Method for predicting overall horizontal force of deepwater jacket based on physical characteristics

Abstract

The invention discloses a deepwater jacket platform overall horizontal force prediction method based on physical characteristics, and belongs to the field of ocean deepwater jacket digital twinning. According to the method, on the basis of a physical calculation formula of the overall horizontal force of the platform, a nonlinear mapping relation between wind, wave and flow environment loads and the overall horizontal force of the platform is captured in combination with a convolutional neural network, and the intelligent prediction method for the overall horizontal force of the deepwater jacket platform based on physical characteristics is established. The method can accurately predict the overall horizontal force of the deepwater jacket platform under the action of ocean environment loads such as wind, waves and flow so as to realize real-time monitoring and performance evaluation of the overall performance state of the platform structure, and can be widely applied to the field of ocean deepwater jacket digital twinning.

Description

Technical Field

[0001] The present invention relates to the field of marine engineering and adopts a method for predicting the overall horizontal force of a deepwater jacket platform based on a convolutional neural network (CNN). Background Art

[0002] With the continuous development of marine resource development and offshore engineering, the stability and safety of deepwater jacket platforms, as key facilities for offshore oil and gas extraction, are crucial to ensuring a stable supply of marine resources and the safety of personnel. Deepwater jacket platforms exhibit significant nonlinear dynamic characteristics under the influence of complex marine environmental loads such as wind, waves, and currents, making the prediction of the platform's overall performance a significant challenge.

[0003] Traditional deepwater jacket platform design and performance assessment methods often rely on empirical formulas or simplified physical models. These methods have limitations when dealing with nonlinear dynamics and struggle to accurately capture the platform's response characteristics under complex marine loads. Furthermore, while traditional numerical simulation methods can provide relatively accurate results, they are computationally expensive and cannot meet the demands of real-time monitoring and rapid response. Summary of the Invention

[0004] To address the challenges of existing technologies, this paper provides a physical-characteristics-based method for predicting the overall horizontal forces of deepwater jacket platforms. This method employs a convolutional neural network (CNN) to predict the overall horizontal forces of deepwater jacket platforms, addressing the existing limitations of inaccuracy and efficiency. By constructing a high-precision overall indicator prediction model, this paper accurately predicts the overall horizontal forces of deepwater jacket platforms subjected to marine loads such as wind, waves, and currents. This model improves prediction accuracy and efficiency, providing technical support for the design, monitoring, and maintenance of deepwater jacket platforms.

[0005] The technical solution adopted by the present invention is: a method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics, which specifically includes the following steps:

[0006] S1. Extract monitoring data and pre-process it according to the structural characteristics of the deepwater jacket platform;

[0007] S2. Calculate the overall horizontal force of the deepwater jacket using the monitoring data in combination with the calculation method and formula for the overall horizontal force of the deepwater jacket;

[0008] S3. Establishing a mapping relationship between monitoring data and overall horizontal force in a deepwater jacket;

[0009] S4. Predicting the overall horizontal force of the deepwater jacket platform using an intelligent prediction algorithm for the overall horizontal force of the deepwater jacket based on physical characteristics. This specifically includes the following sub-steps:

[0010] S4.1 Build a database of the platform's overall performance characteristics and divide the dataset into training, validation, and test sets;

[0011] S4.2 takes the combined loads of wind, wave and current ocean environment as the model input and the overall horizontal force of the deepwater jacket as the model output;

[0012] S4.3 Construct a CNN prediction model structure consisting of an input layer, a convolutional layer, a dropout layer, a fully connected layer, and an output layer;

[0013] S4.4 combines the prediction model and database to iteratively train the CNN prediction model, including parameter initialization, backpropagation, parameter update, and cross-validation;

[0014] S4.5 evaluates and optimizes the overall horizontal force prediction model of the deepwater jacket platform through the above steps to obtain the optimal hyperparameters of the model.

[0015] Furthermore, the step S2 specifically includes the following sub-steps:

[0016] S2.1 Calculate the axial force F of the diagonal brace and support leg attached to the measuring point by using stress and strain sensors installed on the diagonal brace and support leg at the deepwater jacket base X and F T ;

[0017] S2.2 Establish the overall force balance equation of the deepwater jacket to describe the environmental load F and the support reaction F at support points A and B. A 、F B static equilibrium relationship;

[0018] S2.3 Establish the local force balance equation of the deepwater jacket to describe the support reaction force F at support points A and B. A 、F B Related to the brace axial force F XA 、F XB , supporting leg axial force F TA 、F TB static equilibrium relationship;

[0019]

[0020] S2.4 Solve for the environmental load F and deduce the overall horizontal force F of the platform w ;

[0021]

[0022] in: is the horizontal unit vector.

[0023] Furthermore, in step S4.2, the load data includes wind speed, wind direction, flow speed, flow direction, wave direction, wave height, and period.

[0024] The beneficial effects of the present invention are: 1. Compared with traditional deepwater jacket platform design and performance evaluation methods that often rely on empirical formulas or simplified physical models, the present invention can better handle nonlinear problems and accurately capture the response characteristics of the platform under complex marine environment loads.

[0025] 2. Compared with traditional numerical simulation methods, the prediction method in the present invention has low computational cost and can more quickly meet the needs of real-time monitoring and rapid response.

[0026] 3. The present invention elaborates on the calculation method of the key characteristics of the overall horizontal force of the platform, ensuring the accuracy and reliability of the input data and laying the foundation for high-precision prediction of the model.

[0027] 4. The present invention establishes a mapping relationship between multiple monitoring data and the overall horizontal force of the deep-water jacket, taking into account the indicators of the overall horizontal force of the deep-water jacket under the joint load distribution, and can more comprehensively and accurately capture the complex information in the prediction network, thereby more effectively predicting the overall horizontal force of the deep-water jacket platform.

[0028] 5. By constructing a high-precision CNN prediction model, this invention can capture the overall performance of the platform structure, which is crucial for achieving a digital twin of the platform's overall performance. The CNN-based jacket overall horizontal force prediction model achieves a 97.8% accuracy rate for predicting the platform's overall horizontal force. This facilitates more precise monitoring and assessment of the operating status of deepwater jacket platforms in complex marine environments, contributing to the digital and intelligent development of marine engineering and possessing potential engineering application value and innovative significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is the flow chart for constructing the platform's overall horizontal force physical characteristic dataset.

[0030] Figure 2 This is a flowchart of the training and verification of the CNN-based jacket overall horizontal force prediction model.

[0031] Figure 3 This is the fitting effect diagram of the jacket overall horizontal force prediction model based on CNN. DETAILED DESCRIPTION

[0032] The specific embodiments of the present invention are described in further detail below in conjunction with the accompanying drawings. However, it should be understood that the drawings are provided only for a better understanding of the present invention and should not be construed as limiting the present invention.

[0033] A method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics comprises the following steps:

[0034] S1. Extract monitoring data with stable data quality based on the structural characteristics of the deepwater jacket platform and perform preprocessing on it;

[0035] S2. The axial force of the diagonal brace and support leg near the measuring point is calculated by using stress and strain sensors installed on the diagonal brace and support leg at the deepwater jacket base, and the overall horizontal force F of the platform is deduced. w ; Specifically includes the following sub-steps:

[0036] S2.1 Calculate the axial force F of the diagonal brace and support leg near the measuring point using stress and strain sensors installed on the diagonal brace and support leg at the base of the deepwater jacket. X and F T ;

[0037] S2.2 Establish the overall force balance equation of the deepwater jacket to describe the environmental load F and the support reaction F at support points A and B. A 、F B static equilibrium relationship;

[0038] S2.3 Establish the local force balance equation of the deepwater jacket to describe the support reaction force F at support points A and B. A 、F B Related diagonal bracing and supporting leg axial force F XA 、F TA , F XB 、F TB static equilibrium relationship;

[0039]

[0040] S2.4 Solve for the environmental load F and deduce the overall horizontal force F of the platform w ;

[0041]

[0042] in: is the horizontal unit vector.

[0043] S3. Establishing a mapping relationship between monitoring data and overall horizontal force of the platform in the deepwater jacket;

[0044] S4. Use the deepwater jacket overall horizontal force intelligent prediction algorithm based on physical characteristics to accurately predict the overall horizontal force of the deepwater jacket platform. This step specifically includes the following sub-steps:

[0045] S4.1 Build a database of the platform's overall performance characteristic dataset, and divide the total dataset into a training set, a validation set, and a test set;

[0046] S4.2 takes the combined loads of wind, wave and current ocean environment as the model input and the overall horizontal force of the deepwater jacket as the model output;

[0047] S4.3 Construct a CNN prediction model structure consisting of an input layer, a convolutional layer, a dropout layer, a fully connected layer, and an output layer;

[0048] S4.4 combines the prediction model and the database to iteratively train the CNN prediction model, including steps such as parameter initialization, backpropagation, parameter update, and cross-validation;

[0049] S4.5 evaluates and optimizes the overall horizontal force prediction model of the deepwater jacket platform through the above steps to obtain the optimal hyperparameters of the model;

[0050] S4.6 uses the trained deepwater jacket platform overall horizontal force prediction model and uses MAE, RMSE, and MAPE evaluation indicators to evaluate and test the overall horizontal force of the deepwater jacket platform under marine environmental loads such as wind, waves, and currents, proving the rationality of the designed model. Figure 1 This is a flowchart for constructing a dataset of physical characteristics of the platform's overall horizontal forces. By building a high-precision overall indicator prediction model, we achieve accurate prediction of the overall horizontal forces of deepwater jacket platforms under ocean loads such as wind, waves, and currents. During implementation, the collected data on deepwater jacket platforms under different ocean loads is first normalized to meet the input requirements of the CNN model. Next, the platform's overall horizontal forces are extracted from the dataset of overall platform performance characteristics.

[0051] The CNN model architecture consists of an input layer, two convolutional layers, two dropout layers, a fully connected layer, and an output layer. The convolutional layers use one-dimensional convolutions with kernels of 1x3 to accommodate the scale of the data features. In these layers, local features are extracted by sliding the kernels across the input feature map. To improve the model's generalization, two dropout layers are added instead of pooling layers.

[0052] After training is complete, use the trained CNN model to predict the test data and compare the predicted values ​​with the actual values ​​to evaluate the model's accuracy. Use metrics such as MAE and RMSE to test and evaluate model performance. Based on the evaluation results, adjust the model structure or hyperparameters to improve prediction accuracy.

[0053] Example 1

[0054] S1. Combining the structural characteristics of the deepwater jacket platform with sensor monitoring data (wind, wave, current, and member axial force information, etc.), after analysis, select sensors with stable data quality for analysis and modeling, and perform data processing on their data;

[0055] S2. Combined with the calculation method and formula of the horizontal force of deepwater jacket, the overall horizontal force F of the platform is calculated by the axial forces of 8 diagonal braces and 4 legs at the base of Liuhua 11-1. w

[0056] S3. Establish a mapping relationship between monitoring data (environmental load data wind speed, wind direction, flow speed, flow direction, wave direction, wave height, and period) in deepwater jackets and the overall horizontal force of the platform; obtain a characteristic dataset of the overall horizontal force of the platform, and divide it into 70% training set and validation set, and 30% test set.

[0057] S4. Establishing a CNN structure including input layer, convolution layer, dropout layer, fully connected layer and output layer to establish an intelligent prediction model of the overall horizontal force of the deepwater jacket based on physical characteristics;

[0058] S5. Combine the above network prediction model and call the database to iteratively train the CNN network model, including initialization parameters, back propagation, parameter update, cross validation and other steps. The platform sensor network mapping model training process is as follows: Figure 2 As shown;

[0059] S6. Through the training process in the evaluation and tuning of the sensor network mapping model, the optimal hyperparameters of the model are obtained, including 7 input features, 5 convolutional neural network layers, 128 hidden layers, 64 batch size, 1000 training rounds, 20 early stopping rounds, 0.001 learning rate, Adam optimizer, and mean square error loss function.

[0060] The trained CNN-based jacket overall horizontal force prediction model is used to perform predictions using the test set data that did not participate in the training, and the predicted values ​​and actual values ​​are plotted into a fitting curve. Figure 3 The CNN prediction model was fitted to the test data, demonstrating the overall horizontal force Fx and Fy of the platform along the X and Y directions, demonstrating the model's reliability. Verification results on the test data set showed that the CNN-based jacket overall horizontal force prediction model achieved an accuracy of 97.8%.

[0061] The above embodiments are only used to illustrate the present invention, wherein the structure and environmental load of the deepwater jacket are subject to change. Any equivalent transformations and improvements based on the technical solution of the present invention should not be excluded from the scope of protection of the present invention.

Claims

1. A method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics, characterized by: The following steps are involved: S1. Extract monitoring data and pre-process it according to the structural characteristics of the deepwater jacket platform; S2. Calculate the horizontal force of the deepwater jacket using the monitoring data by combining the calculation method and formula of the horizontal force of the deepwater jacket; S3. Establishing a mapping relationship between monitoring data and overall horizontal force of the platform in the deepwater jacket; S4. Predicting the overall horizontal force of the deepwater jacket platform using an intelligent prediction algorithm for the overall horizontal force of the deepwater jacket based on physical characteristics. This specifically includes the following sub-steps: S4.1 Build a database of the platform's overall performance characteristic dataset, and divide the total dataset into a training set, a validation set, and a test set; S4.2 takes the combined loads of wind, wave and current ocean environment as the model input and the overall horizontal force of the deepwater jacket as the model output; S4.3 Construct a CNN prediction model structure consisting of an input layer, a convolutional layer, a dropout layer, a fully connected layer, and an output layer; S4.4 combines the prediction model and the database to iteratively train the CNN prediction model; S4.5 evaluates and optimizes the overall horizontal force prediction model of the deepwater jacket platform through the above steps to obtain the optimal hyperparameters of the model.

2. The method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics according to claim 1, characterized in that: The step S2 specifically includes the following sub-steps: S2.1 Calculate the axial force F of the diagonal brace and support leg near the measuring point using stress and strain sensors installed on the diagonal brace and support leg at the base of the deepwater jacket. X and F T ; S2.2 Establish the overall force balance equation of the deepwater jacket to describe the environmental load F and the support reaction F at support points A and B. A 、F B static equilibrium relationship; S2.3 Establish the local force balance equation of the deepwater jacket to describe the support reaction force F at support points A and B. A 、F B Related to the brace axial force F XA 、F XB , supporting leg axial force F TA 、F TB static equilibrium relationship; ； S2.4 Solve for the environmental load F and deduce the overall horizontal force F of the platform w ; ； in: is the horizontal unit vector.

3. The method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics according to claim 1, characterized in that: In step S4.2, the load data includes wind speed, wind direction, flow speed, flow direction, wave direction, wave height, and period.

4. The method for predicting the overall horizontal force of a deepwater jacket based on physical characteristics according to claim 1, characterized in that: In step S4.4, the CNN prediction model is iteratively trained, including parameter initialization, back propagation, parameter update, and cross-validation.

Images