Event assessment and aid decision-making method and system for offshore work platform

Abstract

The invention provides an event assessment and auxiliary decision-making method and system for an offshore operation platform, and the method comprises the steps: taking platform data in different complex operation scenes as input, employing different key data preprocessing methods, and obtaining event assessment key factors and relationships of the whole offshore platform; an evaluation analysis model is established to analyze and evaluate the whole platform event space, auxiliary decision support data is improved, visual display is carried out with different granularities or different abstract levels, and the function of accurately evaluating and deciding platform events is achieved. According to the method, the decision-making speed is increased by analyzing, concluding and integrating complicated data accumulated by subsystem equipment of the maritime satellite launching and recycling platform and carrying out classified analysis on evaluation data by utilizing an intelligent optimization genetic algorithm; an event evaluation model constructed based on a fusion algorithm is applied to assist the platform and personnel in performing remote diagnosis and decision-making on the operation of the offshore operation platform, so that risk threats caused by emergencies are reduced, and the operation efficiency of the platform is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of event-oriented evaluation and decision-making, and in particular relates to an event evaluation and auxiliary decision-making method and system for an offshore operating platform. Background Art

[0002] With the continuous development and application of offshore satellite launch and recovery platforms, the demand for intelligent operation and maintenance management of operating platforms is also expanding. When the platform encounters an emergency or a marine disaster, it will seriously affect the operation status of the platform. It is necessary to evaluate and deal with the events during the operation in a timely manner to assist in the decision-making of platform operation and maintenance.

[0003] The background of offshore platform assessment includes platform design and construction: the design of offshore platforms must take into account extreme marine environmental conditions, including strong winds, large waves, salt spray corrosion, etc.; structural safety assessment: as the platform ages, structural safety assessments need to be conducted regularly to ensure that the platform meets safety standards; environmental impact assessment: the impact of offshore platforms on the surrounding environment also needs to be assessed, including but not limited to marine ecosystems, water pollution, etc.; risk management: offshore operations face multiple risks, such as platform structure failure, blowout accidents, casualties, etc., so comprehensive risk management is needed.

[0004] The background of decision-making support for offshore operating platforms includes complex operating environments: offshore operating platforms face extreme and changeable environmental conditions, such as strong winds, large waves, deep water, etc., which place high demands on the design, construction, and operation of the platforms; high operating risks: offshore operations involve high-risk activities such as drilling, oil and gas production, and maintenance, which require strict safety management measures; huge amounts of data: a large amount of data is generated during offshore operations, including meteorological information, platform status, operation progress, etc., and effective methods are needed to process and analyze these data.

[0005] With the development of intelligent theories such as artificial intelligence and machine learning, as well as information technology, the integration of remote services such as remote evaluation and graph decision-making through the systematization, standardization, and intelligence of big data has become an inevitable trend for the further development of platform intelligent technology. However, due to the limitations of platform equipment and technical conditions, emergencies or abnormal operations often require detailed evaluation and rapid response measures to ensure the continued effective operation of the platform. How to ensure accuracy and efficiency in the evaluation process has become a hot and difficult issue in platform-assisted decision-making. Event evaluation is affected by many factors, including the severity of the event impact, the professional skills level of the platform equipment, the time required for the evaluation process, and the occurrence of adverse events.

[0006] In recent years, domestic and foreign researchers have studied and evaluated the influencing factors in the operation of different platforms, mainly using three tools: modified sequential assessment, early warning assessment, and event hazard assessment. These tools provide some help in offshore platform operation and maintenance decision-making. Platform early warning assessment research can provide a reference for event operation and maintenance management. However, most of these studies are based on scoring tools such as single platform status characteristics and inspection results, lacking systematic big data intelligent analysis. As the application of intelligent algorithm technology in platform data mining, image processing, and intelligent diagnosis has gradually been welcomed and recognized by the industry, it has become possible to accurately evaluate and make decisions on platform events, thereby maximizing the operating efficiency of offshore platforms. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide an event assessment and auxiliary decision-making method and system for offshore operating platforms, which are used for accurately assessing and making decisions on platform events.

[0008] The technical solution adopted by the present invention to solve the above technical problems is: an event assessment and auxiliary decision-making method for offshore operating platforms, comprising the following steps: S1: Acquire and pre-process the data of offshore operating platforms under different operation scenarios; S2: Detect anomalies or damages of offshore platforms based on knowledge graphs, extract feature datasets from preprocessed data and train event assessment models, and update model parameters to detect and identify events; S3: Evaluate the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; S4: Integrate event assessment results and historical data to visualize and output real-time decision results at different granularity or abstraction levels.

[0009] According to the above scheme, in step S1, the specific steps are: S11: Monitor and obtain real-time data of offshore platforms through sensors; S12: Pre-process key data by establishing a database and analyzing data to obtain key factors and relationships for event assessment of offshore operating platforms.

[0010] Furthermore, in the step S12, the specific steps are: Filter data based on meaning and characteristics, organize, convert and train collected data and calculate the correlation between data; merge or integrate data belonging to different databases based on correlation; find out the data that has the greatest impact on the event through data analysis, and determine whether to collect further data.

[0011] According to the above scheme, in step S2, the specific steps are: S21: Construct an association graph of operating factors of offshore operation platforms, form a set of characteristic factors for analyzing events, extract constituent elements of the knowledge graph from the preprocessed data and store them in the database; S22: Use the knowledge graph to extract features and embed nodes on entity nodes, obtain key information in the data, generate feature representations of platform events, and obtain feature data sets; S23: Combine deep learning, federated learning, and distributed learning methods to build an event assessment model, use the final results of emergencies as training labels, and use the feature datasets extracted by knowledge graph analysis to train the attributes of the event assessment model.

[0012] According to the above scheme, in step S3, the specific steps are: Assume that the fault tree has n sets, and each event in the fault tree is independent, and the probability of an event occurring at time t is F i (t), the probability of occurrence of each event refers to the historical data of the platform, then the probability of event occurrence, that is, the risk result P(t) of the offshore operating platform at time t, is: .

[0013] Furthermore, in the step S3, the events during the platform operation process include failure of the launch platform or premature failure, failure of the observation tower, failure of link coordination, human failure, environmental failure, fuze failure and other failure events.

[0014] According to the above scheme, in step S4, the specific steps are: Analyze fault repair situations using evaluation data; introduce constraints and penalty functions to assist fault decision-making; calculate the total resource consumption for solving a certain fault event based on the fault repair resource constraints of each device on the offshore platform; Plan maintenance resource allocation and sequence decision results, use scatter plots to evaluate data visualization and analyze data clustering and observe data distribution and whether there are abnormal points; Data with two attributes are represented in a plane form; data with three attributes are represented in a three-dimensional form; data sets with multiple attributes are represented using a scatter matrix, where each cell represents the distribution of the data in two dimensions.

[0015] An event assessment and decision support system for offshore platforms. The preprocessing submodule is used to obtain and preprocess the data of the offshore operating platform under different operating scenarios; The event recognition submodule is used to detect anomalies or damages of offshore platforms based on the knowledge graph, extract feature datasets from preprocessed data, train event assessment models, and update model parameters to detect and identify events; The probability assessment submodule is used to assess the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; The visual decision submodule is used to integrate event evaluation results and historical data, and visualize the real-time decision results at different granularities or abstraction levels.

[0016] An event assessment model training system includes a central computing unit and a plurality of local computing units; The local computing unit is used to calculate the gradient of the model locally, encrypt the gradient using technologies including homomorphic encryption and differential privacy, and send the encrypted gradient to the central computing unit. It is also used to decrypt the aggregated results sent back by the central computing unit and update the model parameters with the decrypted gradient to obtain the target model of the event evaluation model under the condition of data isolation. The central computing unit is used to securely aggregate the received encrypted gradients using a federated averaging algorithm and send the aggregation results back to the local computing units.

[0017] A computer memory stores a computer program executable by a computer processor, wherein the computer program executes an event assessment and auxiliary decision-making method for an offshore operating platform.

[0018] The beneficial effects of the present invention are: 1. The present invention provides an event assessment and decision support method and system for offshore operating platforms. The method takes platform data under different complex operating scenarios as input, adopts different key data preprocessing methods, and obtains the key factors and relationships of event assessment for the entire offshore platform. An assessment and analysis model is established to analyze and evaluate the entire platform event space, improve the decision support data, and perform visualization at different granularities or different levels of abstraction, thereby realizing the functions of accurate assessment and decision-making of platform events.

[0019] 2. The present invention analyzes, summarizes and integrates the complex data accumulated by the subsystem equipment of the offshore satellite launch and recovery platform, applies an event assessment model constructed based on a fusion algorithm, and assists the platform and personnel in remote diagnosis and decision-making on the operation of the offshore operating platform, thereby reducing the risk threats caused by emergencies and improving the platform operation efficiency.

[0020] 3. The present invention uses an event assessment model to conduct hazard assessment and send probability calculation for risk events on offshore operating platforms, which plays a significant role as supporting data for auxiliary decision-making; and uses an intelligent optimization genetic algorithm to classify and analyze the assessment data to improve the decision-making speed; the decision is made by judging the hazard and probability of occurrence of the event, and the judgment logic is simple and not prone to errors. The present invention has achieved better results in enhancing the platform data evaluation speed, reducing the complexity of the decision logic, and improving the decision accuracy.

[0021] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0023] Figure 1 is a flow chart of an embodiment of the present invention.

[0024] Figure 2 It is a flowchart of an embodiment of the present invention.

[0025] Figure 3 It is a characteristic representation diagram of the platform event of the embodiment of the present invention.

[0026] Figure 4 It is an input-output diagram of the event evaluation model of the embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0028] Example 1 See also Figure 1 ,The specific steps of an event assessment and decision-making support method for offshore operating platforms are as follows: S1: Acquire and pre-process the data of offshore operating platforms under different operation scenarios; S2: Detect anomalies or damages of offshore platforms based on knowledge graphs, extract feature datasets from preprocessed data and train event assessment models, and update model parameters to detect and identify events; S3: Evaluate the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; S4: Integrate event assessment results and historical data to visualize and output real-time decision results at different granularity or abstraction levels.

[0029] Furthermore, in step S1, the specific steps are: S11: Monitor and obtain real-time data of offshore platforms through sensors; S12: Pre-process key data by establishing a database and analyzing data to obtain key factors and relationships for event assessment of offshore operating platforms.

[0030] Furthermore, in step S12, the specific steps are: filtering data according to meaning and characteristics, organizing, converting and training the collected data and calculating the correlation between the data; merging or integrating data belonging to different databases according to the correlation; finding out the data with the greatest impact on the event through data analysis, and determining whether to collect further data.

[0031] In step S2, the specific steps are: S21: Construct an association graph of operating factors of offshore operation platforms, form a set of characteristic factors for analyzing events, extract constituent elements of the knowledge graph from the preprocessed data and store them in the database; S22: Use the knowledge graph to extract features and embed nodes on entity nodes, obtain key information in the data, generate feature representations of platform events, and obtain feature data sets; S23: Combine deep learning, federated learning, and distributed learning methods to build an event assessment model, use the final results of emergencies as training labels, and use the feature datasets extracted by knowledge graph analysis to train the attributes of the event assessment model.

[0032] In step S3, the specific steps are: Assume that the fault tree has n sets, and each event in the fault tree is independent, and the probability of an event occurring at time t is F i (t), the probability of occurrence of each event refers to the historical data of the platform, then the probability of event occurrence, that is, the risk result P(t) of the offshore operating platform at time t, is: .

[0033] Furthermore, in step S3, events during the platform operation process include failure of the launch platform or premature failure, failure of the observation tower, failure of link coordination, human failure, environmental failure, fuze failure and other failure events.

[0034] In step S4, the specific steps are: Analyze fault repair situations using evaluation data; introduce constraints and penalty functions to assist fault decision-making; calculate the total resource consumption for solving a certain fault event based on the fault repair resource constraints of each device on the offshore platform; Plan maintenance resource allocation and sequence decision results, use scatter plots to evaluate data visualization and analyze data clustering and observe data distribution and whether there are abnormal points; Data with two attributes are represented in a plane form; data with three attributes are represented in a three-dimensional form; data sets with multiple attributes are represented using a scatter matrix, where each cell represents the distribution of the data in two dimensions.

[0035] This embodiment uses platform data under different complex operating scenarios as input, adopts different key data preprocessing methods, and obtains the key factors and relationships of event assessment of the entire offshore platform; establishes an assessment and analysis model to analyze and evaluate the entire platform event space, improves auxiliary decision-making support data, and performs visualization at different granularities or different levels of abstraction, thereby realizing the functions of accurate assessment and decision-making of platform events.

[0036] Example 2 See also Figure 2 The steps of this embodiment are the same as those of embodiment 1, except that each step is applied to a specific example. Specifically, the following steps are included: S1: Acquire and pre-process offshore platform data under different operation scenarios; S11: Offshore platforms are usually equipped with a large number of sensors to monitor various parameters. First, real-time data collection and monitoring are required. Sensors are used to monitor and obtain real-time data of offshore platforms, including the health status and safety-related data of the platform. are k sensors S1, S2, ..., S k The sampling points of each sensor are n sets of data, then:

[0037] S12: Pre-process key data by establishing a database and analyzing data to obtain key factors and relationships for event assessment of offshore operating platforms; Filter out all potentially useful data based on the meaning and characteristics of the data, then organize the data, and perform appropriate conversion and training on the collected data to increase the relevance between the data; τ 2, τ 3, ..., τ k represents the similarity between the sampling points of the 2nd to kth sensors and the sampling point of the 1st sensor, N represents the similarity coefficient between the sampling points, then the kth sensor sampling points The correlation is:

[0038] Establish a database and analyze the data; merge or integrate the data belonging to different databases according to the relevant information. After the database is established, conduct data analysis and find out the data with the greatest impact, and then determine whether it is necessary to further collect more detailed data.

[0039] S2: See Figure 4 , detect anomalies or damages of offshore platforms based on knowledge graphs, extract feature datasets from preprocessed data and train event assessment models, and update model parameters to detect and identify events; S21: Construct a correlation map of platform operation factors to form a set of characteristic factors for analyzing events Y , extract the constituent elements of the knowledge graph such as entities, relationships and attributes from a large number of structured preprocessing platform data records and store them in the database:

[0040] in, is the event characteristic factor set, Including equipment failure type data, operation-related data, and environmental factor change values.

[0041] S22: Using knowledge graph to identify entity nodes X Perform feature extraction and node embedding to obtain key information in the data and generate feature representations of platform events for accurate diagnosis prediction, such as Figure 3 As shown; the graph neural network convolution formula used for feature extraction is as follows:

[0042] in, , is the activation function, W is the training parameter of the graph neural network, A is the adjacency matrix of the entity node, and I N The identity matrix, is the degree matrix whose value is the sum of each row of the adjacency matrix, and Z is the entity node X The embedding vector is the feature representation of each entity.

[0043] S23: Build an event assessment model by combining deep learning, federated learning, and distributed learning methods. Use the final results of emergencies (platform cannot operate, platform has defects that cannot be handled, platform is not damaged) as training labels, and train model attributes using feature data sets extracted from knowledge graph analysis.

[0044] During the training process, the platform calculates the model gradient locally and uses technologies such as homomorphic encryption and differential privacy to protect the gradient information. The encrypted gradient is then sent to the central controller. The central controller uses the federated average algorithm for secure aggregation. The aggregated results are then sent back to the calculation models of each platform. Each model decrypts the received encrypted gradient and uses the decrypted gradient to update the model parameters. Finally, the target model is obtained under the condition of data isolation.

[0045] S3: Evaluate the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; The fault tree analysis method (FTA) based on the offshore platform is used to evaluate the event risk. The probability of occurrence of various events during the platform operation, such as launch station failure or premature failure, observation tower failure, link coordination failure, human failure, environmental failure, fuze failure and other failure events, is calculated. Assuming that the fault tree has n sets and each fault event in the fault tree is independent, the probability of event failure is

[0046] F i (t) is the probability of occurrence of a fault event at time t. The probability of occurrence of each event refers to the past data of the platform. P(t) represents the risk result of the offshore operating platform at time t.

[0047] S4: Integrate event assessment results and historical data to visualize and output real-time decision results at different granularity or abstraction levels; After completing the platform operation event detection and evaluation, the event evaluation results and historical data are integrated to provide real-time decision-making suggestions for operators and users. First, the evaluation data is used to analyze the fault maintenance situation, and the maintenance function is defined as , represents the total resources required for fault repair, where N represents the maintenance stage (10 minutes for one stage), and x represents the resources required for stage maintenance. Then, an intelligent optimization algorithm for fault repair auxiliary decision-making is designed. First, the fault repair item set of each equipment on the offshore platform is defined. , Indicates the maintenance sub-items of each platform equipment and defines resource constraints:

[0048] It is a maintenance item Required resources, is the maintenance resource strength that can be provided by the offshore operating platform in stage k. The algorithm objective function is the resources required for maintenance, that is, the total resource consumption to solve a certain fault event. Constraints and penalty functions are introduced to process fault auxiliary decision-making. The resource constraint limit jump results are counted. If the constraint is followed, the value is 1, otherwise the value is 0. The penalty function is as follows:

[0049] Wherein, e represents the parameter direction, a represents the product of the resource constraint hopping results, and b represents the sum of the resource constraint hopping results.

[0050] The maintenance resource allocation and sequential decision results can be planned, and scatter plots can be used to evaluate data visualization and analyze the clustering of data, and observe the distribution of data and whether there are abnormal points, so that operators and users can make manual decisions. For data with two or three attributes, a plane or three-dimensional representation is used; for data sets with multiple attributes, a scatter matrix is ​​used, and each cell represents the distribution of data in two dimensions.

[0051] This embodiment takes the offshore satellite launch and recovery platform under complex operation scenarios as the object, and designs an event evaluation and decision-making support method based on a fusion algorithm to solve the problems that the platform data is difficult to systematically analyze and the evaluation model is inefficient, so as to achieve the purpose of iterative optimization of the system architecture. In terms of decision-making efficiency, the decision-making support system can quickly process a large amount of data and help decision makers make accurate judgments more quickly; in terms of operational risks, by simulating and predicting various possible situations, the decision-making support system can help formulate safer operational plans; in terms of operational processes, through data analysis and model building, the decision-making support system can discover bottlenecks and room for improvement in the operational process, thereby improving overall efficiency; in terms of emergency response support, in emergency situations, the decision-making support system can provide a rapid emergency response plan to reduce losses.

[0052] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0053] Example 3 This embodiment is used to implement the principles of the above method embodiments to build an event assessment and auxiliary decision-making system for offshore operating platforms, including a preprocessing submodule, an event recognition submodule, a probability assessment submodule and a visualization decision submodule; the preprocessing submodule is used to obtain and preprocess the data of the offshore operating platform under different operating scenarios; the event recognition submodule is used to detect anomalies or damages of the offshore operating platform based on the knowledge graph, extract feature data sets from the preprocessed data and train the event assessment model, and update the model parameters to detect and identify events; the probability assessment submodule is used to assess the probability of occurrence of events on the offshore operating platform based on the fault tree FTA analysis method; the visualization decision submodule is used to integrate event assessment results and historical data, and visualize and output real-time decision results at different granularities or abstract levels.

[0054] Each sub-module is mainly used to implement each step of the method embodiment, which will not be described in detail here.

[0055] Example 4 This embodiment is used to implement the principle of the above method embodiment to build an event assessment model training system, including a central unit and several local units; The local unit is used to calculate the gradient of the model locally, encrypt the gradient using technologies including homomorphic encryption and differential privacy, and send the encrypted gradient to the central unit. It is also used to decrypt the aggregated results sent back by the central unit and update the model parameters with the decrypted gradient to obtain the target model of the event evaluation model under the condition of data isolation. The central unit is used to securely aggregate the received encrypted gradients using a federated averaging algorithm and send the aggregation results back to the local units.

[0056] Assume that the platform local model set is S, and the training model of the subsystem / device is B k S, according to the federated learning method, B k Train locally separately. After training the t+1th batch of samples, upload the training gradients through homomorphic encryption ,in Based on the previous gradient Continue training to get it.

[0057]

[0058] Similarly, the computing models of other platforms will also obtain the model gradients of the t+1th batch and send the encrypted gradients to the central controller, which uses the federated average algorithm for secure aggregation operations:

[0059] Where n is the sum of the sample sizes of each subsystem / equipment, n k For subsystem / equipment B k The number of samples is K, and K is the total number of subsystems / devices involved in model training. The central controller aggregates the results W t +1 Send it to each computing model; each computing model decrypts the received gradient and uses the decrypted gradient result to update its own model parameters and continue the next batch of training. After the training is completed, the model can be used to detect and identify events.

[0060] In terms of intelligence and automation, this embodiment improves the intelligence level of the platform through the Internet of Things, big data and artificial intelligence technologies; in terms of environmental protection technology, the impact of platform operations on the environment is reduced through the research and development of new technologies; in terms of new materials and new technologies, the use of lighter and stronger materials and technologies improves the safety and efficiency of the platform.

[0061] It should be pointed out that, according to the needs of implementation, the various steps / components described in this application can be split into more steps / components, and two or more steps / components or partial operations of steps / components can be combined into new steps / components to achieve the purpose of the present invention.

[0062] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.

Claims

1. An event assessment and decision support method for an offshore operating platform, characterized in that: The following steps are involved: S1: Acquire and pre-process the data of offshore operating platforms under different operation scenarios; S2: Detect anomalies or damages of offshore platforms based on knowledge graphs, extract feature datasets from preprocessed data and train event assessment models, and update model parameters to detect and identify events; S3: Evaluate the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; S4: Integrate event assessment results and historical data to visualize and output real-time decision results at different granularity or abstraction levels.

2. The event assessment and decision-making support method for an offshore operating platform according to claim 1, characterized in that: In the step S1, the specific steps are: S11: Monitor and obtain real-time data of offshore platforms through sensors; S12: Pre-process key data by establishing a database and analyzing data to obtain key factors and relationships for event assessment of offshore operating platforms.

3. The event assessment and decision-making support method for an offshore operating platform according to claim 2, characterized in that: In the step S12, the specific steps are: Filter data based on meaning and characteristics, organize, convert and train collected data and calculate the correlation between data; merge or integrate data belonging to different databases based on correlation; find out the data that has the greatest impact on the event through data analysis, and determine whether to collect further data.

4. The event assessment and decision-making support method for an offshore operating platform according to claim 1, characterized in that: In the step S2, the specific steps are: S21: Construct an association graph of operating factors of offshore operation platforms, form a set of characteristic factors for analyzing events, extract constituent elements of the knowledge graph from the preprocessed data and store them in the database; S22: Use the knowledge graph to extract features and embed nodes on entity nodes, obtain key information in the data, generate feature representations of platform events, and obtain feature data sets; S23: Combine deep learning, federated learning, and distributed learning methods to build an event assessment model, use the final results of emergencies as training labels, and use the feature datasets extracted by knowledge graph analysis to train the attributes of the event assessment model.

5. The event assessment and decision support method for an offshore operating platform according to claim 1, characterized in that: In the step S3, the specific steps are: Assume that the fault tree has n sets, and each event in the fault tree is independent, and the probability of an event occurring at time t is F i (t), the probability of occurrence of each event refers to the historical data of the platform, then the probability of event occurrence, that is, the risk result P(t) of the offshore operating platform at time t, is: 。 6. The event assessment and decision-making support method for an offshore operating platform according to claim 5, characterized in that: In the step S3, the events during the platform operation process include failure of the launch platform or premature failure, failure of the observation tower, failure of the coordination of links, human failure, environmental failure, fuze failure and other failure events.

7. The event assessment and decision-making support method for an offshore operating platform according to claim 1, characterized in that: In the step S4, the specific steps are: Analyze fault repair situations using evaluation data; introduce constraints and penalty functions to assist fault decision-making; calculate the total resource consumption for solving a certain fault event based on the fault repair resource constraints of each device on the offshore platform; Plan maintenance resource allocation and sequence decision results, use scatter plots to evaluate data visualization and analyze data clustering and observe data distribution and whether there are abnormal points; Data with two attributes are represented in a plane form; data with three attributes are represented in a three-dimensional form; data sets with multiple attributes are represented using a scatter matrix, where each cell represents the distribution of the data in two dimensions.

8. An event assessment and decision support system for offshore platforms, characterized in that: The preprocessing submodule is used to obtain and preprocess the data of the offshore operating platform under different operating scenarios; The event recognition submodule is used to detect anomalies or damages of offshore platforms based on the knowledge graph, extract feature datasets from preprocessed data, train event assessment models, and update model parameters to detect and identify events; The probability assessment submodule is used to assess the probability of occurrence of events on offshore platforms based on the fault tree FTA analysis method; The visual decision submodule is used to integrate event evaluation results and historical data, and visualize the real-time decision results at different granularities or abstraction levels.

9. An event assessment model training system, characterized in that: It includes a central computing unit and several local computing units; The local computing unit is used to calculate the gradient of the model locally, encrypt the gradient using technologies including homomorphic encryption and differential privacy, and send the encrypted gradient to the central computing unit. It is also used to decrypt the aggregated results sent back by the central computing unit and update the model parameters with the decrypted gradient to obtain the target model of the event evaluation model under the condition of data isolation. The central computing unit is used to securely aggregate the received encrypted gradients using a federated averaging algorithm and send the aggregation results back to the local computing units.

10. A computer memory, characterized in that: A computer program executable by a computer processor is stored therein, and the computer program executes an event assessment and auxiliary decision-making method for an offshore operating platform as described in any one of claims 1 to 7.

Images