Oil gas processing decision method and system based on big data

Abstract

The invention provides an oil gas processing method and system based on big data, and the method comprises the steps: determining an adopted processing mode through judging whether a target parameter of a certain processing object meets a parameter standard or not, and if yes, taking the processing mode as an implementation processing mode; if not or a processing mode switching intention is received, whether the processing objects and the processing modes are the same is judged, and the execution frequency is recognized, and if not, a first processing mode is selected according to the key target parameters to carry out oil gas processing operation, and the frequency is updated; if yes, selecting a first processing mode for operation if the number of times meets the requirement; and if yes, selecting a second processing mode to carry out oil-gas processing operation by utilizing the target model according to the related parameters. By adopting the scheme, the problems that flexible adjustment is difficult and application is limited in the prior art can be solved, data of different processing modes can be mutually learned and mutually assisted by classified planning processing related data, the processing mode is optimized through the big data technology, and the processing efficiency is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mechanical energy processing and optimization, and particularly to an oil and gas processing method and system based on big data. Background Art

[0002] To achieve circular economy, meet the requirements of energy conservation and emission reduction, and follow the path of sustainable development and low-carbon economy is a major technical problem in the energy and chemical industries. Most chemical and petrochemical enterprises build the scale of air separation gas, syngas, steam and other media and energy according to their own needs. They rely on external purchases for electricity and water. The cost of treating "three wastes" separately is high, and it is difficult to meet the emission and environmental protection requirements. They need to pay high environmental protection fees, which affects the development of the enterprises themselves. In addition, due to the lack of economies of scale in the production of energy media, the energy utilization rate is not high, and the production stability of the enterprises themselves cannot be guaranteed.

[0003] Current oil and gas processing methods are all traditional mechanical fixed processing methods, which are difficult to be adjusted and optimized specifically for different data processing requirements, and cannot quickly adapt to the changes of the processing objects during the processing process. In addition, there are also some solutions in the prior art that use environmental protection big data technology to explore the deep internal relationship between the process control parameters of sulfur recovery units and the concentration of tail gas SO2, use data analysis technology for data extraction, transformation, analysis and specific model operation processing, extract the strong operating variables that affect the SO2 emission concentration of the tail gas of sulfur recovery units and the key data for assisting production decision-making from the parameters, and achieve the purpose of ensuring the stable compliance of the SO2 tail gas of sulfur recovery units and forming an overall treatment plan for SO2 compliance emission under the current working conditions without adding new desulfurization facilities. However, these methods have poor adaptability to the changing processing environment, insufficient processing flexibility, and poor operability.

[0004] Therefore, how to process oil and gas specifically based on this rich resource of big data, and provide an operable processing method that can quickly adapt to the oil and gas processing environment is a problem to be solved.

[0005] The information disclosed in the background art part of the present invention is only intended to deepen the understanding of the general background art of the present invention, and should not be regarded as an admission or any form of implication that this information constitutes the prior art known to those skilled in the art. Summary of the Invention

[0006] To solve the above problems, the present invention provides an oil and gas processing method based on big data. This method determines the processing method to be adopted by judging whether the target parameters of a certain processing object meet the parameter standards. If so, this processing method is used as the implemented processing method; if not or a processing method switching intention is received, it is judged whether it is the same processing object and the same processing method, and the execution times are identified. If not, the first processing method is selected based on the key target parameters for the oil and gas processing operation, and the times are updated; if so, when the times meet the requirements, the first processing method is selected for operation; if so, when the times do not meet the conditions, the second processing method is selected for the oil and gas processing operation according to the relevant parameters using the target model. Adopting this solution can overcome the problems of difficult flexible adjustment and limited application in the prior art, classify and plan the processing-related data so that the data of different processing methods can learn from and assist each other, optimize the processing method through big data technology to achieve high-quality oil and gas processing, and effectively improve the processing efficiency, with better operability and practicability. Preferably, in one embodiment, the method includes:

[0007] Parameter acquisition step S1: Based on a certain oil and gas processing operation, acquire the relevant parameters of the processing object during the processing; among them, include initial parameters, intermediate parameters, and target parameters representing the optimization objectives of the processing method;

[0008] Parameter quality judgment step S2: Based on the current processing object, combined with the preset parameter standards of the corresponding processing method, judge whether the current target parameters meet the corresponding parameter standards. If so, it indicates that the current processing method is qualified, and continue to use this processing method as the implemented processing method; if not or a processing method switching intention is received, then execute step S3;

[0009] Processing method decision step S3: Judge whether it is the same processing object and the same processing method. If so, increment the times value by 1, and judge whether the updated times value is less than the set first number of times. If less, execute step S4; otherwise, execute step S5, where the initial value of the times is 0;

[0010] First processing method selection step S4: Select the key target parameters based on the current processing object, select the corresponding first processing method based on the key target parameters, and enter step S6;

[0011] Second processing method selection step S5: Comprehensively use the relevant parameters corresponding to the current processing object to select the second processing method using the target model, and enter step S6;

[0012] Processing result output step S6: Perform the oil and gas processing operation based on the selected processing method, and update the relevant parameters according to the processing result;

[0013] Among them, the target model is a neural network model obtained by training based on big data through model training step S0.

[0014] Further, in an optional embodiment, in step S2, when it is determined that the target parameter meets the standard conditions, the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment and their corresponding treatment methods are obtained and saved in an associated manner; it is also added as qualified sample data to the big data processing center to guide subsequent or other oil and gas treatments.

[0015] Preferably, in one embodiment, in step S4, one or more target parameters with the farthest distance from the standard conditions in the target parameters are used as key target parameters or key target parameter groups; based on a certain target parameter, the distance between it and the standard conditions is calculated according to the following formula:

[0016]

[0017] where A represents the value of the current target parameter, and U1, U2 represent that the standard conditions of the current target parameter belong to [U1, U2].

[0018] Further, in one embodiment, in step S4, in the process of selecting the corresponding first treatment method based on the key target parameter, it includes:

[0019] If it is a single key target parameter, the treatment method with the strongest processing ability for the current key target parameter is selected as the selected first treatment method;

[0020] If it is a key target parameter group, the treatment methods with the strongest processing ability for each key target parameter in the key target parameter group are respectively selected, and the treatment method with the highest weight is selected as the first treatment method with the number of occurrences as the weight.

[0021] Optionally, in one embodiment, in step S5, when selecting the second treatment method based on the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment, first select a target model with the same intermediate parameter type as the intermediate parameter type of the current treatment object, and then input the initial parameter, intermediate parameter, and target parameter into the selected target model to determine the second treatment method.

[0022] Further, in one embodiment, in step S5, when there is no model with exactly the same intermediate parameter type, one or more target models with the number of intermediate parameter types that are not exactly the same less than the set value are selected.

[0023] Optionally, in one embodiment, when there are multiple second treatment methods obtained by inputting the initial parameter, intermediate parameter, and target parameter into the selected target model, the multiple second treatment methods are sorted based on the degree of coincidence of the intermediate parameter, the number of training times of the model, and the operating cost, so that different processes can select different second treatment methods in order.

[0024] In one embodiment, if the current intermediate parameter type is more than the intermediate parameter type required for input by the model, select the intermediate parameter type required for input by the model from the current intermediate parameter type, and discard the redundant parameter data.

[0025] If the current intermediate parameter type is less than the intermediate parameter type required for input by the model, fill the data of the missing intermediate parameter type required for input with the allowed extreme values.

[0026] In a preferred embodiment, in the model training step, prepare initial parameters, intermediate parameters, target parameters related to oil and gas treatment operations of a set scale and their corresponding processing methods as training sample data, and train the set basic model to obtain a target model; the basic model uses a neural network model, a convolutional neural network, machine learning, or a support vector machine model.

[0027] Based on other aspects of the method described in any one or more of the above embodiments, the present invention also provides a storage medium, on which program code for implementing the method described in any one or more of the above embodiments is stored.

[0028] Based on the application aspect of the method described in any one or more of the above embodiments, the present invention also provides an oil and gas treatment system based on big data, which executes the method described in any one or more of the above embodiments.

[0029] Compared with the closest prior art, the present invention also has the following beneficial effects:

[0030] The present invention provides an oil and gas treatment method and system based on big data. The method first determines whether the target parameter of a certain treatment object meets the parameter standard to determine the adopted treatment method. If so, use this treatment method as the implementation treatment method; if not or a treatment method switching intention is received, then determine whether it is the same treatment object and the same treatment method, and identify the execution times. If not, select the first treatment method for oil and gas treatment operations according to the key target parameters and update the times; if so, select the first treatment method for operations when the times meet the requirements; if so, when the times do not meet the conditions, select the second treatment method for oil and gas treatment operations according to the relevant parameters using the target model. Adopting this solution can facilitate optimizing the treatment method using big data technology, gradually adjust the treatment experience based on big data in the absence of treatment experience, and obtain a treatment method with good treatment effects; in addition, by classifying the data involved in the treatment, the data of different oil and gas treatment methods can learn from each other, greatly improving flexibility, and the training accuracy will increase rapidly with the increase in the data volume, realizing efficient and high-quality oil and gas treatment operations.

[0031] Other features and advantages of the present invention will be set forth in the following description, and in part will be obvious from the description, or can be learned by practicing the present invention. The objectives and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] The drawings are used to provide a further understanding of the present invention, and constitute a part of the description. They are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention. In the drawings:

[0033] Figure 1 is a schematic flowchart of an oil and gas processing method based on big data provided by an embodiment of the present invention;

[0034] Figure 2 is a schematic structural diagram of an oil and gas processing system based on big data provided by another embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] The following will describe in detail the embodiments of the present invention with reference to the drawings and embodiments, so that those skilled in the art of the present invention can fully understand how to apply technical means to solve technical problems and achieve the process of realizing technical effects, and implement the present invention according to the above implementation process. It should be noted that as long as there is no conflict, the various embodiments in the present invention and the various features of each embodiment can be combined with each other, and the technical solutions formed are within the protection scope of the present invention.

[0036] Although the flowchart describes the operations as sequential processing, many of the operations can be performed in parallel, concurrently or simultaneously. The order of the operations can be rearranged. The processing can be terminated when its operations are completed, but there can also be additional steps not included in the drawings. The processing can correspond to a method, function, procedure, subroutine, subprogram, etc.

[0037] The computer device includes a user device and a network device. Among them, the user device or client includes, but is not limited to, a computer, a smart phone, a PDA (Personal Digital Assistant), etc.; the network device includes, but is not limited to, a single network server, a server group composed of multiple network servers, or a cloud composed of a large number of computers or network servers based on cloud computing. The computer device can run alone to implement the present invention, or can be connected to the network and implement the present invention through interaction with other computer devices in the network. The network where the computer device is located includes, but is not limited to, the Internet, a wide area network, a metropolitan area network, a local area network, a VPN network, etc.

[0038] The terms "first", "second", etc. may be used herein to describe various elements, but these elements should not be limited by these terms. These terms are only used to distinguish one element from another. The term "and / or" used herein includes any and all combinations of one or more of the associated listed items. When an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element, or intervening elements may be present.

[0039] The terms used herein are for the purpose of describing particular embodiments only and are not intended to limit the exemplary embodiments. Unless the context clearly dictates otherwise, the singular forms "a", "an" used herein are also intended to include the plural. It should also be understood that the terms "comprises" and / or "comprising" specify the presence of the stated features, integers, steps, operations, elements, and / or components, and do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or combinations thereof.

[0040] Energy has become an increasingly concerned topic in the world today. With the continuous development of the social and economic situation, the coal processing industry has been growing continuously. As a result, the problems of high energy consumption, high pollution, and environmental pressure have become increasingly serious. To achieve circular economy, meet the requirements of energy conservation and emission reduction, and follow the path of sustainable development and low-carbon economy is a major technical problem in the energy and chemical industries. Most chemical and petrochemical enterprises build the scale of air separation gas, syngas, steam and other media and energy according to their own needs. They rely on external purchases for electricity and water, etc. The cost of separately treating "three wastes" is high, and it is difficult to meet the emission and environmental protection requirements. They need to pay high environmental protection fees, which affects the development of the enterprises themselves. In addition, due to the lack of economies of scale in the production of energy media, the energy utilization rate is not high, and the production stability of the enterprises themselves cannot be guaranteed.

[0041] According to research, dozens of flue gas desulfurization process technologies, such as limestone-gypsum wet method, flue gas circulating fluidized bed method, seawater method, desulfurization and dust removal integration, semi-dry method, rotary spray drying method, in-furnace calcium injection with tail-end humidification activation method, electron beam method, etc., have been applied. According to investigations, among the commissioned flue gas desulfurization units, those using the limestone-gypsum wet desulfurization process account for about 92%, the seawater method accounts for about 3%, the flue gas circulating fluidized bed method accounts for about 2%, the ammonia method accounts for about 2%, and others account for about 1%. However, these treatment methods are all traditional mechanical established treatment methods, which are difficult to be adjusted and optimized specifically for different data processing requirements and cannot quickly adapt to the changes of the treatment objects during the treatment process.

[0042] In addition, there are also some solutions in the prior art that use environmental protection big data technology to explore the deep internal relationship between the relevant process control parameters of the sulfur device and the concentration of tail gas SO2. Data extraction, transformation, analysis, and specific model operation processing are carried out using data analysis technology to extract strong operation variables that affect the SO2 emission concentration of the sulfur device tail gas and key data for assisting production decision-making, so as to achieve the purpose of ensuring the stable compliance of the sulfur device tail gas SO2 and forming an overall treatment plan for SO2 compliance emission under the current working conditions without adding new desulfurization facilities. However, these methods have poor adaptability to the changing processing environment, insufficient processing flexibility, and poor operability.

[0043] The researchers of the present invention have found that how to carry out targeted oil and gas processing based on this rich resource of big data and provide an operable processing method that can quickly adapt to the oil and gas processing environment is a problem to be solved.

[0044] In many oil and gas processing operations, the processing objects are large quantities that exist for a long time and have similar qualities. However, due to the speed of knowledge dissemination and equipment, etc., the local processing efficiency cannot be effectively improved. Continuously optimizing the processing method through big data guidance locally can greatly improve the processing efficiency.

[0045] To solve the above problems, the present invention provides an oil and gas processing method and system based on big data. Applying this method can gradually adjust processing experience based on big data without processing experience and determine a processing method with good processing effects; by classifying the data involved in the processing process, different oil and gas processing methods can learn from each other and support each other as optimization, overcoming the defect that the model operation scheme in the prior art only relies on specific local data for learning; inputting different types of parameters into different sub-models and levels significantly improves flexibility, and the accuracy of the training results will increase rapidly as the amount of data increases; in addition, the present invention introduces multiple neural networks for balancing, which can also improve the accuracy of the selection results; a representative parameter configuration method for key target parameters and key target parameter groups is proposed, which greatly simplifies the initial selection complexity on the premise of ensuring the accuracy of the processing operation. Especially, the method based on key target parameter groups has a good effect in solving the problem of multi-objective and multi-model selection.

[0046] Next, the detailed process of the method of the embodiment of the present invention will be described in detail based on the accompanying drawings. The steps shown in the flowchart of the accompanying drawings can be executed in a computer system including a set of computer-executable instructions. Although the logical order of the steps is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than here.

[0047] Embodiment 1

[0048] Figure 1The flowchart of the oil and gas processing method based on big data provided in the first embodiment of the present invention is shown. Referring to Figure 1 it can be seen that the method includes the following steps.

[0049] Parameter acquisition step S1: Based on a certain oil and gas processing operation, acquire the relevant parameters of the processing object during the processing; among them, include initial parameters, intermediate parameters, and target parameters representing the optimization goal of the processing method;

[0050] Parameter quality identification step S2: Based on the current processing object, combined with the preset parameter standard of the corresponding processing method, identify whether the current target parameter meets the corresponding parameter standard. If so, it indicates that the current processing method is qualified, and continue to use this processing method as the implemented processing method; if not or a processing method switching intention is received, execute step S3;

[0051] Processing method decision step S3: Judge whether it is the same processing object and the same processing method. If so, increment the number value by 1, and judge whether the updated number value exceeds the set first number. If it does not exceed, execute step S4; if it exceeds, execute step S5, where the initial value of the number is 0;

[0052] First processing method selection step S4: Select the key target parameter based on the current processing object, and select the corresponding first processing method based on the key target parameter, and enter step S6;

[0053] Second processing method selection step S5: Select the second processing method using the target model by integrating the relevant parameters corresponding to the current processing object, and enter step S6;

[0054] Processing result output step S6: Perform the oil and gas processing operation based on the selected processing method, and update the relevant parameters according to the processing result;

[0055] Among them, the target model is a neural network model obtained by training based on big data through model training step S0.

[0056] The solution of the present invention has the following advantages: (1) Gradually adjust the processing experience based on big data without processing experience and obtain a processing method with good processing effect; (2) Classify the data involved in the processing so that different oil and gas processing methods can learn from each other and spread to each other, which is different from the defect of only relying on local data for learning in the prior art; (3) Input different types of parameters into different sub-models and levels, which greatly improves the flexibility, and the training accuracy will increase rapidly with the increase of the data volume; (4) Introduce multiple neural networks for balancing, and can also improve the accuracy of the selection result. (5) Propose a representation method for key target parameters and key target parameter groups, which greatly simplifies the complexity of the primary selection and ensures a certain accuracy. Especially the method based on the key target parameter group has a good effect in solving the multi-objective and multi-model selection.

[0057] Preferably, in one embodiment, in the parameter acquisition step S1, the entire processing process involves three types of parameters, including: initial parameters, intermediate parameters, and target parameters; where: the initial parameters can be set as the initial detection parameters of the oil and gas processing object before processing; the intermediate parameters are set as the intermediate parameters in the oil and gas processing process related to the oil and gas processing method. For example, the intermediate parameters can include: the processing temperature of the processing equipment, the real-time capacitance measurement value of the key equipment, the sulfur vapor gas temperature, etc.; the target parameters are the general target parameters for the oil and gas processing that are not closely related to the oil and gas processing method. For example, the target parameters can include: output energy consumption, unit output energy consumption, unit energy consumption output, power generation efficiency of oil and gas, and processing pollution index, etc.; where the target parameters reflect the target optimization guiding direction of the processing method.

[0058] For various types of oil and gas processing methods, there are usually a large number of parameters involved. In the embodiment of the present invention, various parameters are classified and planned according to the model selection and the processing process, so as to simplify the parameter types and enable the selection of the processing method based on big data.

[0059] Next, execute step S2 to determine whether the target parameter meets the standard conditions. If not (does not meet), enter step S3; preferably: if so (meets), then continue to adopt the current processing method; when the user hopes to see other processing methods, enter step S3 according to the user feedback;

[0060] Based on the current processing object, combined with the preset parameter standard of the corresponding processing method, determine whether the current target parameter meets the corresponding parameter standard. If so, it indicates that the current processing method is qualified, and continue to use this processing method as the implementation processing method; if not or a processing method switching intention is received, then execute step S3;

[0061] In step S2, when it is determined through identification that the target parameter meets the standard conditions, obtain the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment and their corresponding treatment methods, and associate and save them; also add them as qualified sample data to the big data processing center to guide subsequent or other oil and gas treatments.

[0062] In practical applications, when it is the case, obtain and associate and save the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment and their corresponding treatment methods; when the target parameter meets the standard conditions, save it as qualified sample data and use it as new big data to guide subsequent or other oil and gas treatments; upload the qualified sample data to a preset big data processing center;

[0063] Among them, the standard conditions correspond to the current treatment method and / or treatment object. Optionally, the standard conditions are artificially set conditions. In practical applications, the standard conditions can be set according to the regulations and conditions of the treatment standard file corresponding to the treatment object. The standard conditions corresponding to different treatment methods for the same treatment object can be set separately.

[0064] For the same treatment object, that is, the initial parameters of the treatment object are exactly the same, or for the same treatment object and the same treatment method is adopted, the present invention performs identification and provides a corresponding judgment switching strategy; through this same judgment, the selection lock-in rigidity is effectively avoided.

[0065] Further, in step S3, determine the number of times of entering this step for the same treatment object. When the number of times is less than the first number, enter step S4; otherwise, enter step S5;

[0066] Through step S3, immediately identify the number of occurrences of the situation in step S2 where "for the same treatment object, its target parameter does not meet the standard conditions"; for example, if the first number is set to 2, the initial value of the number of times is 0. The above situation is set to 1 for the first time. At this time, normally enter step S4. If it is determined that the number of times is equal to or greater than 2, enter step S5.

[0067] Specifically, in step S3, when it is determined whether it is the same treatment object and the same treatment method, if so, increment the number value by 1, and determine whether the updated number value is less than the set first number. If it is less, execute step S4; otherwise, execute step S5, where the initial value of the number of times is 0;

[0068] Among them, determine whether it is for the same treatment object by determining whether the initial parameters are the same, including the type and value of the initial parameters. Considering that the selection method corresponding to step S4 is relatively fast and simple, while the selection method corresponding to step S5 has a higher operation complexity and is even less accurate than the selection in step S4 when the model training is insufficient. Therefore, in practical applications, the choice of whether to run the more complex selection can be made based on the above considerations.

[0069] Preferably, the first number is a preset value; the first number can be set to 2;

[0070] Alternatively, when it is determined that the number of times of entering this step by using the same processing method for the same processing object is less than the first number, enter step S4;

[0071] When there are fewer optional processing methods, using this processing method can reduce the number of attempts and directly adopt the more precisely selected processing method.

[0072] Further, when executing step S4, determine the key target parameter, select the first processing method based on the key target parameter, and enter step S6;

[0073] In step S4, based on the current processing object, the corresponding key target parameter is selected. Optionally, in one embodiment, one or more target parameters that deviate the farthest from the standard condition among the target parameters are used as the key target parameter or the key target parameter group; based on a certain target parameter, calculate its distance from the standard condition according to the following formula:

[0074]

[0075] where A represents the value of the current target parameter, and U1, U2 represent that the standard condition of the current target parameter belongs to [U1, U2].

[0076] In the step S4, in the process of selecting the corresponding first processing method based on the key target parameter, it includes:

[0077] If it is a single key target parameter, select the processing method with the strongest processing ability for the current key target parameter as the selected first processing method;

[0078] If it is a key target parameter group, respectively select the processing method with the strongest processing ability for each key target parameter in the key target parameter group, and use the number of occurrences as the weight to select the processing method with the highest weight as the first processing method.

[0079] The researchers of the present invention found that when there are many target parameters that do not meet the standard conditions and there are many processing methods, selecting a single key target parameter at a time will cause oscillations in the selection of processing methods. Therefore, a selection method for the key target parameter group is proposed;

[0080] Alternatively, when the number of target parameters that do not meet the standard conditions exceeds the first quantity and the number of candidate processing methods exceeds the second quantity, the determination of the key target parameter is to determine the key target parameter group; where the first quantity and the second quantity are preset values;

[0081] The determination of the key target parameter group is specifically as follows: Determine the target parameters among the target parameters whose distances from the standard conditions are the top N to form the key target parameter group; the distance calculation method is the same as above, and N is a preset value; for example: N = 3, the target parameter group is (A1, A2, A3), and the target parameter group is represented by an ordered tuple;

[0082] Correspondingly, when selecting the first processing method based on the key target parameter group, optionally, specifically, select the first processing method based on the key target parameter group, sequentially select the processing method based on each key target parameter in the key target parameter group, and put the selected processing methods into the processing method set. When the same processing method is added again, weight the processing method; take the processing method with the highest weight in the processing method set as the first processing method;

[0083] When the weights of all processing methods are the same, take the processing method selected based on the first key target parameter in the key target parameter group as the first processing method.

[0084] Further, when performing step S5, select the second processing method based on the initial parameters, intermediate parameters, and target parameters corresponding to the current oil and gas treatment, and enter step S6;

[0085] Select the target model based on the intermediate parameters of the current processing object, and input the initial parameters, intermediate parameters, and target parameters into the selected target model to obtain the second processing method; specifically, first select the target model whose intermediate parameter type of the model input is the same as that of the intermediate parameters of the current processing object, and then input the initial parameters, intermediate parameters, and target parameters into the selected target model to determine the second processing method.

[0086] Preferably, if the current intermediate parameter type is more than the intermediate parameter type required for model input, select the intermediate parameter type required for model input from the current intermediate parameter type, and discard the redundant parameter data;

[0087] If the current intermediate parameter type is less than the intermediate parameter type required for model input, fill the missing intermediate parameter type data required for input with the allowed extreme values.

[0088] Further, in an optional embodiment, in step S5, when there are multiple second processing methods obtained by inputting the initial parameters, intermediate parameters, and target parameters into the selected target model, sort the multiple second processing methods based on the degree of coincidence of the intermediate parameters, the number of training times of the model, and the operating cost, so that different processes can select different second processing methods in order.

[0089] Preferably, when setting the input of relevant parameters, use hierarchical input, and input the three types of parameters into different levels or sub-modules of the model;

[0090] Select a target model from a preset model library based on the intermediate parameters. Optionally, specifically, select a model whose intermediate parameter type required for the model is exactly the same as the intermediate parameter type targeted for the current processing as the target model;

[0091] Preferably, when there is no model with exactly the same intermediate parameter type, select a model with a roughly the same type as the target model; roughly the same means that the number of intermediate parameter types that are not exactly the same between the two is less than a set threshold;

[0092] In this way, when the current intermediate parameter type is more than the intermediate parameter type required for input, use the current intermediate parameter type data as the data of the intermediate parameter type required for input, and discard the redundant parameter data. Conversely, fill the missing intermediate parameter type data required for input with the allowed extreme values; filling the extreme values causes the accurate dependence of the model on the default parameters to decrease accordingly in the default case, making the trend tend to select the correct model.

[0093] Preferably, when the number of target models is less than 1, that is, when no target model is selected, increase the size of the set threshold; conversely, when the number of target models is greater than 5, reduce the size of the preset value;

[0094] Preferably, when there are multiple target models, sort the multiple second processing methods obtained from the multiple target models; the sorting method is to sort according to the degree of coincidence of the intermediate parameters, the number of training times of the model, the running cost, etc.

[0095] Further, when performing step S6, process the processing object based on the first processing method selected in step S4 above or the second processing method selected in step S5. After the processing is completed, obtain the processing result and return to step S1;

[0096] Preferably, when there are multiple first processing methods or second processing methods, select one processing method for processing;

[0097] The selection method is to select the processing method with the highest ranking or manual selection; the basis for manual selection can be according to the adaptability of the equipment or manpower.

[0098] The method further includes: step EXS1: perform model training based on big data;

[0099] The steps specifically include: there is one or more models, and different models have different model training requirements; when the amount of big data meets the model training requirements, perform training based on the training data including initial parameters, intermediate parameters, and target parameters and their corresponding processing methods and the model that meets the training requirements;

[0100] In the model training step, initial parameters, intermediate parameters, and target parameters related to oil and gas processing operations of a set scale and their corresponding processing methods are prepared as training sample data, and a set basic model is trained to obtain a target model; the basic model adopts a neural network model, a convolutional neural network, machine learning, or a support vector machine model.

[0101] The complexity and adaptability of each model are different, and the model update is very fast. The same data can be used for the training of multiple models. Making full use of big data can improve the data reusability and operating efficiency.

[0102] More importantly, although increasing the data volume is beneficial to improving the model accuracy, the training data of the model will only be substantially improved when it reaches a certain quantity. On the contrary, training the model is time-consuming and laborious.

[0103] Therefore, the present invention proposes to set different training requirements for different models, so as to improve the data reusability while improving the training efficiency, and greatly improve the operating efficiency.

[0104] Preferably, the model training requirements include the number of newly added training data and the requirements that specific type data in the training data need to meet; for example, the requirements that the initial parameters need to meet.

[0105] It is required that the processing temperatures of the processing devices in the initial parameters all need to meet a certain temperature range; in this way, the trained model can be widely used when selected. For example, it is required that the temperature of the processing device enables most implementers to use common devices to achieve, reducing the resource requirements; one training data may correspond to one or more processing methods.

[0106] Preferably, the initial parameters, intermediate parameters, and target parameters corresponding to different processing methods are stored in separate logical storage spaces; for example, the initial parameters, intermediate parameters, and target parameters corresponding to the same processing method are stored in a separate data table.

[0107] Preferably, when multiple processing methods correspond to the same data entry, all the initial parameters, intermediate parameters, and target parameters are saved in the same logical space, and each initial parameter, intermediate parameter, and target parameter is associated and stored with the index values of different processing methods; among them: the index values are stored in one bit of the binary number, and the index values corresponding to multiple processing methods together form a binary number; for example, there are two processing methods, A and B, where A is stored in the second bit and B is stored in the first bit, and the binary number 10 identifies the corresponding A processing method, and the binary number 11 identifies the corresponding A and B processing methods; this storage method improves the storage efficiency in the case of data reuse and is adapted to the subsequent model training input.

[0108] Preferably, the model includes multiple sub-models, and the multiple sub-models are located in one or more different levels. The initial parameters, intermediate parameters, and target parameters are respectively input into different sub-models; these parameters are relatively independent of each other. Through this model setting, the flexibility is greatly improved, and the training accuracy will increase rapidly as the amount of data increases; by introducing multiple neural networks for balancing, the accuracy of the segmentation result can also be improved.

[0109] Preferably, the model is a two-layer neural network model; the initial parameter, intermediate parameter, and target parameter are respectively input into the first, second, and third sub-models; the outputs of the first, second, and third sub-models are used as the input of the fourth sub-model, and the second processing method is determined based on the output of the fourth sub-model;

[0110] Preferably, the model is a three-layer neural network model; the initial parameter is input into the first sub-model, the output of the first sub-model and the intermediate parameter are used as the input of the second sub-model, and the output of the second sub-model and the target parameter are used as the input of the third sub-model;

[0111] Preferably, the binary number formed by the index values is used as the input of the sub-model; specifically, the binary number formed by the index values is used as the input of the second sub-model, and is jointly input into the second sub-model with the intermediate parameter and the output of the first sub-model;

[0112] Preferably, the sub-model is an artificial intelligence model such as a neural network model, a convolutional neural network, machine learning, a support vector machine, etc.;

[0113] Preferably, the output of the model is a binary classification, and its output value of 0-1 represents the classification result.

[0114] The researchers of the present invention considered that there are differences in the relevant parameters of different types of oil and gas treatment methods. For multiple types of oil and gas treatment methods, a large number of relevant parameters must be considered. By classifying them according to the matching treatment process and operation model, the parameter types can be appropriately simplified, so as to make decisions on the treatment method based on big data while controlling the operation complexity, and achieve reliable and comprehensive oil and gas treatment.

[0115] In actual application, for the historical oil and gas treatment operations after the oil and gas treatment is completed, the relevant parameters of the oil and gas treatment object for that time are obtained; in an optional embodiment, three types of parameters involved in the entire treatment process are obtained, including: initial parameters, intermediate parameters, and target parameters.

[0116] Among them, the initial parameters are the initial detection parameters of the oil and gas treatment object before treatment; the intermediate parameters are the intermediate parameters in the oil and gas treatment process related to the oil and gas treatment method, for example, may include: the treatment temperature of the treatment equipment, the real-time capacitance measurement value of the key equipment, the sulfur vapor gas temperature, etc.; the target parameters are the general target parameters targeted by the oil and gas treatment that are not closely related to the oil and gas treatment method, for example, may include the output energy consumption, the unit output energy consumption, the unit energy consumption output, the power generation efficiency of the oil and gas, the treatment pollution index, etc.; the target parameters have the significance of guiding the optimization direction of the treatment method.

[0117] Adopting the solution of the above-mentioned embodiment of the present invention has at least the following application advantages: (1) Gradually adjust the treatment experience based on big data in the absence of treatment experience and obtain a treatment method with good treatment effect; (2) By classifying the data involved in the treatment, different oil and gas treatment methods can learn from each other and spread to each other, which is different from the defect of only relying on local data for learning in the prior art; (3) Input different types of parameters into different sub-models and levels, which greatly improves the flexibility, and the training accuracy will increase rapidly with the increase of the data volume; (4) Introduce multiple neural networks for balancing, and can also improve the accuracy of the selection result. (5) Propose the representation methods of key target parameters and key target parameter groups, which greatly simplifies the complexity of the primary selection and ensures a certain degree of accuracy. Especially based on the method of key target parameter groups, it has a good effect in solving the problem of multi-objective and multi-model selection.

[0118] For the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should know that the present invention is not limited by the described action sequence, because according to the present invention, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0119] It should be noted that in other embodiments of the present invention, the method can also be combined with one or several of the above-mentioned embodiments to obtain a new oil and gas treatment method based on big data, so as to achieve efficient and accurate data processing for oil and gas treatment operations.

[0120] The present invention also provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the processor executes the oil and gas treatment method based on big data as described in the above-mentioned embodiment.

[0121] Embodiment 2:

[0122] In the embodiments of the present invention disclosed above, the method has been described in detail. The method of the present invention can be implemented by various forms of devices or systems. Therefore, based on other aspects of the method described in any one or more of the above embodiments, the present invention also provides an oil and gas processing system based on big data, which is used to execute the oil and gas processing method based on big data described in any one or more of the above embodiments. Specific embodiments are given below for detailed description.

[0123] Specifically, Figure 2 The structural schematic diagram of the oil and gas processing system based on big data provided in the embodiments of the present invention is shown in Figure 2 As shown, the system includes:

[0124] A parameter acquisition module, configured to acquire relevant parameters of the processing object during the processing based on a certain oil and gas processing operation; wherein, it includes initial parameters, intermediate parameters, and target parameters representing the optimization target of the processing method;

[0125] A parameter quality identification module, configured to identify whether the current target parameter meets the corresponding parameter standard based on the current processing object and in combination with the preset parameter standard of the corresponding processing method. If so, it indicates that the current processing method is qualified, and the current processing method continues to be used as the implemented processing method; if not or a processing method switching intention is received, the processing method decision module is enabled;

[0126] A processing method decision module, configured to determine whether it is the same processing object and the same processing method. If so, increment the count value by 1, and determine whether the updated count value is less than the set first count value. If it is less, the first processing method selection module is enabled; otherwise, the first processing method selection module is enabled, where the initial value of the count is 0;

[0127] A first processing method selection module, configured to select key target parameters based on the current processing object, select the corresponding first processing method based on the key target parameters, perform the oil and gas processing operation based on the selected processing method, and update relevant parameters according to the processing result;

[0128] A second processing method selection module, configured to select a second processing method using the target model based on the relevant parameters corresponding to the current processing object, perform the oil and gas processing operation based on the selected processing method, and update relevant parameters according to the processing result;

[0129] Wherein, the target model is a neural network model trained based on big data using the model training module.

[0130] Further, in an alternative embodiment, the parameter quality identification module is configured to, when it is determined through identification that the target parameter meets the standard conditions, obtain the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment and their corresponding treatment methods, and associate and save them; and also add them as qualified sample data to the big data processing center to guide subsequent or other oil and gas treatments.

[0131] Preferably, in one embodiment, the treatment method decision module is configured to use one or more target parameters in the target parameter that deviate the farthest from the standard conditions as the key target parameter or key target parameter group; based on a certain target parameter, calculate its distance from the standard conditions according to the following formula:

[0132]

[0133] where A represents the value of the current target parameter, and U1 and U2 represent that the standard condition of the current target parameter belongs to [U1, U2].

[0134] In one embodiment, the treatment method decision module determines whether it is the same treatment object according to the initial parameter. If the initial parameter of the current treatment object is the same as that of the previous treatment object, it indicates that the two are the same treatment object.

[0135] Further, in one embodiment, the first treatment method selection module selects the corresponding first treatment method based on the key target parameter according to the following logic:

[0136] If it is a single key target parameter, select the treatment method with the strongest processing ability for the current key target parameter as the selected first treatment method;

[0137] If it is a key target parameter group, respectively select the treatment method with the strongest processing ability for each key target parameter in the key target parameter group, and use the number of occurrences as the weight to select the treatment method with the highest weight as the first treatment method.

[0138] Optionally, in one embodiment, the second treatment method selection module is configured to select the second treatment method based on the initial parameter, intermediate parameter, and target parameter corresponding to the current oil and gas treatment. First, select the target model whose input intermediate parameter type is the same as the intermediate parameter type of the current treatment object, and then input the initial parameter, intermediate parameter, and target parameter into the selected target model to determine the second treatment method.

[0139] Further, in one embodiment, the second treatment method selection module is configured to, when there is no model with exactly the same intermediate parameter type, select one or more target models with the number of intermediate parameter types that are not exactly the same less than the set value.

[0140] Specifically, in an optional embodiment, the second processing mode selection module is configured to, if the current intermediate parameter type is more than the intermediate parameter type required for input by the model, select the intermediate parameter type required for input by the model from the current intermediate parameter type, and discard the redundant parameter data;

[0141] If the current intermediate parameter type is less than the intermediate parameter type required for input by the model, fill the data of the missing intermediate parameter type required for input with the allowed extreme values.

[0142] Optionally, in an embodiment, when there are multiple second processing modes obtained by inputting the initial parameter, intermediate parameter, and target parameter into the selected target model, sort the multiple second processing modes based on the degree of coincidence of the intermediate parameter, the number of training times of the model, and the running cost, so as to enable different processes to select different second processing modes in sequence.

[0143] In a preferred embodiment, the model training module is configured to prepare the initial parameter, intermediate parameter, and target parameter related to the oil and gas processing operation of a set scale and their corresponding processing modes as training sample data, and train the set basic model to obtain the target model; the basic model adopts a neural network model, a convolutional neural network, machine learning, or a support vector machine model.

[0144] In the oil and gas processing system based on big data provided by the embodiments of the present invention, each module or unit structure can operate independently or in combination according to the actual parameter setting requirements and operation processing requirements to achieve the corresponding technical effects.

[0145] The present invention also provides an oil and gas processing system, which is characterized in that it includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the processor is provided with each functional structure of the oil and gas processing system based on big data, and the memory stores instructions executable by the one processor, and the instructions are executed by the at least one processor to enable the at least one processor to execute the oil and gas processing method based on big data as described in the above embodiments.

[0146] The present invention also provides a computer device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, which is characterized in that when the processor executes the computer program, it implements the oil and gas processing method based on big data as described in the above embodiments.

[0147] It should be understood that the embodiments disclosed in the present invention are not limited to the specific structures, processing steps, or materials disclosed herein, but should extend to equivalent alternatives of these features understood by those of ordinary skill in the relevant art. It should also be understood that the terms used herein are only for the purpose of describing specific embodiments and do not mean limitation.

[0148] As used in the specification, "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the phrase "an embodiment" that appears throughout the specification does not necessarily refer to the same embodiment.

[0149] Although the embodiments disclosed in the present invention are as described above, the above content is only an embodiment adopted for the convenience of understanding the present invention and is not intended to limit the present invention. Any person skilled in the art within the technical field to which the present invention pertains may make any modifications and variations in the form and details of the implementation without departing from the spirit and scope disclosed by the present invention. However, the scope of patent protection of the present invention shall still be subject to the scope defined by the appended claims.

Claims

1. An oil and gas processing method based on big data, characterized in that The method includes: Parameter acquisition step S1: Based on a certain oil and gas treatment operation, acquire relevant parameters of the treatment object during the treatment process; among them, it includes initial parameters, intermediate parameters, and target parameters characterizing the optimization goal of the treatment method. Parameter quality identification step S2: Based on the current treatment object, combined with the preset parameter standards of the corresponding treatment method, identify whether the current target parameter meets the corresponding parameter standards. If so, it indicates that the current treatment method is qualified, and continue to use this treatment method as the implemented treatment method; if not or a treatment method switching intention is received, then execute step S3. Treatment method decision step S3: Determine whether it is the same treatment object and the same treatment method. If so, increment the number value by 1, and determine whether the updated number value is less than the set first number. If it is less, execute step S4; otherwise, execute step S5, where the initial value of the number is 0. First treatment method selection step S4: Based on the current treatment object, select the key target parameter, and based on the key target parameter, select the corresponding first treatment method, and enter step S6. Second treatment method selection step S5: Comprehensively utilize the relevant parameters corresponding to the current treatment object, and select the second treatment method using the target model, and enter step S6. Treatment result output step S6: Perform oil and gas treatment operations based on the selected treatment method, and update the relevant parameters according to the treatment results. Among them, the target model is a neural network model obtained by training based on big data through model training step S0.

2. The method according to claim 1, wherein In step S2, when it is identified that the target parameter meets the standard conditions, acquire the initial parameters, intermediate parameters, and target parameters corresponding to the current oil and gas treatment and their associated treatment methods, and save them; also add them as qualified sample data to the big data processing center to guide subsequent or other oil and gas treatments.

3. The method according to claim 1, characterized in that, In step S4, take one or more target parameters with the farthest distance from the standard conditions among the target parameters as the key target parameter or key target parameter group; based on a certain target parameter, calculate its distance from the standard conditions according to the following formula: Where, A represents the value of the current target parameter, and U1, U2 represent that the standard condition of the current target parameter belongs to [U1, U2].

4. The method according to claim 1, wherein In the said step S4, during the process of selecting the corresponding first treatment method based on the key target parameter, it includes: If it is a single key target parameter, select the treatment method with the strongest processing ability for the current key target parameter as the selected first treatment method. If it is a key target parameter group, respectively select the treatment method with the strongest processing ability for each key target parameter in the key target parameter group, and use the occurrence times as the weight, and select the treatment method with the highest weight as the first treatment method.

5. The method according to claim 1, characterized in that In the said step S5, when selecting the second treatment method based on the initial parameters, intermediate parameters, and target parameters corresponding to the current oil and gas treatment, first select the target model with the same intermediate parameter type as the intermediate parameter type of the current treatment object for model input, and then input the initial parameters, intermediate parameters, and target parameters into the selected target model to determine the second treatment method.

6. The method according to claim 1, characterized in that, In the step S5, when there is no model with exactly the same intermediate parameter types, one or more target models with the number of intermediate parameter types that are not exactly the same less than a set value are selected.

7. The method according to claim 1, wherein In the step S5, when there are multiple second processing methods obtained by inputting the initial parameters, intermediate parameters, and target parameters into the selected target model, the multiple second processing methods are sorted based on the degree of coincidence of the intermediate parameters, the number of training times of the model, and the running cost, so that different processes can select different second processing methods in order.

8. The method according to claim 1, characterized in that In the step S5, if the current intermediate parameter types are more than the intermediate parameter types required to be input by the model, the intermediate parameter types required to be input by the model are selected from the current intermediate parameter types, and the redundant parameter data is discarded; If the current intermediate parameter types are less than the intermediate parameter types required to be input by the model, the data of the missing intermediate parameter types required to be input is filled with the allowed extreme values.

9. The method according to claim 1, wherein In the model training step, initial parameters, intermediate parameters, and target parameters related to oil and gas processing operations of a set scale and their corresponding processing methods are prepared as training sample data, and a set basic model is trained to obtain a target model; the basic model uses a neural network model, a convolutional neural network, machine learning, or a support vector machine model.

10. A storage medium, characterized in that, The storage medium stores program code that can implement the method according to any one of claims 1 to 9.

11. An oil and gas processing system based on big data, characterized in that, The system executes the method according to any one of claims 1 to 9.

Images