CN-122022074-A - Accident early warning method and system for FPSO mooring system under extreme weather condition

Abstract

The invention belongs to the technical field of oil and gas exploration and development, and relates to an accident early warning method and system of an FPSO mooring system under extreme weather conditions, wherein the accident early warning method comprises the steps of establishing an accident scene library according to extreme weather data; the method comprises the steps of grading accidents in an accident scene library according to expert opinions, fusing the expert opinions through a fuzzy mathematic method to obtain the weight of each expert opinion, combining the expert opinion with the weight to obtain a key node accident prediction result of an FPSO mooring system, and inputting extreme weather condition data and the key node accident prediction result which change along with time into a dynamic Bayesian network model to obtain an accident early warning result of the FPSO mooring system. The method has the advantages that the accident occurrence probability of the whole mooring system, the subsystem and the equipment unit under the extreme weather condition is calculated, the accident early warning is realized, and the data support is provided for field emergency protection.

Inventors

• LI DA
• CAO YANG
• ZHANG YUE
• XU SHUAI
• CHEN KUNLIANG
• YANG YAQI
• CHEN CHUANGANG
• ZHOU WEI
• LIU TAO
• WANG KUITAO
• WANG CHUNQING
• HAO JINGMIN
• HU ZHONGQIAN
• ZHANG HAIJUAN

Assignees

• 中国海洋石油集团有限公司
• 中海油研究总院有限责任公司

Dates

Publication Date

20260512

Application Date

20260410

Claims (10)

1. 1. An accident pre-warning method for an FPSO mooring system under extreme weather conditions is characterized by comprising the following steps: Establishing an accident scene library according to extreme weather data; grading the accidents in the accident scene library according to expert opinion; fusing expert opinions through a fuzzy mathematic method to obtain the weight of each expert opinion; combining expert opinion with the weight to obtain a key node accident prediction result of the FPSO mooring system; and inputting the extreme weather condition data changing along with time and the critical node accident prediction result into a dynamic Bayesian network model to obtain an accident early warning result of the FPSO mooring system.
2. 2. The method for early warning of an accident in an FPSO mooring system under extreme weather conditions according to claim 1, wherein the method for establishing an accident scene library is as follows: Obtaining extreme weather data according to weather forecast, and carrying out fine division; classifying the results of the refined division by combining the actual conditions of the FPSO mooring system; and establishing an extreme weather scene related to the accident according to the grading result.
3. 3. The method for early warning of an accident of an FPSO mooring system under extreme weather conditions according to claim 2, wherein the extreme weather includes lightning, wind, waves and currents, the fine division is divided into lightning intensity and lightning density, wind into wind speed, waves into effective wave height, and currents into flow velocity.
4. 4. The FPSO mooring system accident pre-warning method according to claim 1, wherein the expert opinions are classified into seven grades of very low, medium, high and very high, and the expert opinions are converted by triangle blur numbers and trapezoid blur numbers.
5. 5. The method for early warning of an FPSO mooring system accident in extreme weather conditions according to claim 4, wherein the membership function of the expert opinion is: Wherein, the Is a membership function with very low expert opinion; Is a membership function with low expert opinion; Is expert opinion as a lower membership function; is a membership function with moderate expert opinion; The expert opinion is a higher membership function; Is a membership function with high expert opinion; is a membership function with high expert opinion, and x is a quantitative value of an object to be evaluated in a specific evaluation dimension.
6. 6. The method for early warning of an FPSO mooring system accident under extreme weather conditions according to claim 4, wherein the method for calculating the weight of each expert opinion is as follows: establishing an evaluation result similarity formula among experts; calculating a similarity matrix between every two experts according to the similarity formula; calculating the average value of consistency degree of the evaluation result of the expert i and the evaluation results of other experts according to the similarity matrix; and calculating the sum weight of the expert according to the average value of the consistency degree of the expert i.
7. 7. The FPSO mooring system accident pre-warning method under extreme weather conditions according to claim 6, wherein the similarity formula is: Wherein the evaluation result of expert i is expressed as The evaluation result of expert j is expressed as , Is the similarity of the evaluation opinions of expert i and expert j.
8. 8. The FPSO mooring system accident pre-warning method under extreme weather conditions according to claim 6, wherein the expert's sum weights are: Wherein, the Is the sum weight of the expert, Is the initial weight of the expert, Is a weight obtained based on the evaluation result, which is obtained by dividing the average value of the degree of coincidence of the expert i by the average value of the degree of coincidence of all the experts, Is a coordination coefficient used to balance the expert initial weight with the adjustment parameters that assess the relative importance of the weights.
9. 9. The FPSO mooring system accident pre-warning method according to claim 1, wherein the dynamic bayesian network model describes the process of the FPSO mooring system changing under severe environmental conditions through a series of time slices, each time slice reflects the state of the mooring system and its affiliated facilities at the current moment, and each time slice contains state variables and observation variables, wherein the state variables refer to discrete dynamic nodes, the observation variables refer to static nodes, and the dynamic bayesian network model satisfies markov assumptions and observation assumptions: The Markov assumption is that the state variable in the current state is related to the last state only and is unrelated to all states except the last state; The observation assumes that the variable value of the observation variable depends only on the current state and is independent of the state at other moments.
10. 10. An accident pre-warning system for an FPSO mooring system in extreme weather conditions, comprising: The weather data acquisition module is used for establishing an accident scene library according to the extreme weather data; the grading module is used for grading the accidents in the accident scene library according to expert opinion; The weight confirmation module is used for fusing expert opinions through a fuzzy mathematic method to obtain the weight of each expert opinion; The static prediction result module is used for combining expert opinion with the weight to obtain a key node accident prediction result of the FPSO mooring system; And the dynamic prediction result module is used for inputting the extreme weather condition data changing along with time and the critical node accident prediction result into a dynamic Bayesian network model to obtain an accident early warning result of the FPSO mooring system.

Description

Accident early warning method and system for FPSO mooring system under extreme weather condition Technical Field The invention relates to an accident early warning method and system for an FPSO (floating production, storage and offloading) mooring system under extreme weather conditions, and belongs to the technical field of oil and gas exploration and development. Background The FPSO mooring system is core equipment for offshore oil and gas operation safety, and a complex constraint system is formed by anchor chains, steel cables, polyester cables, connectors and accessory equipment of the anchor chains, the steel cables, the polyester cables and the connectors and is used for resisting marine environment loads such as wind, waves, currents and the like. In severe environments with high temperature and high humidity in deep open sea and frequent strong typhoons, mooring systems in long-term service face multiple failure risks including effective sectional area reduction caused by corrosion, fatigue damage accumulation caused by alternating load and overload risks in extreme weather sea conditions. Further, accidents such as anchor chain breakage, fire disaster, explosion and the like of the mooring system are increased, and the service safety of the mooring system is seriously threatened. In order to judge the health condition of the mooring system under extreme weather conditions, the traditional risk assessment method mainly comprises static analysis methods such as accident trees, event trees and the like, but the methods are only suitable for static health condition analysis and accident prevention of the mooring system, and the dynamic health condition of the mooring system is difficult to judge. However, the extreme weather conditions are continuously changed along with the time line, and the method cannot embody the characteristic of dynamic risk change, so that the method is not suitable for the field of early warning of mooring system accidents under the extreme weather conditions. In order to overcome the defects of the static analysis method, a dynamic analysis model of time variation such as a dynamic accident tree, a Markov chain, a dynamic Bayesian network and the like provides a new thought for research. However, implementation of the dynamic analysis model function described above requires determining the probability of occurrence of the basic event. The probability of occurrence of many basic events is unknown in production practice. The risk conditions for the mooring system overall system, subsystems and component facilities under extreme weather conditions are difficult to obtain from historical data. Therefore, in order to grasp the real-time health condition of the mooring system under the extreme weather condition and realize the early warning of the accident, it is necessary to provide an accident early monitoring and early warning method which can accurately judge the occurrence probability of each basic event under the extreme weather condition and dynamically judge the health condition of the mooring system so as to realize the purposes of early discovery and early prevention. Disclosure of Invention Aiming at the problems, the invention aims to provide an accident early warning method and system for an FPSO mooring system under extreme weather conditions, which can accurately early warn the health condition of the mooring system and auxiliary equipment under the extreme weather conditions and realize the service safety of the FPSO. The technical scheme includes that an accident scene library is built according to extreme weather data, accidents in the accident scene library are classified according to expert opinions, the weight of each expert opinion is obtained through fuzzy mathematic method fusion of the expert opinion, the expert opinion is combined with the weight to obtain a key node accident prediction result of the FPSO mooring system, and extreme weather condition data and the key node accident prediction result which change along with time are input into a dynamic Bayesian network model to obtain the accident early warning result of the FPSO mooring system. The method for establishing the accident scene library comprises the steps of obtaining extreme weather data according to weather forecast, carrying out fine division, grading the result of the fine division by combining the actual situation of the FPSO mooring system, and establishing an extreme weather scene related to an accident according to the grading result. Further, the extreme weather includes lightning, wind, waves and currents, the fine division is divided into lightning intensity and lightning density, wind into wind speed, waves into effective wave height, and currents into flow velocity. Further, the expert opinions are classified into seven grades of very low, medium, high and very high, and are converted through triangle blur numbers and trapezoid blur numbers. Further, the membership functions of t