time t With as the independent variable and annual oil and gas production as the dependent variable, a prediction model for annual oil and gas production is established as follows: ;in, Indicates annual oil and gas production. a, b, c These are model constants. m This is a classification factor for the model.

[0045] Based on historical oil and gas parameter data, the least squares method was used to fit the parameters in the model.

[0046] S3: Resilience Assessment of Converged Technologies and Economic Indicators

[0047] Resilience refers to a system's ability to absorb shocks, maintain function, and recover after being subjected to disturbances. Integrating technical and economic indicators, a real-time resilience assessment of the subsea production system is conducted from two aspects: performance retention and production stability. The performance retention metric is remaining service life, and the production stability metric is oil and gas production. Changes in remaining service life and oil and gas production during the resilience assessment are shown below. Figure 2 As shown.

[0048] Performance retention measures the degree to which an underwater production system maintains its functionality after being disturbed. Specifically, it is the proportion of the system's remaining real-time lifetime over its entire lifespan compared to its remaining lifetime under conditions without any disturbance. The real-time performance retention of an underwater production system at time t is calculated as follows: ;in, RUL ( t )express t The real-time remaining service life of an underwater production system is the remaining service life under the combined effects of natural degradation and external impacts. rul ( t )express t The expected remaining service life of the underwater production system at any given time, i.e., the remaining service life under conditions of only natural degradation. t 0 is the starting point of the performance resilience assessment cycle, which is generally 0.

[0049] The stability of oil and gas production in subsea production systems is largely dependent on the system's operational status. Accurately characterizing the impact of system performance degradation on production is crucial for system resilience assessment and condition-based maintenance optimization. This paper introduces performance impact factors and constructs a correction function for oil and gas production in subsea production systems to describe the deviation of actual production from the ideal state. The corrected production after considering the impact of performance degradation is... for ;in, rul ( t ) indicates that the system at time t t The remaining useful life, Φ(t) is the performance impact function, which reflects the proportion of the actual degradation state on the production capacity.

[0050] Production stability describes the degree of fluctuation in oil and gas production before and after a disturbance event. Specifically, it represents the proportion of oil and gas production retained at the current moment compared to the production before the disturbance. The real-time production stability of a subsea production system at time t is... R pro ( t ) calculated as To achieve a comprehensive evaluation of system resilience, a comprehensive resilience evaluation model for underwater production systems integrating technology and economy is constructed based on the evaluation results of performance retention and production stability.

[0051] The real-time resilience of an underwater production system is calculated as follows: RE(t) = λ·R rul (t)+(1-λ)·R pro (t) ;in, λ As a weighting coefficient for the technical dimension, it can be flexibly adjusted according to the management objectives of different operational stages to achieve a balance between security and economy.

[0052] S4: Resilience-Driven Condition-Based Maintenance Decision Optimization Method

[0053] When an underwater production system is subjected to external disturbances, its remaining service life and oil and gas production will decrease to varying degrees. When the system's real-time resilience reaches the maintenance threshold, maintenance activities should be carried out. After maintenance, the system's performance and production will recover significantly, but will not reach the initial level before the disturbance during the maintenance cycle. This type of maintenance activity is called incomplete maintenance. The effect of incomplete maintenance is described by the service life rollback factor. The degree of imperfect maintenance is related to maintenance costs and the degree of system aging. Generally speaking, the higher the maintenance cost, the better the maintenance quality. On the other hand, the degree of system aging is another important factor affecting the service life improvement factor. In the early stages of the system's service life, significant performance improvements can be achieved with relatively small maintenance costs. As the system ages, the optimization effect obtained from the same maintenance cost will gradually weaken. The service life rollback factor is calculated as follows: pm =1-( cm / CR) (T-RUL(t)) / MRL ;in, pm This represents the system's maintenance life reduction factor. cm This indicates the planned maintenance costs, depending on the situation. CR This indicates the replacement cost. T This indicates the total lifespan of the underwater production system. MRL The system's average remaining lifespan, after non-compliance maintenance, is: RUL(t+1) = T - (T - RUL(t)) * pm; where, RUL(t +1 ) express t +1 indicates the real-time remaining service life of the underwater production system after the completion of any maintenance activity.

[0054] Based on the real-time remaining service life of the subsea production system after the completion of condition-based maintenance activities, the oil and gas production of the subsea production system can be updated after the completion of the maintenance activities. The remaining service life and oil and gas production recovery process of the subsea production system are presented in a stepped manner. The height of each step represents the lifespan or production of the component system restored by the maintenance activity, and the width of each step represents the time required to repair the component. The maintenance time of the subsea production system is determined based on the experience data of previous maintenance. According to the weight of the components, the maintenance time of each component can be reasonably allocated and the maintenance effect of each component can be evaluated, thereby constructing an accurate recovery process.

[0055] Since incomplete repairs make it difficult to restore valves to their initial state, their performance tends to decline after each repair. As the number of repairs increases, the reliability of the components gradually decreases, eventually requiring replacement to restore their performance. To improve the reliability of the components, they need to be replaced after multiple incomplete repair activities, i.e., a complete repair approach should be adopted. Based on past repair experience, underwater production system components should be replaced after a cumulative total of 3 incomplete repairs.

[0056] Over the entire service life of a component, the maintenance cost includes both incomplete maintenance costs and complete maintenance costs. ;in, C ALL The total cost of condition-based maintenance and replacement activities for underwater production systems. M j For components j The number of repairs depending on the situation. R j For components j Number of replacements; t mj For components j The time required to perform a condition-based maintenance; t rj For components j The time required to carry out a change activity; C mj For components j The unit time cost of performing condition-based maintenance. C rj For components j The unit time cost of replacement.

[0057] By analyzing the real-time dynamic changes in the resilience of the underwater production system throughout its entire service life, and using the quantified resilience value as the key independent variable, a decision model is established with the dual optimization objectives of maximizing system resilience and minimizing maintenance costs. A multi-objective particle swarm optimization algorithm is used to perform parallel search of the maintenance decision space. The game relationship between objectives is handled through non-dominated ranking and elite retention strategies. An equilibrium solution is selected from the Pareto optimal solution set, and the optimal resilience value is finally output as the condition-based maintenance trigger threshold.

[0058] Example 2

[0059] like Figure 3 As shown, the underwater production system includes a hydraulic power unit 101, a main control station 102, a power unit 103, an above-water umbilical cable terminal assembly 104, an umbilical cable 105, an underwater umbilical cable terminal assembly 106, an underwater control module 107, and a wellhead 108.

[0060] The hydraulic power unit 101 is connected to the surface umbilical cable terminal assembly 104 via a hydraulic hose bundle to provide hydraulic power to the underwater equipment. The main control station 102 is connected to the surface umbilical cable terminal assembly 104 via a fiber optic cable to issue operating commands and transmit underwater status data to the underwater equipment. The power unit 103 is connected to the surface umbilical cable terminal assembly 104 via a power cable to provide power to the underwater equipment. The surface umbilical cable terminal assembly 104 is directly connected to the umbilical cable 105 to transmit hydraulic power, operating commands, and power to the underwater equipment. The umbilical cable 105 is directly connected to the underwater umbilical cable terminal assembly 106 to distribute hydraulic power, operating commands, power, and injected chemicals. The underwater umbilical cable terminal assembly 106 is connected to the underwater control module 107 via a jumper cable to transmit control signals and hydraulic power. The underwater control module 107 is directly connected to the Christmas tree 108 to control the Christmas tree 108.

[0061] A resilience-driven condition-based maintenance decision optimization method for underwater production systems is proposed and applied to a resilience-driven condition-based maintenance decision optimization system for underwater production systems.

[0062] like Figure 4 As shown, the resilience-driven underwater production system condition-based maintenance decision optimization system includes a remaining useful life prediction module 201, a system natural degradation performance prediction unit 202, a component natural degradation performance prediction subunit 203, a component weight evaluation subunit 204, a system performance prediction unit under external shock influence 205, an external shock arrival time prediction subunit 206, an external shock intensity prediction subunit 207, an oil and gas production prediction module 208, a real-time resilience assessment module 209, a technical resilience assessment unit 210, an economic resilience assessment unit 211, a service life regression factor calculation unit 212, a maintenance cost calculation unit 213, a condition-based maintenance decision module 214, and a maintenance decision output module 215.

[0063] The remaining useful life prediction module 201 includes a system natural degradation performance prediction unit 202 and a system performance prediction unit 205 under external shock influence; the system natural degradation performance prediction unit 202 includes a component natural degradation performance prediction subunit 203 and a component weight evaluation subunit 204; the system performance prediction unit 205 under external shock influence includes an external shock arrival time prediction subunit 206 and an external shock intensity prediction subunit 207; the real-time resilience assessment module 209 includes a technical resilience assessment unit 210 and an economic resilience assessment unit 211; and the condition-based maintenance decision module 214 includes a service life rollback factor calculation unit 212 and a maintenance cost calculation unit 213.

[0064] The remaining service life prediction module 201 predicts the remaining service life of the system based on its performance under natural degradation conditions and external shock effects. The system natural degradation performance prediction unit 202 calculates the system's performance under natural degradation conditions based on the component natural degradation performance prediction results and component weight evaluation results. The component natural degradation performance prediction subunit 203 is connected to the main control station 102 via a cable and uses historical and monitoring data to predict the performance of each component of the underwater production system under natural degradation conditions. The component weight evaluation subunit 204 is connected to the component natural degradation performance prediction subunit 203 via a cable and is used to calculate the importance of each component in the underwater production system. The system performance prediction unit 205 under external shock effects calculates the performance of the underwater production system under external shock effects based on the predicted arrival time and intensity of the external shock. The external shock arrival time prediction subunit 206 simulates the frequency of external shocks based on historical marine environmental data. The external shock intensity prediction subunit 207 is connected to the external shock arrival time prediction subunit 206 via a cable and is used to simulate the intensity of each external shock. Strength; Oil and gas production prediction module 208 is connected to remaining useful life prediction module 201 via cable, used to predict oil and gas production under the combined influence of natural degradation conditions and external shocks; Real-time resilience assessment module 209 assesses the comprehensive resilience of the subsea production system in real time based on technical resilience assessment results and economic resilience assessment results; Technical resilience assessment unit 210 is connected to remaining useful life prediction module 201 via cable, used to assess technical resilience with remaining useful life as an indicator; Economic resilience assessment unit 211 is connected to oil and gas production prediction module 208 via cable, used to assess economic resilience with oil and gas production as an indicator; Service life regression factor calculation unit 212 is used to calculate the effect of each incomplete maintenance activity; Maintenance cost calculation unit 213 is used to calculate the total cost of maintenance of the subsea production system; Condition-based maintenance decision module 214 is connected to real-time resilience assessment module 209 via cable, used to make condition-based maintenance decisions under the dual constraints of resilience and cost; Maintenance decision output module 215 is connected to condition-based maintenance decision module 214 via cable, used to output the results of resilience-driven condition-based maintenance decision optimization.

[0065] During the operation of the resilience-driven underwater production system condition-based maintenance decision optimization system, the remaining service life prediction module 201 predicts the remaining service life of the system based on its performance under natural degradation conditions and external shocks; the oil and gas production prediction module 208 predicts the oil and gas production under the combined influence of natural degradation conditions and external shocks; the real-time resilience assessment module 209 assesses the comprehensive resilience of the underwater production system in real time based on the technical resilience assessment results and economic resilience assessment results; the condition-based maintenance decision module 214 makes condition-based maintenance decisions under the dual constraints of resilience and cost; and the maintenance decision output module 215 outputs the results of the resilience-driven condition-based maintenance decision optimization and transmits them to the main control station 102 for display, so that relevant maintenance personnel can extract relevant information to formulate condition-based maintenance plans.

[0066] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A resilience-driven method for optimizing condition-based maintenance decisions in underwater production systems, characterized in that, include: S1. Predicting the remaining service life of the system: Modeling the natural degradation performance of components based on the Weibull distribution, quantifying the random impact degradation by combining the Poisson-normal impact model, and superimposing the total degradation performance of the system to predict the remaining service life; S2. Predicting oil and gas production: Fitting historical production data with a type I generalized mathematical model to establish a production prediction function; S3. Real-time resilience assessment: Integrating technical and economic indicators, calculating the comprehensive resilience value through the performance retention function and the production stability function; S4. Maintenance decision optimization: Using the real-time resilience value as an input variable, establishing a dual-objective decision model that maximizes system resilience and minimizes maintenance costs, and using a multi-objective optimization algorithm to solve for the optimal maintenance trigger threshold; S1 includes: Calculating the natural degradation performance of components through the Weibull cumulative distribution function. In the formula, F j (t) represents the performance of the j-th component at time t, typically the component's failure probability, where t represents the time variable, and η j Let β be the scale parameter of the j-th component. j The shape parameter of the j-th component; also includes: the average degradation performance of the component over its entire lifecycle, expressed as: In the formula, The average degradation performance of a component over its entire lifecycle is represented by T, where T represents the total lifecycle length of the underwater production system, and F represents the average degradation performance of the component over its entire lifecycle. j (t) represents the performance of the j-th component at time t; calculate the degradation sensitivity of component j. In the formula, N represents the number of components in the underwater production system; W represents the weight percentage of the j-th component with respect to the system degradation performance. j for The system's natural degradation performance is a weighted sum of the degradation performance of each component. In the formula, D N (t) represents the natural degradation performance of the system, W j (t) represents the weight coefficient of the j-th component, and N represents the total number of components; in the Poisson-normal shock model, the number of shock events follows a homogeneous Poisson process. In the formula, N(t) represents the total number of events occurring within the time interval [0,t], c represents the possible values ​​of the random variable N(t), and λ is the intensity parameter, which remains constant throughout the time interval; the degradation amount of a single impact follows a normal distribution; the cumulative degradation performance of the system caused by the external impact event is... ; where Y k Let the system degradation performance caused by the k-th impact event follow a normal distribution with mean μY and standard deviation σY; the total degradation performance of the system is the sum of the cumulative degradation performance caused by natural degradation and the cumulative degradation performance caused by the external impact event; when the system performance degrades to 0, the system reaches a failure state, and the remaining service life of the system is expressed as the time interval between the first time the system degradation performance reaches 0, which can be calculated as: RUL=inf{t:D(t)≤0|D(0)>0}; the generalized mathematical model of type I of S2 is... In the formula, Let S represent the annual oil and gas production, a, b, and c be model constants, and m be the model classification factor; the parameters a, b, c, and m are determined by fitting historical production data using the least squares method; in S3, the performance retention function is: In the formula, RUL(t) represents the real-time remaining service life of the underwater production system at time t', i.e., the remaining service life under the combined effects of natural degradation and external shocks; rul(t) represents the expected remaining service life of the underwater production system at time t', i.e., the remaining service life under only natural degradation conditions; t0 is the starting point of the performance resilience assessment period, which is generally 0; the production stability function is: In the formula, R pro (t) represents the production stability at time t'. Indicates annual oil and gas production. The actual output is adjusted for performance impact factors; the real-time resilience of the underwater production system is calculated as: RE(t) = λ·Rrul(t) + (1-λ)·R pro (t); where λ is the weighting coefficient of the technology dimension; the actual output Corrected by performance impact factor In the formula, rul(t) represents the remaining service life of the system at time t`, and Φ(t) is the performance impact function, reflecting the proportion of the actual degradation state on the production capacity.

2. The resilience-driven condition-based maintenance decision optimization method for underwater production systems according to claim 1, characterized in that, The S4 uses a service life rollback factor to describe the effect of incomplete maintenance: pm=1-(cm / CR)(T-RUL(t)) / MRL; where pm represents the system's service life reduction factor, cm represents the planned cost of condition-based maintenance, CR represents the replacement cost, T represents the total life cycle length of the underwater production system, and MRL represents the system's average remaining life; the remaining life after maintenance is updated as: RUL(t+1)=T-(T-RUL(t))*pm; where RUL(t+1) represents the real-time remaining life of the underwater production system after the condition-based maintenance activity ends at time t+1.

3. The resilience-driven condition-based maintenance decision optimization method for underwater production systems according to claim 2, characterized in that, The repair cost includes both incomplete repair costs and complete repair costs. In the formula, C ALL The total cost of condition-based maintenance and replacement activities for underwater production systems, M j R represents the number of condition-based repairs required for component j. j t represents the number of times component j is replaced. mj The time required for a condition-based maintenance of component j, t rj The time required for component j to perform a replacement activity, C mj C represents the unit time cost of performing condition-based maintenance on component j. rj M represents the unit time cost of replacing component j; where M j R j These represent the number of incomplete repairs and the number of replacements for component j, respectively.

4. The resilience-driven condition-based maintenance decision optimization method for underwater production systems according to claim 3, characterized in that, After three incomplete repairs are performed on the same component, a complete repair is performed.

5. A resilience-driven underwater production system condition-based maintenance decision optimization system, used to implement the resilience-driven underwater production system condition-based maintenance decision optimization method according to claim 4, characterized in that, include: The remaining useful life prediction module (201) includes a system natural degradation performance prediction unit (202) and a system performance prediction unit under external shock influence (205); the system natural degradation performance prediction unit (202) includes a component natural degradation performance prediction subunit (203) and a component weight evaluation subunit (204), used to calculate weighted natural degradation performance; the system performance prediction unit under external shock influence (205) includes an external shock arrival time prediction subunit (206) and an external shock intensity prediction subunit (207), used to establish a Poisson-normal shock model; the oil and gas production prediction module (208) is connected to the remaining useful life prediction module (201) via a cable, and predicts production based on a generalized mathematical model; real-time resilience The evaluation module (209) includes a technical resilience evaluation unit (210) and an economic resilience evaluation unit (211), which are connected to the remaining useful life prediction module (201) and the oil and gas production prediction module (208) via cables, respectively, to calculate the comprehensive resilience value by integrating performance retention and production stability; the condition-based maintenance decision module (214) is connected to the real-time resilience evaluation module (209) via cables to establish a dual-objective decision model; the maintenance decision output module (215) is connected to the condition-based maintenance decision module (214) via cables to output the optimal maintenance trigger threshold to the main control station (102); wherein, each module realizes data transmission via cables, and the system natural degradation performance prediction unit (202) adopts Weibull distribution modeling component degradation.

Images