Tight gas gathering and transportation pipeline corrosion prediction method, system, equipment and medium
By randomly combining influencing factors and optimizing them using neural networks and data augmentation models, the problem of failing to effectively consider SRB content and the combined effect of multiple factors in existing technologies has been solved, achieving more accurate corrosion prediction for tight gas gathering and transportation pipelines.
Patent Information
- Application Number
- CN202411072048.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-06
- Publication Date
- 2026-02-06
Smart Images
Figure CN121483433A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the technical field of pipeline corrosion prediction, and particularly relates to a tight gas gathering pipeline corrosion prediction method, system, device and medium. BACKGROUND
[0002] Unconventional natural gas as a kind of new energy has the characteristics of clean and green. It is widely distributed and has abundant reserves. With the gradual increase of unconventional natural gas development intensity, single well production and construction scale are increasing, and corrosion problems of gathering pipelines are also increasingly prominent. Tight gas fields usually use gas-liquid mixed transportation process, which can cause a large amount of condensate oil and water in the gathering pipeline, forming an oil-gas-water three-phase coexistence system. At the same time, the use of fracturing fluid in the production process leads to the existence of SRB in the environment. Therefore, the corrosion environment of tight gas gathering pipeline is complex, and there are many factors affecting the gathering pipeline, and each factor has mutual influence. At present, there are many studies on the corrosion rate prediction model of CO2 and H2S, but few studies consider the comprehensive effect of SRB content and different factors on the corrosion degree.
[0003] Therefore, the present application aims to provide a tight gas gathering pipeline corrosion prediction method, system, device and medium to solve the above-mentioned related problems. SUMMARY
[0004] The technical problem to be solved by the present application is that the existing technology does not consider the influence of SRB content and the comprehensive effect of different factors on the corrosion degree. The purpose is to provide a tight gas gathering pipeline corrosion prediction method, system, device and medium. By randomly arranging and combining multiple influencing factors, multiple different dimension and different influencing factor combinations are obtained to predict the influence of the comprehensive action of multiple influencing factors. Then, the multiple influencing factor combinations are input into the neural network model for training and testing. The root mean square error of multiple pipeline corrosion speed prediction models is output by using the test results. By selecting the influencing factor combination input into the prediction model with the smallest root mean square error, the correlation between the influencing factor and the pipeline corrosion is improved, and the prediction accuracy is improved. At the same time, the data enhancement model is used to enhance the selected influencing factor combination, and the enhanced influencing factor combination is used to iteratively optimize the model, thereby further improving the accuracy of the prediction model.
[0005] The present application is realized by the following technical scheme:
[0006] A tight gas gathering pipeline corrosion prediction method, the method comprising:
[0007] Obtaining a plurality of influence factor combinations of different dimensions, constructing a training set and a test set based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by random permutation and combination of a plurality of influence factors, and each influence factor contains at least one influence data;
[0008] Training the plurality of influence factor combinations in the training set by inputting them into a preset neural network model respectively to obtain a plurality of pipeline corrosion speed prediction models;
[0009] Inputting the plurality of influence factor combinations in the test set into the corresponding pipeline corrosion speed prediction models respectively, outputting a plurality of prediction results, and obtaining the root mean square error of the plurality of pipeline corrosion speed prediction models based on the prediction results;
[0010] Extracting the influence factor combination input into the pipeline corrosion speed prediction model with the smallest root mean square error in the training set, inputting the influence factor combination into a preset data enhancement model for data enhancement processing to obtain a data-enhanced influence factor combination;
[0011] Using the data-enhanced influence factor combination to iteratively optimize and train the pipeline corrosion speed prediction model to obtain an optimal pipeline corrosion speed prediction model, and using the optimal pipeline corrosion speed prediction model to predict the pipeline corrosion degree.
[0012] Further, the preset neural network model adopts an ANN artificial neural network model.
[0013] Further, the preset data enhancement model adopts a CGAN generative adversarial network model.
[0014] Further, the plurality of influence factor combinations of different dimensions are obtained based on the plurality of influence factor combinations, and the plurality of influence factor combinations are obtained by random permutation and combination of a plurality of influence factors, specifically as follows:
[0015] Obtaining a plurality of influence factors, obtaining a plurality of influence factor combinations of different dimensions by random permutation and combination, and constructing a training set and a test set based on the plurality of influence factor combinations; wherein the dimension refers to the number of influence factors in the influence factor combination.
[0016] Further, based on the data-enhanced influence factor combination, the corresponding pipeline corrosion speed prediction model is inputted for iteratively optimized training to obtain an optimal pipeline corrosion speed prediction model, and the optimal pipeline corrosion speed prediction model is used to predict the pipeline corrosion degree, specifically as follows:
[0017] Dividing the data-enhanced influence factor combination into an optimization data set and an optimization validation set;
[0018] The optimized data set is input into a corresponding pipeline corrosion rate prediction model for iterative optimization training until the prediction error between the output prediction result and the optimized verification set is less than a preset error threshold, that is, the optimal pipeline corrosion rate prediction model is obtained, and the optimal pipeline corrosion rate prediction model is used to predict the pipeline corrosion degree.
[0019] Further, the influencing factors include pipeline material, CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and real internal corrosion rate value of the pipeline at the current position.
[0020] Further, the dimension of the combination of influencing factors includes 6, 7, 8, 9, 10 and 11.
[0021] The application also provides a dense gas gathering and transportation pipeline corrosion prediction system, characterized in that the system comprises:
[0022] A data set construction module is configured to obtain a plurality of combinations of influencing factors with different dimensions, and construct a training set and a test set based on the plurality of combinations of influencing factors, wherein the plurality of combinations of influencing factors are obtained by random arrangement and combination of a plurality of influencing factors, and each influencing factor contains at least one influencing data;
[0023] A model training module is configured to train a plurality of pipeline corrosion rate prediction models by inputting the plurality of combinations of influencing factors in the training set into a preset neural network model respectively;
[0024] A model testing module is configured to input the plurality of combinations of influencing factors in the test set into corresponding pipeline corrosion rate prediction models respectively, output a plurality of prediction results, and obtain the root mean square error of the plurality of pipeline corrosion rate prediction models based on the prediction results;
[0025] A data enhancement module is configured to extract the combination of influencing factors in the training set input into the pipeline corrosion rate prediction model with the smallest root mean square error, input the combination of influencing factors into a preset data enhancement model for data enhancement processing, and obtain the combination of influencing factors after data enhancement;
[0026] A model optimization module is configured to use the combination of influencing factors after data enhancement to iteratively optimize and train the pipeline corrosion rate prediction model, and obtain the optimal pipeline corrosion rate prediction model;
[0027] A corrosion prediction module is configured to use the optimal pipeline corrosion rate prediction model to predict the pipeline corrosion degree.
[0028] Further, the preset neural network model adopts an ANN artificial neural network model.
[0029] Further, the preset data enhancement model adopts a CGAN (Conditional Generative Adversarial Network) to generate an adversarial network model.
[0030] Further, a plurality of influence factor combinations of different dimensions are obtained, and a training set and a test set are constructed based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by random permutation and combination of a plurality of influence factors, and specifically, the plurality of influence factor combinations are obtained by random permutation and combination of the plurality of influence factors.
[0031] A plurality of influence factors are obtained, a plurality of influence factor combinations of different dimensions are obtained by random permutation and combination, and a training set and a test set are constructed based on the plurality of influence factor combinations; wherein the dimension refers to the number of influence factors in the influence factor combination.
[0032] Further, the influence factor combination after data enhancement is input into a corresponding pipeline corrosion speed prediction model for iterative optimization training to obtain an optimal pipeline corrosion speed prediction model, and the optimal pipeline corrosion speed prediction model is used to predict the pipeline corrosion degree.
[0033] The influence factor combination after data enhancement is divided into an optimization data set and an optimization verification set.
[0034] The optimization data set is input into a corresponding pipeline corrosion speed prediction model for iterative optimization training until the prediction error between the output prediction result and the optimization verification set is less than a preset error threshold, and an optimal pipeline corrosion speed prediction model is obtained, and the optimal pipeline corrosion speed prediction model is used to predict the pipeline corrosion degree.
[0035] Further, the influence factors include pipeline material, CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and real internal corrosion rate value of the pipeline at the current position.
[0036] Further, the dimension of the influence factor combination includes 6, 7, 8, 9, 10 and 11.
[0037] The application also provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and the processor implements the steps of the method according to any one of the above embodiments when executing the computer program.
[0038] The application also provides a computer readable storage medium, which stores a computer program, and the computer program implements the steps of the method according to any one of the above embodiments when executed by a processor.
[0039] The application also provides a computer program product comprising instructions, which, when executed by a computer device cluster, cause the computer device cluster to perform the method according to any one of the above embodiments.
[0040] Compared with the prior art, the present application has the following advantages and beneficial effects:
[0041] In the present application, by randomly arranging and combining multiple influencing factors, multiple different dimension and different influencing factor combinations are obtained to achieve the prediction of the influence generated by the comprehensive action of multiple influencing factors; the multiple influencing factor combinations are input into the neural network model for training and testing, the root mean square error of multiple pipeline corrosion speed prediction models is output by using the test results, the influencing factor combination input into the prediction model with the minimum root mean square error is selected to improve the correlation between the influencing factor and the pipeline corrosion and improve the prediction accuracy; at the same time, the data enhancement model is used to enhance the selected influencing factor combination, and the enhanced influencing factor combination is used to iteratively optimize the model, thereby further improving the accuracy of the prediction model. BRIEF DESCRIPTION OF DRAWINGS
[0042] In order to more clearly illustrate the technical solutions of the example embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some embodiments of the present application, and therefore should not be regarded as a limitation on the scope, and for those skilled in the art, other related drawings can also be obtained without creative labor. In the drawings:
[0043] Figure 1 The method flow chart of the dense gas gathering pipeline corrosion prediction method in the present embodiment;
[0044] Figure 2 The system module diagram of the dense gas gathering pipeline corrosion prediction system in the present embodiment;
[0045] Figure 3 The flowchart of the CGAN generated adversarial network model enhancing influence data in the dense gas gathering pipeline corrosion prediction method in the present embodiment;
[0046] Figure 4 The structure diagram of the CGAN generated adversarial network model in the dense gas gathering pipeline corrosion prediction method in the present embodiment;
[0047] Figure 5 The structure diagram of the ANN neural network in the dense gas gathering pipeline corrosion prediction method in the present embodiment;
[0048] Figure 6 The structure diagram of the computer device in the present embodiment. DETAILED DESCRIPTION
[0049] Exemplary embodiments of the present disclosure are described herein below with reference to the accompanying drawings, in which various details are set forth to facilitate an understanding of the present disclosure. It should be apparent to those skilled in the art that various changes and modifications can be made to the embodiments described herein without departing from the scope of the present disclosure. Also, descriptions of well-known functions and constructions are omitted for clarity and conciseness.
[0050] In the present disclosure, the terms "first", "second", and the like are used to describe various elements only and do not intend to limit the positional relationship, the time relationship, or the importance of the elements, and the terms are used only to distinguish one element from another element. In some examples, the first element and the second element can refer to the same instance of the element, and in some cases, based on the context of the description, they can also refer to different instances.
[0051] The terms used in the description of various examples in the present disclosure are only for the purpose of describing the specific examples and are not intended to be limiting. Unless the number of elements is specifically limited, the element can be one or more than one, if the number of elements is not specifically limited, unless the context clearly indicates otherwise. In addition, the term "and / or" used in the present disclosure encompasses any one of the listed items and all possible combinations thereof.
[0052] In order to facilitate the understanding of the embodiments of the present application, first, some terms related to the present application are explained and described.
[0053] Tight gas, also known as tight sandstone gas, refers to natural gas in sandstone formations with a permeability of less than 0.1 md (or less than 0.5 mD, the specific value varies depending on different definitions).
[0054] A gathering pipeline is a long-distance pipeline for transporting different types of gaseous and liquid substances, which is usually made of steel or polyethylene pipe and has the ability to withstand high pressure. In the process of oil and gas production and processing, gathering pipelines are widely used to transport oil, natural gas, ethylene, propylene, and other chemical products, and liquefied natural gas.
[0055] ANN artificial neural network model is a complex network composed of a large number of simple elements (i.e. neurons) connected to each other, with high nonlinearity, capable of complex logical operations and nonlinear relationship implementation.
[0056] CGAN (Conditional Generative Adversarial Network) is a variant of GAN (Generative Adversarial Network) that introduces conditional information into the generation process, allowing the generated data to meet specific conditions or requirements. The CGAN model consists of two main parts: the Generator and the Discriminator.
[0057] The Root Mean Squared Error (RMSE) is a commonly used indicator to measure the difference between the predicted value and the true value of the model. In practical applications, it can also be understood as the square root of the square of the deviation between the predicted value and the true value divided by the number of observations n. RMSE can better reflect the precision of measurement and is an important indicator to measure the prediction accuracy of the model.
[0058] Example 1
[0059] In combination Figure 1 As shown in the embodiment, a method for predicting corrosion of a dense gas gathering pipeline is disclosed, the method comprising:
[0060] S1: Obtain a plurality of different dimensionality influence factor combinations, and construct a training set and a test set based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by randomly arranging and combining a plurality of influence factors, and each influence factor contains at least one influence data, specifically:
[0061] Obtain a plurality of influence factors, obtain a plurality of different dimensionality influence factor combinations by randomly arranging and combining the plurality of influence factors, and construct a training set and a test set based on the plurality of influence factor combinations; wherein the dimensionality refers to the number of influence factors in the influence factor combination, and each influence factor contains at least one influence data;
[0062] It should be noted that in the embodiment, the number of influencing factors is 12, which specifically includes pipeline material, CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and the real internal corrosion rate value of the pipeline at the current position; at the same time, according to the principle that the number of influencing factors in the predicted corrosion rate model is greater than half of the number of all influencing factors, the dimension of the influencing factor combination includes 6 dimensions, 7 dimensions, 8 dimensions, 9 dimensions, 10 dimensions and 11 dimensions; first, the influencing factors are classified according to the pipeline material, and after classification, there are 11 influencing factors left, and then the 11 influencing factors are randomly combined, so that the influencing factor combination with a dimension of 6 has 462 groups, the influencing factor combination with a dimension of 7 has 330 groups, the influencing factor combination with a dimension of 8 has 165 groups, the influencing factor combination with a dimension of 9 has 55 groups, the influencing factor combination with a dimension of 10 has 11 groups, and the influencing factor combination with a dimension of 11 has 1 group, and a total of 1024 groups.
[0063] For example, 50 groups of influence data of dense gas gathering and transportation pipelines of L360 pipe material are collected, the 50 groups of influence data are classified according to the influencing factors to obtain 12 influencing factors, then the 12 influencing factors are randomly arranged and combined according to the above dimension to obtain 1024 groups of influencing factor combinations, and finally the 1024 groups of influencing factor combinations are divided into a training set and a test set, and the ratio of the data amount of the training set to the test set is 9:1.
[0064] S2: training is performed by inputting the multiple influencing factor combinations in the training set into the ANN artificial neural network model respectively to obtain multiple pipeline corrosion rate prediction models;
[0065] Specifically, in the embodiment, the relationship of the constructed ANN artificial neural network model is as follows:
[0066]
[0067] In the formula, f is an activation function, w is a weight matrix, x is an input vector, and b is a bias term. i i
[0068] For example, the above 1024 groups of influencing factor combinations are input into the ANN artificial neural network model for training to obtain 1024 pipeline corrosion rate prediction models.
[0069] S3: inputting the multiple influencing factor combinations in the test set into the corresponding pipeline corrosion rate prediction model respectively, outputting multiple prediction results, and obtaining the root mean square error of the multiple pipeline corrosion rate prediction models based on the prediction results;
[0070] It should be noted that the combination of factors in the test set input into each pipeline corrosion rate prediction model corresponds one-to-one to the combination of factors in the training set, for example, the combination of factors in the training set for training the pipeline corrosion rate prediction model is 6-dimensional, specifically including CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, and temperature, and the combination of factors in the test set for testing the pipeline corrosion rate prediction model should also be 6-dimensional, specifically including CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, and temperature.
[0071] Specifically, in the present embodiment, the plurality of combinations of factors in the test set are input into the corresponding pipeline corrosion rate prediction model, and the plurality of prediction results are output, and the root mean square error RMSE is used to judge the error between the prediction result and the test set. Each pipeline corrosion rate prediction model will obtain a corresponding root mean square error, and the root mean square error RMSE relationship used is as follows:
[0072]
[0073] In the formula, N is the number of samples, y i is the actual value of the test set, is the predicted value of the prediction result;
[0074] For example, the pipeline corrosion rate prediction model obtained by the 8-dimensional combination of factors including CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, temperature, water content, and gas-liquid ratio has the smallest root mean square error between the predicted corrosion rate and the actual value.
[0075] S4: Extract the combination of factors in the training set input into the pipeline corrosion rate prediction model with the smallest root mean square error, and input it into the CGAN generative adversarial network model for data enhancement processing to obtain the data-enhanced combination of factors;
[0076] Specifically, in the present embodiment, referring to Figures 3-4 , the CGAN generative adversarial network model is used to generate influence data similar to real data, and the structure of the CGAN generative adversarial network model used is specifically as follows:
[0077]
[0078] In the formula, x is the real training sample input into the discriminator network D, z is random noise conforming to a Gaussian distribution, and y is the input condition shared by the generator network G and the discriminator network D;
[0079] Exemplarily, the influence data in the 8-dimensional influence factor combination is trained by the CGAN generated adversarial network model, 30 groups of influence data are added on the basis of the original 50 groups of influence data, and the influence data is considered as real data, so that the influence factor combination after data enhancement is obtained;
[0080] S5: inputting the influence factor combination after data enhancement into the corresponding pipeline corrosion rate prediction model for iterative optimization training to obtain an optimal pipeline corrosion rate prediction model, and using the optimal pipeline corrosion rate prediction model to predict the pipeline corrosion degree, specifically:
[0081] The influence factor combination after data enhancement is divided into an optimization data set and an optimization verification set; the optimization data set is input into the corresponding pipeline corrosion rate prediction model for iterative optimization training until the prediction error between the output prediction result and the optimization verification set is less than a preset error threshold, that is, the optimal pipeline corrosion rate prediction model is obtained, and the optimal pipeline corrosion rate prediction model is used to predict the pipeline corrosion degree.
[0082] It should be noted that in the embodiment, the influence factor combination after data enhancement is divided into an optimization data set and an optimization verification set according to a ratio of 9:1, the optimization data set is input into the corresponding pipeline corrosion rate prediction model for iterative optimization training, the root mean square difference between the prediction result output by the pipeline corrosion rate prediction model and the actual value of the optimization verification set is used to judge the error size, whether the error is less than the preset error threshold is judged, if not satisfied, the training is returned, if satisfied, the trained ANN model is output, that is, the optimal pipeline corrosion rate prediction model is obtained.
[0083] Exemplarily, 80 groups of influence data are randomly allocated as 72 groups of training set and 8 groups of test set; the influence data in the 72 groups of training set is used to iteratively optimize and train the corresponding pipeline corrosion rate prediction model, and the root mean square difference between the prediction result and the actual value is calculated; if the root mean square difference satisfies the preset value, the optimal pipeline corrosion rate prediction model is obtained, if the root mean square difference does not satisfy the preset value, the training is returned, until the root mean square difference satisfies the preset value.
[0084] The application also provides a compact gas gathering pipeline corrosion prediction system, characterized in that the system comprises:
[0085] The data set construction module 100 is used for acquiring a plurality of influence factor combinations with different dimensions, and constructing a training set and a test set based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by randomly arranging and combining a plurality of influence factors;
[0086] The model training module 200 is configured to train the plurality of pipeline corrosion rate prediction models by inputting the plurality of influence factor combinations in the training set into the preset neural network model respectively.
[0087] The model testing module 300 is configured to input the plurality of influence factor combinations in the test set into the corresponding pipeline corrosion rate prediction model respectively, output a plurality of prediction results, and obtain the root mean square error of the plurality of pipeline corrosion rate prediction models based on the prediction results.
[0088] The data enhancement module 400 is configured to extract the influence factor combination in the training set input into the pipeline corrosion rate prediction model with the minimum root mean square error, input the influence factor combination into the preset data enhancement model for data enhancement processing, and obtain the data-enhanced influence factor combination.
[0089] The model optimization module 500 is configured to input the data-enhanced influence factor combination into the corresponding pipeline corrosion rate prediction model for iterative optimization training to obtain the optimal pipeline corrosion rate prediction model.
[0090] The corrosion prediction module 600 is configured to predict the pipeline corrosion degree by using the optimal pipeline corrosion rate prediction model.
[0091] Further, the preset neural network model adopts an ANN artificial neural network model.
[0092] Further, the preset data enhancement model adopts a CGAN generative adversarial network model.
[0093] Further, a plurality of influence factor combinations of different dimensions are obtained, and the training set and the test set are constructed based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by randomly arranging and combining a plurality of influence factors, and specifically:
[0094] A plurality of influence factors are obtained, a plurality of influence factor combinations of different dimensions are obtained by randomly arranging and combining the plurality of influence factors, and the training set and the test set are constructed based on the plurality of influence factor combinations; wherein the dimension refers to the number of influence factors in the influence factor combination.
[0095] Further, the data-enhanced influence factor combination is divided into an optimization data set and an optimization verification set.
[0096] Further, the data-enhanced influence factor combination is divided into an optimization data set and an optimization verification set.
[0097] The optimized data set is input into a corresponding pipeline corrosion rate prediction model for iterative optimization training until the prediction error between the output prediction result and the optimized verification set is less than a preset error threshold, that is, an optimal pipeline corrosion rate prediction model is obtained, and the optimal pipeline corrosion rate prediction model is used to predict the pipeline corrosion degree.
[0098] Further, the influencing factors include pipeline material, CO2 partial pressure, H2S partial pressure, SRB number, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and real internal corrosion rate value of the pipeline at the current position.
[0099] Further, the dimension of the influencing factor combination includes 6, 7, 8, 9, 10, and 11.
[0100] It should be noted that the modules in the system of Embodiment 2 correspond to the steps in the method of Embodiment 1, and the steps in the method of Embodiment 1 have been described in detail in Embodiment 1, and the content of the modules in the system will not be described in detail in Embodiment 2.
[0101] Embodiment 3
[0102] The embodiment also provides a computer device including a memory 1005 and a processor 1001, the memory 1005 stores a computer program, and the processor 1001 implements the steps of the method of any one of the above embodiments when executing the computer program.
[0103] It should be noted that the processor 1001 is configured to execute the steps in the method embodiments described above according to instructions in the program code. Alternatively, the processor 1001 implements the functions of each module / unit in each system / device embodiment described above when executing the computer program.
[0104] Specifically, in the embodiment, the computer program can be divided into one or more modules / units, and the one or more modules / units are stored in the memory 1005 and executed by the processor 1001 to complete the present application. The one or more modules / units can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the terminal device.
[0105] The terminal device can be a desktop computer, a notebook computer, a palm computer, and a cloud server, etc. The terminal device can include, but is not limited to, the processor 1001, the memory 1005. Those skilled in the art can understand that it does not constitute a limitation on the terminal device, and can include more or fewer components than the illustration, or combine certain components, or different components, for example, the terminal device can also include an input / output device 1003, a network access device 1002, a bus 1006, etc.
[0106] The processor 1001 can be a central processing unit (CPU), and can also be other general-purpose processors 1001, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. The general-purpose processor 1001 can be a microprocessor or the processor can also be any conventional processor.
[0107] The memory 1005 can be an internal storage unit of the terminal device, such as a hard disk or a memory of the terminal device. The memory 1005 can also be an external storage device 1004 of the terminal device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, or the like. Further, the memory 1005 can include both an internal storage unit and an external storage device 1004 of the terminal device. The memory 1005 is used to store computer programs and other programs and data required by the terminal device. The memory 1005 can also be used to temporarily store data that has been output or will be output.
[0108] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, system and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described herein.
[0109] Embodiment 4
[0110] The embodiment provides a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program is executed by a processor to implement the steps of the method in any one of the above embodiments.
[0111] The computer readable storage medium, for example, can be, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium include: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), registers, a hard disk, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. A computer readable storage medium can be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
[0112] An exemplary storage medium is coupled to the processor such that the processor can read information from, and write information to, the storage medium. Of course, the storage medium can be a part of the processor. Consistent with the teachings of the present disclosure, a storage medium can be implemented using any appropriate media, including but not limited to optical, magnetic or semiconductor technologies, or any suitable combination of the foregoing. In the context of this document, a "computer readable storage medium" can be any tangible medium that can contain, or store programming for use by or in connection with an instruction execution system, apparatus, or device.
[0113] The specific implementation described above is illustrative of specific embodiments of the present application. However, various modifications can be made without departing from the spirit and scope of the present application. For example, the above-described embodiments can be implemented in a variety of contexts. Although the present application has been described in relation to particular embodiments, many details of construction and operation will be apparent to those skilled in the art in light of this disclosure, numerous other embodiments will also be apparent and can be made without departing from the spirit and scope of the application.
Claims
1. A method for predicting corrosion in tight gas gathering and transportation pipelines, characterized in that the method... The method comprises the following steps: obtaining a plurality of influence factor combinations of different dimensions, and constructing a training set and a test set based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by randomly arranging and combining a plurality of influence factors, and each influence factor comprises at least one influence data; training the plurality of influence factor combinations in the training set by inputting them into a preset neural network model to obtain a plurality of pipeline corrosion rate prediction models; inputting the plurality of influence factor combinations in the test set into corresponding pipeline corrosion rate prediction models to output a plurality of prediction results, and obtaining the root mean square error of the plurality of pipeline corrosion rate prediction models based on the prediction results; extracting the influence factor combination input into the pipeline corrosion rate prediction model with the minimum root mean square error in the training set, inputting the influence factor combination into a preset data enhancement model for data enhancement processing to obtain an enhanced influence factor combination; iteratively optimizing and training the pipeline corrosion rate prediction model by using the enhanced influence factor combination to obtain an optimal pipeline corrosion rate prediction model, and predicting the pipeline corrosion degree by using the optimal pipeline corrosion rate prediction model.
2. The method of claim 1, wherein the method further comprises: The preset neural network model adopts an ANN artificial neural network model.
3. The method of claim 1, wherein the method further comprises: The preset data enhancement model adopts a CGAN generative adversarial network model.
4. The method of claim 1, wherein the method further comprises: The method comprises the following steps: obtaining a plurality of influence factors, and obtaining a plurality of influence factor combinations of different dimensions by randomly arranging and combining the plurality of influence factors, and constructing a training set and a test set based on the plurality of influence factor combinations; wherein the dimension refers to the number of influence factors in the influence factor combination.
5. The method of claim 1, wherein the method further comprises: The method comprises the following steps: dividing the enhanced influence factor combination into an optimization data set and an optimization verification set; inputting the optimization data set into the corresponding pipeline corrosion rate prediction model for iterative optimization training until the prediction error between the output prediction result and the optimization verification set is less than a preset error threshold, thereby obtaining the optimal pipeline corrosion rate prediction model, and predicting the pipeline corrosion degree by using the optimal pipeline corrosion rate prediction model.
6. The method of claim 4, wherein the method further comprises: The influence factors include pipeline material, CO2 partial pressure, H2S partial pressure, SRB quantity, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and real internal corrosion rate value of the pipeline at the current position.
7. The method of claim 4, wherein the method further comprises: The dimension of the influence factor combination includes 6, 7, 8, 9, 10, and 11.
8. A system for corrosion prediction in a dense phase gas gathering pipeline, the system comprising: The method comprises the following steps: a data set construction module is configured to obtain a plurality of influence factor combinations of different dimensions, and construct a training set and a test set based on the plurality of influence factor combinations, wherein the plurality of influence factor combinations are obtained by randomly arranging and combining a plurality of influence factors, and each influence factor comprises at least one influence data; The model training module is configured to train the preset neural network model by inputting multiple influence factor combinations in the training set into the preset neural network model respectively, to obtain multiple pipeline corrosion speed prediction models. The model testing module is configured to input multiple influence factor combinations in the test set into corresponding pipeline corrosion speed prediction models respectively, to output multiple prediction results, and obtain root mean square errors of the multiple pipeline corrosion speed prediction models based on the prediction results. The data enhancement module is configured to extract an influence factor combination input into the pipeline corrosion speed prediction model with the minimum root mean square error in the training set, input the influence factor combination into a preset data enhancement model for data enhancement processing, and obtain an enhanced influence factor combination. The model optimization module is configured to perform iterative optimization training on the pipeline corrosion speed prediction model by using the enhanced influence factor combination, to obtain an optimal pipeline corrosion speed prediction model. The corrosion prediction module is configured to predict the pipeline corrosion degree by using the optimal pipeline corrosion speed prediction model.
9. A system for corrosion prediction in a dense phase gas gathering pipeline according to claim 8, wherein, The preset neural network model is an ANN artificial neural network model.
10. The system for corrosion prediction of a gathering pipeline for a tight gas according to claim 8, wherein The preset data enhancement model is a CGAN generative adversarial network model.
11. A system for corrosion prediction in a dense phase gas gathering pipeline according to claim 8, wherein, Multiple influence factor combinations of different dimensions are obtained, and a training set and a test set are constructed based on the multiple influence factor combinations. Multiple influence factors are obtained, multiple influence factor combinations of different dimensions are obtained by random arrangement and combination of the multiple influence factors, and a training set and a test set are constructed based on the multiple influence factor combinations.
12. A system for corrosion prediction in a dense phase gas gathering pipeline according to claim 8, wherein, Based on the enhanced influence factor combination, iterative optimization training is performed on the corresponding pipeline corrosion speed prediction model to obtain an optimal pipeline corrosion speed prediction model, and the optimal pipeline corrosion speed prediction model is used to predict the pipeline corrosion degree. The enhanced influence factor combination is divided into an optimization data set and an optimization verification set. The optimization data set is input into the corresponding pipeline corrosion speed prediction model for iterative optimization training until the prediction error between the output prediction result and the optimization verification set is less than a preset error threshold, and the optimal pipeline corrosion speed prediction model is obtained.
13. A system for corrosion prediction in a dense phase gas gathering pipeline according to claim 11, wherein, The influence factors include pipeline material, CO2 partial pressure, H2S partial pressure, SRB quantity, pH, liquid flow rate, gas flow rate, temperature, pressure, water content, gas-liquid ratio, and real internal corrosion rate value of the pipeline at the current position.
14. The system for corrosion prediction of a gathering pipeline for a tight gas according to claim 11, wherein The dimensions of the influence factor combinations include 6, 7, 8, 9, 10, and 11.
15. A computer device comprising a memory and a processor, the memory storing a computer program, characterized in that, The processor executes the computer program to implement the steps of the method of any one of claims 1 to 7.
16. A computer readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the steps of the method of any one of claims 1 to 7.
17. A computer program product comprising instructions, characterized in that, When the instructions are executed by the computer device cluster, the computer device cluster is caused to perform the method of any one of claims 1 to 7.