Prediction method, system, equipment and medium for marine transportation liquefied natural gas unloading composition
By constructing a prediction model for offshore transport of LNG and using artificial neural networks and machine learning algorithms to screen component influencing factors, the accuracy of prediction of LNG composition changes is solved, and transportation efficiency and safety are improved.
Patent Information
- Application Number
- CN202411407812.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-10
- Publication Date
- 2024-12-27
AI Technical Summary
The prior art cannot accurately predict the changes in the heating components of liquefied natural gas during offshore transportation, resulting in the different components of the liquefied natural gas at the unloading port from the loading port, affecting transportation efficiency and safety.
By obtaining the composition data and component impact data of the loading port and the unloading port, screening the component impact factors with high correlation, using artificial neural networks and machine learning algorithms to build a prediction model for the unloading composition of the liquefied natural gas at sea transport, and training the model to Predicted liquefied natural gas composition at the unloading port.
It achieves efficient and accurate prediction of the composition of liquefied natural gas, improves the reliability and efficiency of liquefied natural gas transportation, and provides reliable data support.
Smart Images
Figure CN119204909A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of data processing methods specifically suitable for prediction purposes, and in particular to a prediction method, system, equipment and medium for unloading composition of liquefied natural gas transported at sea. Background Art
[0002] Liquefied natural gas is a form of liquid that is cooled to extremely low temperatures (about minus 162 degrees Celsius). After liquefaction, liquefied natural gas can be transported over long distances by means of transportation such as ships, railways or trucks. LNG carriers are usually specially designed with good insulation and safety measures to ensure that they maintain a constant low temperature during transportation and avoid leakage or accidents. After arriving at the destination, liquefied natural gas can be re-gasified and converted back into natural gas to supply the energy market or industrial use.
[0003] Although the cargo tanks are highly insulated, heat input from the surrounding environment is unavoidable, which will vaporize the liquid and produce boiling gas. The vaporization of LNG is not uniform because each LNG component has a different boiling point. This phenomenon is called "aging" and causes the composition of LNG to change over time. Therefore, the composition and properties of LNG at the discharge port (destination port) will be slightly different from that at the loading port (origin port).
[0004] Some studies have established a simple mathematical model for the aging process of LNG during storage and use, and obtained the variation pattern of LNG fraction concentration, but only considered the situation where the components contain methane and ethane, and did not consider other trace components, resulting in the results not being consistent with the actual situation, making this method impractical. Other studies have predicted the composition of LNG at the unloading port based on a constant evaporation rate, but this method is difficult to implement accurate heat transfer calculations, resulting in the inability to accurately predict the composition of LNG at the unloading port.
[0005] Therefore, there is an urgent need for a new method, system, equipment and medium for predicting the unloading composition of liquefied natural gas transported at sea. Summary of the invention
[0006] The present invention provides a method, system, equipment and medium for predicting the unloading composition of liquefied natural gas transported at sea, so as to solve the defect that the prior art cannot efficiently and accurately predict the changes of the heated components of liquefied natural gas, thus restricting the transportation of natural gas.
[0007] The present invention provides a method for constructing a prediction model for the unloading composition of liquefied natural gas transported at sea, comprising: Obtain LNG composition data at the loading port, LNG composition data at the unloading port, and LNG component impact data; According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the liquefied natural gas component influence factors correlated with the liquefied natural gas composition of the unloading port are screened out, and the liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factors are obtained from the liquefied natural gas component influence data; The liquefied natural gas composition data and liquefied natural gas component influencing factor data at the loading port are used as sample data, and the liquefied natural gas composition data at the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea.
[0008] In one embodiment, the liquefied natural gas composition data includes concentration data of any one of the following or any combination thereof: methane, ethane, propane, n-butane, isobutane, n-propane, isopropane, and nitrogen. In one embodiment, the liquefied natural gas component influencing data includes any one of the following or any combination thereof: liquefied natural gas composition at the loading port, temperature at the unloading port, pressure and name of the loading port and the unloading port, liquid volume and filling volume, sailing time, aging volume, ship name, and trip number.
[0009] In one embodiment, the liquefied natural gas component influencing factors that are correlated with the liquefied natural gas composition at the unloading port are screened out based on the liquefied natural gas composition data at the loading port, the liquefied natural gas composition data at the unloading port, and the liquefied natural gas component influencing data, including: According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the correlation coefficient of each liquefied natural gas component influence factor in the liquefied natural gas component influence data is obtained by using the first expression; According to the correlation coefficient of each LNG component influencing factor in the LNG component influencing data, the LNG component influencing factors that are correlated with the LNG composition at the unloading port are screened out.
[0010] In one embodiment, the first expression is: , In the first expression, Indicates the liquefied natural gas component impact factor in the liquefied natural gas component impact data Factors affecting liquefied natural gas components The correlation coefficient of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components The average value of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components The average value of . In one embodiment, the LNG component influencing factors correlated with the LNG composition at the unloading port include any one of the following or any combination thereof: LNG composition at the loading port, temperature at the unloading port, voyage time, and aging amount.
[0011] In one embodiment, the liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea, including: The LNG composition data and LNG component influencing factor data of the loading port are used as sample data, and the LNG composition data of the unloading port is used as label data. The first prediction model of the unloading composition of LNG transported by sea is obtained by training using an artificial neural network (ANN). The liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data. The second marine transport liquefied natural gas unloading composition prediction model is trained using a machine learning algorithm. The prediction performances of the first marine transport liquefied natural gas unloading composition prediction model and the second marine transport liquefied natural gas unloading composition prediction model are compared to obtain a final marine transport liquefied natural gas unloading composition prediction model.
[0012] In one embodiment, the marine transport liquefied natural gas unloading composition prediction model includes an input layer, a hidden layer and an output layer, wherein the liquefied natural gas composition data and the liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data, and the marine transport liquefied natural gas unloading composition prediction model is trained, including: Receive sample data and label data through the input layer, and convert the sample data and label data into sample vectors and label vectors; The activation function is used in the hidden layer to make the marine liquefied natural gas unloading composition prediction model learn the mapping relationship between the sample vector and the label vector; The prediction result of the sample vector corresponding to the label vector is output through the output layer.
[0013] In one embodiment, the liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea, including: According to the prediction results of the sample data by the marine transport liquefied natural gas unloading composition prediction model and the label data corresponding to the sample data, the prediction deviation is obtained and the prediction deviation is corrected.
[0014] In one embodiment, the prediction deviation is obtained by using the prediction result of the sample data and the label data corresponding to the sample data according to the prediction model of the unloading composition of liquefied natural gas transported by sea, including: According to the prediction results of the sample data by the marine transport liquefied natural gas unloading composition prediction model and the label data corresponding to the sample data, the prediction deviation is obtained, and the prediction deviation is corrected using the second expression, wherein the second expression is: , In the second expression, Indicates The deviation of the eigenvector after correction, Indicates feature vectors, Indicates The predicted deviation between the predicted value and the label value of the feature vector.
[0015] In one embodiment, the liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea, including: In the process of training the prediction model for the unloading composition of liquefied natural gas transported at sea, the SHAP value method is used to analyze and quantify the contribution of each feature in the sample data to the prediction of the label data.
[0016] The present invention also provides a method for predicting the unloading composition of liquefied natural gas transported at sea, comprising: Receiving liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and liquefied natural gas component influencing factor data corresponding to the liquefied natural gas component influencing factor selected by the method for constructing a composition prediction model for unloading liquefied natural gas transported by sea as described in any one of the above items; According to the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and the data of the influencing factors of the liquefied natural gas components, the liquefied natural gas unloading composition prediction model for sea transport is obtained by the method for constructing the sea transport liquefied natural gas unloading composition prediction model described in any of the above items, and the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port is predicted.
[0017] The present invention also provides a prediction system for the unloading composition of liquefied natural gas transported at sea, comprising: A data receiving module, used to: receive liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port sent from at least one terminal and liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factor selected by the method for constructing a composition prediction model for unloading liquefied natural gas transported by sea as described in any one of the above items; A prediction module is used to predict the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port according to the liquefied natural gas composition data of the liquefied natural gas to be tested and the liquefied natural gas component influencing factor data, and to obtain the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port by using the prediction model for unloading composition of liquefied natural gas to be tested obtained by the method for constructing the prediction model for unloading composition of liquefied natural gas to be tested described in any one of the above items; The data output module is used to send the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port back to at least one terminal.
[0018] The present invention also provides an electronic device, comprising a processor and a memory storing a computer program, wherein when the processor executes the computer program, it implements any of the above-mentioned methods for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea and / or the method for predicting the unloading composition of liquefied natural gas transported by sea.
[0019] The present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any of the above-mentioned methods for constructing a model for predicting the unloading composition of liquefied natural gas transported by sea and / or a method for predicting the unloading composition of liquefied natural gas transported by sea.
[0020] The present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute any of the above-mentioned methods for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea and / or the method for predicting the unloading composition of liquefied natural gas transported by sea.
[0021] The present invention provides a method, system, device and medium for predicting the unloading composition of liquefied natural gas transported by sea. The method uses historical data of transporting liquefied natural gas from a loading port to a unloading port to screen out liquefied natural gas component influencing factors (such as temperature, voyage time, aging amount, etc.) that have a strong correlation with the liquefied natural gas composition at the unloading port, and learns the mapping relationship between the liquefied natural gas component data and the liquefied natural gas component data at the loading port through a model, so as to train a high-precision prediction model for the unloading composition of liquefied natural gas transported by sea. The model can be used to efficiently and accurately predict the liquefied natural gas composition data at the unloading port, provide reliable data support for the transportation of liquefied natural gas, and effectively improve the transportation efficiency of liquefied natural gas. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 A schematic flow chart of a method for predicting the unloading composition of liquefied natural gas transported at sea provided by the present invention.
[0024] Figure 2 A schematic diagram of the structure of a prediction system for unloading liquefied natural gas in marine transportation provided by the present invention.
[0025] Figure 3 This is a schematic structural diagram of an electronic device provided by the present invention. DETAILED DESCRIPTION
[0026] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments, and they should not be understood as limitations on the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. In the description of the present invention, it should be understood that the terms used are only for descriptive purposes and cannot be understood as indicating or implying relative importance.
[0027] Combine the following Figure 1-Figure 3 The present invention describes a method, system, device and medium for predicting the unloading composition of liquefied natural gas transported by sea. It should be noted that the execution subject of the method for predicting the unloading composition of liquefied natural gas transported by sea provided by the present invention can be any network side device / terminal side device that meets the technical requirements, such as a prediction device for the unloading composition of liquefied natural gas transported by sea, etc.
[0028] Figure 1 The figure is a flow chart of the method for predicting the unloading composition of liquefied natural gas transported by sea provided by the present invention. Figure 1 The present invention provides a method for predicting the unloading composition of liquefied natural gas transported by sea, which may include: Step S110, obtaining the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, wherein the liquefied natural gas composition data includes the concentration data of any one of the following items or any combination thereof: methane, ethane, propane, n-butane, isobutane, n-propane, isopropane, nitrogen, and the liquefied natural gas component influence data includes any one of the following items or any combination thereof: the liquefied natural gas composition of the loading port, the unloading port temperature, the pressure and name of the loading port and the unloading port, the liquid volume and filling volume, the sailing time, the aging volume, the ship name, and the trip number, all of which can be obtained by manual measurement or calculation.
[0029] Step S120: based on the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, screen out the liquefied natural gas component influence factors correlated with the liquefied natural gas composition of the unloading port, and obtain the liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factors from the liquefied natural gas component influence data, wherein the liquefied natural gas component influence factors correlated with the liquefied natural gas composition of the unloading port include any one of the following items or any combination thereof: the liquefied natural gas composition of the loading port, the temperature of the unloading port, the sailing time, and the aging amount.
[0030] In one embodiment, step S120 may include: According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the correlation coefficient of each liquefied natural gas component influence factor in the liquefied natural gas component influence data is obtained by using the first expression; According to the correlation coefficient of each LNG component influencing factor in the LNG component influencing data, the LNG component influencing factors that are correlated with the LNG composition at the unloading port are screened out; Among them, the first expression is: , In the first expression, in the first expression, Indicates the liquefied natural gas component impact factor in the liquefied natural gas component impact data Factors affecting liquefied natural gas components The correlation coefficient of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components The average value of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components The average value of .
[0031] The Pearson correlation coefficient PCC² is used to screen the LNG component influencing factors that are correlated with the LNG composition at the unloading port. The range of PCC² is always from 0 to 1. When the value of PCC² is close to 1, it indicates a strong correlation. In this embodiment, the PCC² threshold is limited to 0.81 to remove influencing factors with low correlation, and the removal does not cause any obvious loss of input information.
[0032] Step S130: Using the liquefied natural gas composition data and the liquefied natural gas component influencing factor data of the loading port as sample data, and using the liquefied natural gas composition data of the unloading port as label data, a training model is obtained to obtain a prediction model for the unloading composition of liquefied natural gas transported by sea.
[0033] In one embodiment, step S130 may include: The LNG composition data and LNG component influencing factor data of the loading port are used as sample data, and the LNG composition data of the unloading port is used as label data. The first prediction model of the unloading composition of LNG transported by sea is obtained by training using an artificial neural network (ANN). The liquefied natural gas composition data and the liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data. A machine learning algorithm (a random forest algorithm in this embodiment) is used to train and obtain a second marine transport liquefied natural gas unloading composition prediction model; The prediction performance of the first marine transport liquefied natural gas unloading composition prediction model and the second marine transport liquefied natural gas unloading composition prediction model are compared to obtain the final marine transport liquefied natural gas unloading composition prediction model, wherein the evaluation of the model prediction performance can be evaluated by the average absolute relative deviation (AARD), root mean square error (RMSE) and mean absolute error (MAE), R square, etc., wherein It corresponds to the true value of the predicted liquefied natural gas component impact factor a.
[0034]
[0035] In one embodiment, the marine transport liquefied natural gas unloading composition prediction model (in this embodiment, the first marine transport liquefied natural gas unloading composition prediction model) includes an input layer, a hidden layer and an output layer. Step S130 may include: receiving sample data and label data through an input layer, and converting the sample data and label data into sample vectors and label vectors, wherein the input layer extracts features from the data, combines feature information with weights to form a feature vector, so as to convert the sample data and label data into sample vectors and label vectors; The mapping relationship between the sample vector and the label vector is learned by using an activation function (ReLU function in this embodiment) in the hidden layer to enable the marine transportation liquefied natural gas unloading composition prediction model to learn the mapping relationship between the sample vector and the label vector; The prediction result of the sample vector corresponding to the label vector is output through the output layer.
[0036] In one embodiment, step S130 may further include: According to the prediction results of the sample data by the marine transport liquefied natural gas unloading composition prediction model and the label data corresponding to the sample data, the prediction deviation is obtained, and the prediction deviation is corrected using the second expression, wherein the second expression is: , In the second expression, Indicates The deviation of the eigenvector after correction, Indicates feature vectors, Indicates The predicted deviation between the predicted value and the label value of the feature vector.
[0037] During the model training process, the LNG composition data of the unloading port predicted each time is kept relevant to the corresponding training data, and then the deviation of each movement is calculated, the deviation is corrected, and the model training parameters are optimized. At this time, the deviation will gradually approach 1, and a sufficiently stable feature vector can be obtained, making the prediction accuracy of the marine transport LNG unloading composition prediction model higher.
[0038] In this embodiment, the training data is divided into a training set, a validation set, and a test set according to 8:1:1. The model is trained with the training set, the validation set is preliminarily validated, and then applied to the test set to analyze the model performance. The result obtained by applying the model trained with the training set to the test set is compared with the composition corresponding to the original test set, and then the error is calculated. The training set is fully trained 10 times in total, and the accuracy is tested on the test set each time. The model parameters with the highest test accuracy are the optimal model parameters and are saved as the optimal model. By detecting the accuracy of the first marine transport liquefied natural gas unloading composition prediction model, the RMSE error calculation results are shown in Table 1.
[0039] Table 1
[0040] The second marine transport LNG unloading composition prediction model was further applied to analyze the LNG composition at the unloading port. The RMSE error calculation results are shown in Table 2.
[0041] Table 2
[0042] In summary, for the two models, neural network and random forest, in terms of predicting the composition of liquefied natural gas, the overall data error calculation results show that the second marine transport liquefied natural gas unloading composition prediction model has a better prediction effect.
[0043] In one embodiment, step S130 may further include: In the process of training the prediction model for the unloading composition of LNG transported at sea, the SHAP value method is used to analyze and quantify the contribution of each feature in the sample data to the prediction of the label data, so as to verify the accuracy of the LNG component influencing factors screened out by the PCC² coefficient that are correlated with the LNG composition at the unloading port.
[0044] Step S140, receiving the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and the liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factor screened in step S120; Step S150: According to the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and the data of the influencing factors of the liquefied natural gas components, the marine transported liquefied natural gas unloading composition prediction model obtained in step S130 is used to predict the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port.
[0045] The method for predicting the unloading composition of liquefied natural gas transported by sea provided by the present invention uses historical data of transporting liquefied natural gas from a loading port to a unloading port to screen out liquefied natural gas component influencing factors (such as temperature, sailing time, aging amount, etc.) that have a strong correlation with the liquefied natural gas composition at the unloading port, and learns the mapping relationship between the liquefied natural gas component data and the liquefied natural gas component factor data at the loading port and the liquefied natural gas component data at the unloading port through a model, and trains a high-precision prediction model for the unloading composition of liquefied natural gas transported by sea, which can be used to efficiently and accurately predict the liquefied natural gas composition data at the unloading port, provide reliable data support for the transportation of liquefied natural gas, and effectively improve the transportation efficiency of liquefied natural gas.
[0046] The following is a description of the prediction system for the unloading composition of liquefied natural gas transported at sea provided by the present invention. The prediction system for the unloading composition of liquefied natural gas transported at sea described below and the prediction method for the unloading composition of liquefied natural gas transported at sea described above can be referred to each other.
[0047] Reference Figure 2 The present invention provides a prediction system for unloading composition of liquefied natural gas transported at sea, which may include: A data receiving module, used to: receive liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port sent from at least one terminal and liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factor selected by the method for constructing a composition prediction model for unloading liquefied natural gas transported by sea as described in any one of the above items; A prediction module is used to predict the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port according to the liquefied natural gas composition data of the liquefied natural gas to be tested and the liquefied natural gas component influencing factor data, and to obtain the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port by using the prediction model for unloading composition of liquefied natural gas to be tested obtained by the method for constructing the prediction model for unloading composition of liquefied natural gas to be tested described in any one of the above items; The data output module is used to send the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port back to at least one terminal.
[0048] Figure 3 An example of a physical structure diagram of an electronic device is shown in FIG. Figure 3 As shown, the electronic device may include: a processor 810, a communication interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication interface 820 and the memory 830 communicate with each other through the communication bus 840. The processor 810 may call the logic instructions in the memory 830 to execute any of the above-mentioned methods for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea and / or the prediction method for the unloading composition of liquefied natural gas transported by sea.
[0049] In addition, the logic instructions in the above-mentioned memory 830 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.
[0050] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea and / or the method for predicting the unloading composition of liquefied natural gas transported by sea as described in any of the above items.
[0051] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, is implemented to execute the method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea and / or the method for predicting the unloading composition of liquefied natural gas transported by sea as described in any of the above items.
[0052] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.
[0053] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.
[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for constructing a prediction model for the unloading composition of liquefied natural gas transported at sea, characterized in that: include: Obtain LNG composition data at the loading port, LNG composition data at the unloading port, and LNG component impact data; According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the liquefied natural gas component influence factors correlated with the liquefied natural gas composition of the unloading port are screened out, and the liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factors are obtained from the liquefied natural gas component influence data; The liquefied natural gas composition data and liquefied natural gas component influencing factor data at the loading port are used as sample data, and the liquefied natural gas composition data at the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea.
2. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to claim 1, characterized in that: The liquefied natural gas composition data includes the concentration data of any one of the following or any combination thereof: methane, ethane, propane, n-butane, isobutane, n-propane, isopropane, nitrogen; and / or, The data affecting LNG composition include any one of the following or any combination thereof: LNG composition at the loading port, temperature at the unloading port, pressure and name of the loading port and unloading port, liquid volume and filling volume, voyage time, aging volume, ship name, and trip number.
3. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to claim 2, characterized in that: According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the liquefied natural gas component influence factors correlated with the liquefied natural gas composition of the unloading port are screened out, including: According to the liquefied natural gas composition data of the loading port, the liquefied natural gas composition data of the unloading port and the liquefied natural gas component influence data, the correlation coefficient of each liquefied natural gas component influence factor in the liquefied natural gas component influence data is obtained by using the first expression; According to the correlation coefficient of each LNG component influencing factor in the LNG component influencing data, the LNG component influencing factors that are correlated with the LNG composition at the unloading port are screened out.
4. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to claim 3, characterized in that: The first expression is: ， In the first expression, in the first expression, Indicates the liquefied natural gas component impact factor in the liquefied natural gas component impact data Factors affecting liquefied natural gas components The correlation coefficient of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components The average value of Indicates the impact factor of liquefied natural gas components The value of Indicates the impact factor of liquefied natural gas components the average value of; and / or, The liquefied natural gas component influencing factors that are correlated with the liquefied natural gas composition at the unloading port include any one of the following or any combination thereof: the liquefied natural gas composition at the loading port, the temperature at the unloading port, the voyage time, and the aging amount.
5. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to any one of claims 1 to 4, characterized in that: The liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea, including: The liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data. The first marine transport liquefied natural gas unloading composition prediction model is trained using an artificial neural network. The liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data. The second marine transport liquefied natural gas unloading composition prediction model is trained using a machine learning algorithm. The prediction performances of the first marine transport liquefied natural gas unloading composition prediction model and the second marine transport liquefied natural gas unloading composition prediction model are compared to obtain a final marine transport liquefied natural gas unloading composition prediction model.
6. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to any one of claims 1 to 4, characterized in that: The prediction model for the unloading composition of liquefied natural gas transported by sea includes an input layer, a hidden layer and an output layer. The liquefied natural gas composition data and the liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data. The prediction model for the unloading composition of liquefied natural gas transported by sea is trained, including: Receive sample data and label data through the input layer, and convert the sample data and label data into sample vectors and label vectors; The activation function is used in the hidden layer to make the marine liquefied natural gas unloading composition prediction model learn the mapping relationship between the sample vector and the label vector; The prediction result of the sample vector corresponding to the label vector is output through the output layer.
7. The method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea according to any one of claims 1 to 4, characterized in that: The liquefied natural gas composition data and liquefied natural gas component influencing factor data of the loading port are used as sample data, and the liquefied natural gas composition data of the unloading port is used as label data to train a prediction model for the unloading composition of liquefied natural gas transported by sea, including: According to the prediction results of the sample data by the marine transport liquefied natural gas unloading composition prediction model and the label data corresponding to the sample data, the prediction deviation is obtained, and the prediction deviation is corrected using the second expression, wherein the second expression is: , In the second expression, Indicates The deviation of the eigenvector after correction, Indicates feature vectors, Indicates The predicted deviation between the predicted value and the label value of the feature vector; and / or, In the process of training the prediction model for the unloading composition of liquefied natural gas transported at sea, the SHAP value method is used to analyze and quantify the contribution of each feature in the sample data to the prediction of the label data.
8. A method for predicting the unloading composition of liquefied natural gas transported by sea, characterized in that: include: Receiving liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and liquefied natural gas component influence factor data corresponding to the liquefied natural gas component influence factor screened by the method for constructing a marine transport liquefied natural gas unloading composition prediction model according to any one of claims 1 to 7; According to the liquefied natural gas composition data of the liquefied natural gas to be tested at the loading port and the data of the influencing factors of the liquefied natural gas components, the liquefied natural gas unloading composition prediction model for sea transport is obtained by the method for constructing the sea transport liquefied natural gas unloading composition prediction model described in any one of claims 1 to 7, and the liquefied natural gas composition data of the liquefied natural gas to be tested at the unloading port is predicted.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, it implements the method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea as described in any one of claims 1 to 7 and / or the method for predicting the unloading composition of liquefied natural gas transported by sea as described in claim 8.
10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it implements the method for constructing a prediction model for the unloading composition of liquefied natural gas transported by sea as described in any one of claims 1 to 7 and / or the method for predicting the unloading composition of liquefied natural gas transported by sea as described in claim 8.