Perforation optimization cloud computing method based on micro-service technology

Abstract

The invention discloses a perforation optimization cloud computing method based on a micro-service technology, and the method comprises the steps: achieving the real-time transmission of perforation data through Kafka, Spark and Hbase, carrying out the field perforation data analysis, and optimizing the distributed storage of data, so as to construct a perforation cloud computing environment; multi-source data are integrated to carry out data output standardization and each sub-process decoupling, all the sub-processes can be independently operated, a set of perforation optimization basic process template is abstracted, then customization is carried out according to each well type, and the reusability and flexibility of a software architecture are improved; through combination of a perforation database system and computing micro-service, all perforation optimization computing processes are transferred to a server side for operation, the constraint of a client side is liberated, optimization computing and optimization service are realized, interconnection and sharing are realized anytime and anywhere, the problem of interconnection and intercommunication of data and settlement results among large-span regions is solved, performance limitation of field equipment is also solved, and the system is suitable for popularization and application. And finally, all data and results are fully utilized, and energy conservation and efficiency improvement are realized.

Description

Technical Field

[0001] The present invention relates to a cloud computing method for perforation optimization, and in particular to a cloud computing method for perforation optimization based on microservice technology, belonging to the technical field of intelligent perforation operation. Background Art

[0002] The earliest oil production method was open hole oil production or screen oil production. With the emergence of cementing technology, perforation oil production technology was developed. Since 1932, when perforation technology was applied in the exploration and development of oil and gas fields, perforating bullets have developed from the initial bullet type to the currently widely used shaped perforating bullets. Perforating bullets are divided into two categories: deep penetration and large aperture, which can meet the completion perforation needs of normal temperature to high temperature formations. Perforation completion is the most widely used completion method at home and abroad. It uses the energy of pyrotechnics or other energy sources to shoot open casing, cement ring and formation. Among them, the downhole operation of communicating oil and gas flow channels is called perforation.

[0003] At present, with the development of science and technology, software for perforation calculation has also been put into use. However, the existing perforation technology was developed early, the software technology used is backward, the perforation calculation software system is scattered, the function is single, the scope of application is small, the use efficiency is low, the theories, models, and rules used are outdated, and they cannot be updated in time to reflect the new technologies and methods of perforation. The software platform has poor adaptability and has certain requirements for equipment hardware. In addition, the data involved are scattered and fragmented and cannot be shared in time. The project takes a lot of time, the geographical span is large, and the well types are complex and diverse.

[0004] The prior art CN110597522A discloses "a cloud computing big data platform system based on microservices", which mainly solves the following problems: the big data analysis platform runs on the Linux operating system, is distributedly deployed, and has a complex structure; business personnel need to manually complete system management and platform operation and maintenance through command lines or system built-in tools, and cannot focus on solving business problems, which seriously affects work efficiency; CN116303698A discloses "a method for constructing an open application platform for perforating big data", which mainly solves the problem that the forecast research work on the opening of oil and gas reservoirs using big data technology is still relatively rare.

[0005] Therefore, a perforation optimization calculation method that meets practical needs is urgently needed. Summary of the invention

[0006] The present invention mainly solves the problem of comprehensive application and effective management of various data in perforation design optimization and detection operations, and proposes a perforation optimization cloud computing method based on microservice technology. By defining a format template, the flexible application of data and business related to perforation optimization and detection is realized, and comprehensive comparison is made to integrate relevant perforation design optimization microservice modules such as data intelligent extraction, segmentation and clustering optimization design, crack analysis, geological sweet spots, row gun design, and optimization evaluation under a unified platform. According to the actual application requirements of perforation design optimization and detection engineering operations, based on the Windows platform, existing business functions can be realized by simply clicking and selecting, without switching between multiple software or memorizing too many commands. Business personnel can select corresponding application modules or functions according to business processes or actual business needs, and display the results in the form of data or graphics.

[0007] In order to achieve the above technical objectives, the following technical solutions are proposed: The first objective of this technical solution is to propose a cloud computing method for perforation optimization based on microservice technology, comprising the following steps: S1: Deploy the terminal software with a graphical interface on the working computer to provide a visual interface for perforation optimization, and deploy the cloud service including database storage and cloud computing business on the cloud server; S2: Through Kafka, Spark and Hbase, real-time transmission of perforation data, on-site perforation data analysis and optimized data distributed storage are carried out to build a perforation cloud computing environment; S3: Integrate multi-source data and standardize data output; decouple each sub-process so that they can run independently, abstract a set of basic process templates for perforation optimization, and then customize them according to each well type to improve the reusability and flexibility of the software architecture; S4: Decouple computing from business. The terminal software is compatible with both local computing and cloud computing. The original computing part is embedded in the business layer. The local computing program is called during optimization. Now the computing abstraction layer is introduced to provide computing services for the business. The business layer is unaware of how the underlying layer implements computing. Even if the computing method changes, the business architecture can remain unchanged, realizing the decoupling of computing from the business layer and maintaining the stability of the architecture. The computing abstraction layer is compatible with both local computing and cloud computing. S5: Optimize neural network parameters through back propagation, and optimize company business parameters through back propagation and parameter optimization framework; S6: Evaluation of segmented clustering schemes through MLP / CNN and back propagation, and generation of graded clustering schemes through reinforcement learning and neural networks.

[0008] Furthermore, in step S1, the cloud service can be divided into three modules: business layer, transmission layer and storage layer; All types of computing services belong to the business layer; The transport layer is used to send and receive data between the terminal and the cloud service; The core of the storage layer is the database, which stores the data required by each terminal software.

[0009] Furthermore, the storage layer includes data cleaning and high-performance data warehouse storage based on Hbase+ Spark+ Kafka technology; In step S3, the multi-source data is integrated to standardize the data output, that is, it is performed on the cloud service. After the multi-source data is integrated, data preprocessing is performed to complete the data integration; Among them, multi-source data includes perforating equipment data, reservoir information data, operation process data and design process data, and data preprocessing includes data cleaning, data conversion and data standardization.

[0010] Furthermore, the transport layer in the cloud service adopts two types of transmission schemes according to the different data sizes: For services with less request data, directly package the data in the request, using RPC and / or HTTP; For businesses with large request data, the request and the data body are separated and the data is transmitted separately.

[0011] The core computing part of the cloud service business layer is divided into two categories: Static processing program: For algorithmic programs that do not require state maintenance, executable programs can be used; Dynamic microservices: For business processing that requires intermediate state maintenance, they are deployed as independent microservices; Task routing and distribution management in cloud business: Business request types distribute tasks to different business programs or microservices.

[0012] Furthermore, in step S3, a set of basic process templates is abstracted for perforation optimization: divided into two categories: basic service category and core business category, with a total of 8 sub-modules, each sub-module defines standardized input and output, and the sub-modules are decoupled, so that any one or several sub-modules can be executed independently.

[0013] Furthermore, after the multi-source data is integrated in step S3, it enters the data center; the data center includes basic layer data, summary layer data and algorithm library, basic layer data: mapping configuration is performed on the integrated multi-source data, after the data is analyzed by the perforation storage model, real-time processing and integration are performed, and then the data is transmitted to the summary layer data; Summary layer data: data mining, including perforation effect summary, perforation data mart, conventional well optimization, segmentation and clustering optimization, and master index management; Algorithm libraries: including MLlib, TensorFlow, and PySpark.

[0014] Furthermore, the intelligent application in the terminal software registers the microservice and performs access control on the summary layer data through the REST interface.

[0015] Furthermore, the performance of the REST interface is measured by the P90 percentile of the latency data. The P90 percentile is the critical value that separates the slowest ten percent of latency data. The P90 percentile of latency can more stably reflect the performance of microservices in a cloud environment. The response time R1 of the entry service is actually determined by the response time of all microservices involved in the call and the processing time of the entry service itself, which can be expressed as formula (1): R1=f(R2, R3, ⋯, Rn) + S1 (1); In formula (1), R1 represents the response time of the portal service, "R2, R3, ⋯, Rn" represent the response time of multiple background services called by the portal service, S1 represents the time required for the portal service to process the request, and f is a function involving linear transformation and maximum transformation. The weight of the linear transformation is determined by how many times other services are called each time the portal service is called. Whether there is a maximum transformation is determined by whether the portal service uses parallel or serial calls to other services multiple times. If the call of the entry service does not involve any other services, the response time R1 is only related to S1. The response time R2, R3, ⋯, Rn of a call to a non-entry service can also be derived from formula (1). The final response time R1 can be expressed as formula (2): R1=g(S1, S2, ⋯, Sn) (2); In formula (2), Sn represents the time required for service i to process a request, g is a function with nested linear transformation and maximum transformation, and the processing time Si of service i for a request can be expressed as formula (3): Si = Wi + Pi (3); In formula (3), Wi represents the waiting time, Pi represents the time required to actually process the request, and the waiting time is related to the number of requests per second of the service instance.

[0016] The second objective of the present technical solution is to provide: a computer-readable storage medium having a computer program stored thereon, and the computer program, when executed by a processor, implements the steps of the above-mentioned perforation optimization cloud computing method based on microservice technology.

[0017] The third objective of the present technical solution is to provide: an information data processing terminal for a perforation optimization cloud computing method based on microservice technology.

[0018] The beneficial technical effects brought about by adopting this technical solution are: The present invention combines the perforation database system and computing microservices to transfer all perforation optimization calculation processes to the server side for operation, freeing the constraints of the client side, realizing optimization calculation and optimization services, and interconnecting and sharing anytime and anywhere. It not only solves the problem of interconnection and interoperability of data and settlement results between large-span areas; it also solves the problems of performance limitations of on-site equipment and data islands, making full use of all data and results to save energy and increase efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a work flow chart related to the present invention; Figure 2 This is a schematic diagram of an information processing terminal involved in Embodiment 7 of the present invention; Figure 3 This is a schematic diagram of a data preprocessing terminal involved in Embodiment 8 of the present invention; Figure 4 This is a schematic diagram of the explanation report data processing involved in Example 8 of the present invention; Figure 5 This is a schematic diagram of table data processing involved in Example 8 of the present invention; Figure 6 Schematic diagram of well logging data processing involved in Example 8 of the present invention; Figure 7 This is a schematic diagram of some data fields involved in Example 8 of the present invention; Figure 8 This is a schematic diagram of effect evaluation involved in Example 9 of the present invention; Fig. 9 This is a schematic diagram of the effect evaluation scoring involved in Example 9 of the present invention. DETAILED DESCRIPTION

[0020] The following is a clear and complete description of the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Example 1 This embodiment provides a cloud computing method for perforation optimization based on service technology, such as Figure 1 As shown, the following steps are included: S1: Deploy the terminal software with a graphical interface on the working computer to provide a visual interface for perforation optimization, and deploy the cloud service including database storage and cloud computing business on the cloud server; S2: Through Kafka, Spark and Hbase, real-time transmission of perforation data, on-site perforation data analysis and optimized data distributed storage are carried out to build a perforation cloud computing environment; S3: Integrate multi-source data and standardize data output; decouple each sub-process so that they can run independently, abstract a set of basic process templates for perforation optimization, and then customize them according to each well type to improve the reusability and flexibility of the software architecture; S4: Decouple computing from business. The terminal software is compatible with both local computing and cloud computing. The original computing part is embedded in the business layer. The local computing program is called during optimization. Now the computing abstraction layer is introduced to provide computing services for the business. The business layer is unaware of how the underlying layer implements computing. Even if the computing method changes, the business architecture can remain unchanged, realizing the decoupling of computing from the business layer and maintaining the stability of the architecture. The computing abstraction layer is compatible with both local computing and cloud computing. S5: Optimize neural network parameters through back propagation, and optimize company business parameters through back propagation and parameter optimization framework; S6: Evaluation of segmented clustering schemes through MLP / CNN and back propagation, and generation of graded clustering schemes through reinforcement learning and neural networks.

[0022] This perforation optimization cloud computing method based on microservice technology combines the perforation database system and computing microservices to transfer all perforation optimization calculation processes to the server-side for operation, freeing the constraints of the client, realizing optimized calculation and optimization services, and sharing them anytime and anywhere. It solves the problem of interconnection and interoperability of data and settlement results between large-span areas, solves the performance limitations of on-site equipment, solves the problem of data islands, makes full use of all data and results, and saves energy and increases efficiency.

[0023] Example 2 On the basis of Example 1, this example further limits the cloud service to further illustrate the present invention.

[0024] In step S1, the cloud service can be divided into three modules: business layer, transmission layer and storage layer; Business layer module: including various computing services; Transport layer module: used for sending and receiving data between the terminal and the cloud service; Storage layer module: stores the data required by each terminal software.

[0025] Among them, the storage layer module includes data cleaning and high-performance data warehouse storage based on Hbase+ Spark+ Kafka technology; In step S3, the multi-source data is integrated for data output standardization, which is performed on the cloud service. After the multi-source data is integrated, data preprocessing is performed to complete the data integration; the multi-source data includes perforating equipment data, reservoir information data, operation process data and design process data, and data preprocessing includes data cleaning, data conversion and data standardization.

[0026] The transport layer module in the cloud service uses two types of transmission solutions depending on the size of the data: For services with less request data, directly package the data in the request, using RPC and / or HTTP; For businesses with large request data, the request and the data body are separated and the data is transmitted separately.

[0027] The core computing part of the cloud service business layer module is divided into two categories: Static processing program: For algorithmic programs that do not require state maintenance, executable programs can be used; Dynamic microservices: For business processing that requires intermediate state maintenance, they are deployed as independent microservices; Task routing and distribution management in cloud business: Business request types distribute tasks to different business programs or microservices.

[0028] Example 3 On the basis of Embodiments 1-2, this embodiment further defines a set of basic process templates abstracted from perforation optimization in step S3 to further illustrate the present invention.

[0029] Perforating optimization abstracts a set of basic process templates: divided into two categories: basic service category and core business category, with a total of 8 sub-modules. Each sub-module defines standardized input and output. The sub-modules are decoupled, and any one or several sub-modules can be executed independently.

[0030] Example 4 On the basis of Examples 1-3, this example further limits the multi-source data integration in step S3 to further illustrate the present invention.

[0031] After the multi-source data is integrated in step S3, it enters the data center; the data center includes basic layer data, summary layer data and algorithm library, basic layer data: mapping configuration of the integrated multi-source data, after the data is analyzed by the perforation storage model, real-time processing and integration, and then the data is transmitted to the summary layer data; Summary layer data: data mining, including perforation effect summary, perforation data mart, conventional well optimization, segmentation and clustering optimization, and master index management; Algorithm libraries: including MLlib, TensorFlow, and PySpark.

[0032] Example 5 On the basis of embodiments 1-4, this embodiment further limits the terminal software to further illustrate the present invention.

[0033] The intelligent application in the terminal software registers the microservice and controls the access to the summary layer data through the REST interface. The performance of the REST interface is measured by the P90 percentile of the latency data. The P90 percentile is the critical value that separates the slowest 10% latency data. The P90 percentile of latency can more stably reflect the performance of microservices in the cloud environment. The response time R1 of the entry service is actually determined by the response time of all microservices involved in the call and the processing time of the entry service itself, which can be expressed as formula (1): R1=f(R2, R3, ⋯, Rn) + S1 (1); In formula (1), R1 represents the response time of the portal service, "R2, R3, ⋯, Rn" represent the response time of multiple background services called by the portal service, S1 represents the time required for the portal service to process the request, and f is a function involving linear transformation and maximum transformation. The weight of the linear transformation is determined by how many times other services are called each time the portal service is called. Whether there is a maximum transformation is determined by whether the portal service uses parallel or serial calls to other services multiple times. If the call of the entry service does not involve any other services, the response time R1 is only related to S1. The response time R2, R3, ⋯, Rn of a call to a non-entry service can also be derived from formula (1). The final response time R1 can be expressed as formula (2): R1=g(S1, S2, ⋯, Sn) (2); In formula (2), Sn represents the time required for service i to process a request, g is a function with nested linear transformation and maximum transformation, and the processing time Si of service i for a request can be expressed as formula (3): Si = Wi + Pi (3); In formula (3), Wi represents the waiting time, Pi represents the time required to actually process the request, and the waiting time is related to the number of requests per second of the service instance.

[0034] Example 6 This embodiment also provides: a computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the steps of the above-mentioned perforation optimization cloud computing method based on microservice technology are implemented.

[0035] Example 7 This embodiment also provides: an information data processing terminal (such as Figure 2 As shown in the figure), the sending, receiving and parsing between the client and the cloud system are realized through the predefined restful API interface; the routing distributed micro front-end distributes different businesses to different, independent front-end applications through routing, realizing the decoupling and unification of all functions.

[0036] Example 8 Based on Example 2, this example is performed in a cloud service, and after integrating multi-source data, data preprocessing is performed to complete data integration, such as Figure 3-7 shown.

[0037] Among them, multi-source data include perforating equipment data, reservoir information data, operation process data and design process data; Data preprocessing includes data cleaning, data conversion and data standardization. That is, identifying errors in fields and how to correct them, encoding text (strings) into numbers (vectors) for neural network recognition and use; understanding text, identifying fields and parameters, and outputting results. The processing flow includes: 1) Select the required file in the data cleaning and preprocessing interface; 2) Data files are two-dimensionalized to prepare data for calculation; the Levenshtein distance involved is defined as: 3) Algorithm matching and pre-cleaning; For example, for the explanation report data, the fields are summarized; duplicate data, special characters, typos, etc. are filtered out through the bag-of-words model, and then unified encoding conversion is performed, semantic simplification is performed, and finally a CSV file is generated, appended to the new data template, and output to generate a .csv file; By searching the table data, you can identify irregular and unfilled annotations, give a reference annotation, and output a .csv file; The word embedding encoding of the logging data field is performed, and the fuzzy approximate matching algorithm is used to match the table fields. The correlation between the uncollected fields and the existing template fields is appended to the new data template in an objective and manually intervened way, and then output; the output generates a .csv file.

[0038] The data preprocessing module is used to clean, organize and convert different files, and finally generate relevant template data and save it in the cloud server.

[0039] Example 9 This embodiment also provides: artificial intelligence perforation effect evaluation, such as Figure 8-9 As shown, specifically including: 1. Generate segmented and clustered optimization design results by integrating, processing and calculating the required data from different sources; Second, the optimization design results and production logging parameters, microseismic event response, operation monitoring (temperature, pressure, acceleration during perforation), and oil test production (liquid volume, gas volume) are used as data sets to conduct online perforation effect analysis through the big data platform, and the evaluation neural network SR-CNN is used to score the perforation effect; Assist artificial intelligence in the forward optimization of perforation design training, further improve the accuracy of the results, and provide guarantees for efficiency savings and increased production.

Claims

1. A perforation optimization cloud computing method based on microservice technology, characterized in that: The steps include: S1: Deploy the terminal software with a graphical interface on the work computer, and deploy the cloud service including database storage and cloud computing business on the cloud server; S2: Through Kafka, Spark and Hbase, real-time transmission of perforation data, on-site perforation data analysis and optimized data distributed storage are carried out to build a perforation cloud computing environment; S3: Integrate multi-source data and standardize data output; decouple each sub-process, abstract a set of perforation optimization basic process templates, and then customize them according to each well type; S4: Decouple computing from business. The terminal software is compatible with both local computing and cloud computing. The original computing part is embedded in the business layer. When optimizing the processing, the local computing program is called, and the computing abstraction layer is introduced to provide computing services for the business. S5: Optimize neural network parameters through back propagation, and optimize company business parameters through back propagation and parameter optimization framework; S6: Evaluation of segmented clustering schemes through MLP / CNN and back propagation, and generation of graded clustering schemes through reinforcement learning and neural networks.

2. The perforation optimization cloud computing method based on microservice technology according to claim 1 is characterized in that: In step S1, the cloud service can be divided into a business layer module, a transport layer module and a storage layer module; Business layer module: including various computing services; Transport layer module: used for sending and receiving data between the terminal and the cloud service; Storage layer module: stores the data required by each terminal software.

3. The perforation optimization cloud computing method based on microservice technology according to claim 2 is characterized in that: The storage layer module includes data cleaning and data warehouse storage based on Hbase+ Spark+ Kafka technology.

4. The perforation optimization cloud computing method based on microservice technology according to claim 2 or 3 is characterized in that: In step S3, the multi-source data is integrated and the data output is standardized, which specifically includes: after the multi-source data is integrated in the cloud service, data preprocessing is performed to complete data integration; Among them, multi-source data include perforating equipment data, reservoir information data, operation process data and design process data; Data preprocessing includes data cleaning, data conversion and data standardization.

5. The perforation optimization cloud computing method based on microservice technology according to claim 4 is characterized in that: The transport layer module in the cloud service adopts two types of transmission schemes according to the different data sizes: For services with less request data, directly package the data in the request, using RPC and / or HTTP; For businesses with large request data, the request and the data body are separated and the data is transmitted separately.

6. The perforation optimization cloud computing method based on microservice technology according to claim 5 is characterized in that: The core computing program of the business layer module in the cloud service is divided into static processing programs and dynamic microservices; Static processing program: For algorithmic programs that do not require state maintenance, executable programs can be used; Dynamic microservices: For business processing that requires intermediate state maintenance, they are deployed as independent microservices; Task routing and distribution management in cloud business: Business request types distribute tasks to different business programs or microservices.

7. The perforation optimization cloud computing method based on microservice technology according to claim 1 is characterized in that: In step S3, a set of basic process templates for perforation optimization is abstracted: the templates are divided into two categories: basic service category and core business category, with a total of 8 sub-modules. Each sub-module defines standardized input and output. The sub-modules are decoupled, and any one or several sub-modules can be executed independently.

8. The perforation optimization cloud computing method based on microservice technology according to claim 1 or 7, characterized in that: After the multi-source data is integrated in step S3, it enters the data center; the data center includes basic layer data, summary layer data and algorithm library, wherein: Base layer data: Map and configure the integrated multi-source data, analyze the data through the perforation storage model, process and integrate it in real time, and then transfer the data to the summary layer data; Summary layer data: data mining, including perforation effect summary, perforation data mart, conventional well optimization, segmentation and clustering optimization, and master index management; Algorithm libraries: including MLlib, TensorFlow, and PySpark.

9. The perforation optimization cloud computing method based on microservice technology according to claim 1 is characterized in that: The intelligent application in the terminal software performs microservice registration and performs access control on the summary layer data through the REST interface.

10. The perforation optimization cloud computing method based on microservice technology according to claim 9 is characterized in that: The performance of the REST interface is measured by the P90 percentile of the latency data, which is the critical value that separates the slowest ten percent of latency data.

Images