RU2598785C2 - System for statistical and neural network analysis of telemetry data of oil-field objects - Google Patents

Abstract

FIELD: computer engineering.

SUBSTANCE: invention relates to computer engineering, used in oil industry, in particular, to information systems for automation of control of an oil-extracting enterprise. Disclosed system comprises: data selection unit, database for storage of telemetric information, user list adjusting unit parameter request processing unit reporting unit, neural network analysis unit, unit for calculating mathematical expectation, standard deviation calculation unit, asymmetry calculation unit, correlation calculation unit, schedule display unit, interference vector display unit, Maryland matrix display unit, heat map display unit, database of reference data, mapping parameter processing unit, raw data preparation unit, standardised data preparation unit, unit for data preparation in logarithmic coordinates, data preparation unit for mathematical calculations, calculation parameter editing unit, oil production parameter calculation unit.

EFFECT: technical result is creation of a system of statistical processing, aggregation and imaging of data obtained from telemetry systems, in order to obtain information which is suitable for solving tasks of controlling technological processes.

1 cl, 1 dwg

Description

The system of statistical and neural network analysis of telemetry data of oilfield facilities (NPOs) is intended for statistical processing and aggregation of data received from telemetry systems and for obtaining information suitable for solving problems of regulation of technological processes. The system is an analytical tool that provides the system user with access to information in a convenient form for perception.

The invention relates to the field of computer technology used in the oil industry, and in particular to information systems for automating the management of an oil producing enterprise.

As a prototype, “Automation system for selection and statistical data analysis” was selected (RF patent for utility model No. 123557). The system is designed to select, process and administer access to information contained in one or more databases having a different structure and presentation form.

The system, which is close in structure and functionality to the invention, provides for access to data stored in the database, uses mathematical analysis algorithms, and has the ability to build and save reports to the user's computer. The disadvantages of this system include the implementation of mathematical calculations on a client computer, which entails the transfer of a large amount of data from the server computer to the client computer, as well as the need to install additional software on the client computer. This system also does not provide tools for user interaction with a topographic map, in particular, the ability to select objects necessary for analysis on the map.

The tasks to which the invention is directed is the creation of a system of statistical processing, aggregation and visualization using web-based data (density, temperature, volume, pressure, flow rate of process fluids, etc.) obtained from telemetry systems in order to obtain information suitable for solving problems of regulation of technological processes.

The task is achieved in that a system of statistical and neural network analysis of telemetry data of oilfield objects, including a data selection unit, a telemetry information storage database, a user list settings unit, a parameter request processing unit, a report generation unit, a neural network analysis unit, a mathematical expectation calculation unit, deviation calculation unit, asymmetry calculation unit, correlation calculation unit, graph display unit, inter vector display unit according to the invention additionally includes a database of reference information, a unit for processing cartographic parameters, a unit for preparing raw data, a unit for preparing normalized data, a unit for preparing data in logarithmic coordinates, a unit for preparing data for mathematical calculations, a block for editing calculation parameters, a block for calculating oil production parameters, and the block for processing parameter requests by one-way communication is connected with a database for storing telemetric information, a unit for processing parameter requests with two-way communication is connected to a database of reference information, a unit for processing cartographic parameters with one-way communication is connected to a database for storing telemetric information, a unit for processing cartographic parameters with two-way communication is connected with a database for reference information, a storage database one-way telemetry information is connected to the data selection unit, the data selection unit is one-way communication connected to the raw data preparation unit, the normalized data preparation unit, the data preparation unit in logarithmic coordinates, the data preparation unit for mathematical calculations, the raw data preparation unit for one-way communication is connected to the graph display unit, the normalized data preparation unit for one-way communication is connected to the graph display unit, the data preparation unit in logarithmic coordinates is connected by one-way communication with the graph display unit, the database is reference of information via a two-way communication is connected to a user list settings block, a calculation parameters editing block, an oil production parameters editing block, a one-way communication help database is connected to a correlation calculation block, a data preparation unit for mathematical calculations by a one-way communication is connected to a neural network analysis block, a mathematical calculation block standby, standard deviation calculation unit, asymmetry calculation unit, correlation calculation unit, non a one-way communication analysis is connected to the interaction vector display unit, the mathematical expectation calculation unit by one-way connections is connected to the Maryland matrix display unit and the heat map display unit, the standard deviation calculation unit by one-way connections is connected to the Maryland matrix display unit and the heat map display unit, the asymmetry calculation unit unilaterally connected to the Maryland matrix display unit and the heat map display unit, p a one-way correlation calculator is connected to a Maryland matrix display unit and a heat map display unit, a one-way graph display unit is connected to a reporting unit, a one-way interaction vector display unit is connected to a reporting unit, a Maryland matrix display unit is connected to the reporting unit by one-way , the heat map display unit is connected by one-way communication with the reporting unit.

The ability to access the system from any computer on the corporate network without installing additional software on it is achieved by using web technologies in the development of the application. The server part is a web application based on Liferay Portal technology; telemetry database and reference information database (SI), built on the basis of Oracle Database; as well as a set of PL / SQL procedures within the Oracle database that perform all operations on the data. The client part of the system is an automated workstation that allows the end user to interact with the system: enter parameters to start the statistical analysis procedures and obtain calculation results using visual forms. This approach allows not to impose high performance requirements on the user's computer, since all calculations are performed on the server.

The ability to perform all mathematical calculations on the server is achieved by the implementation of all the calculation blocks in the server part of the application within the database. The user receives only the calculation results, which allows to reduce the time of information transfer from the server side of the application to the client.

The figure 1 presents a diagram of a system of statistical and neural network analysis of telemetry data of oilfield objects, which consists of the following blocks:

1. Block processing requests for parameters

2. The processing unit of the cartographic parameters,

3. Telemetry information storage database,

4. Background information database,

5. The data selection unit,

6. Raw data preparation unit,

7. The unit for the preparation of normalized data,

8. Data preparation unit in logarithmic coordinates,

9. The data preparation unit for mathematical calculations,

10. Block neural network analysis,

11. The unit for calculating the mathematical expectation,

12. The unit for calculating the standard deviation,

13. The asymmetry calculation unit,

14. The unit for calculating the correlation,

15. The graph display unit,

16. The block display vectors of mutual influence,

17. The display unit matrix of Maryland,

18. The heat map display unit,

19. Reporting unit,

20. Block settings user lists,

21. Block for editing calculation parameters,

22. Block editing the parameters of oil production.

The parameter request processing block (1) is a set of visual forms: lists of wells, production parameters with the ability to select the necessary elements, additional design parameters, and a form for entering dates.

The block for processing cartographic parameters (2) is a topographic map with the ability to select wells by clicking on a particular well or by selecting a section for immediately selecting a group of wells.

Telemetry information storage database (DB) (3) is a telemetry data storage database. This repository is organized on the principle of storing all incoming values without any preliminary processing.

The database (DB) of reference information (4) is a database where a set of possible parameters of calculation algorithms is stored; list of oil production parameters involved in the calculations; technological schemes that store information about reservoir pressure maintenance wells located on the same pressure line; information about system users, their access rights to it.

The data selection block (5) is a set of server procedures for selecting data from the telemetry information storage database (3) in accordance with the parameters defined by the user in the Parameter Request Processing Unit (1) and the Cartographic Parameter Processing Unit (2).

The block for preparing raw data (6) is a server procedure designed to filter data. Data filtering is a selective deletion of points, so that when the received data is displayed on the graph, the nature of the curve is preserved.

The normalized data preparation unit (7) is a server procedure for filtering and normalizing time series.

The data preparation unit in logarithmic coordinates (8) is a server procedure designed to filter and convert data into logarithmic coordinates.

The data preparation unit for mathematical calculations (9) is a set of server procedures designed for pre-processing and formatting telemetry data. Data preparation includes building crosstabs, sorting, filtering, and aggregating data.

The neural network analysis unit (10) is a server procedure. This procedure uses neural network algorithms for classifying objects according to the similarity of processes occurring on them at a given time interval.

The mathematical expectation calculation unit (11) is a server procedure for calculating the mathematical expectation of sample values. The arithmetic mean is used as an estimate of the mathematical expectation of the sample.

The standard deviation calculation unit (12) is a server procedure for calculating the standard deviation of sample values. The standard deviation is the amount of dispersion of the sample values relative to its mathematical expectation.

The asymmetry calculation block (13) is a server procedure for calculating the asymmetry coefficient of sample values. This value determines the asymmetry of the data distribution.

, коэффициент ранговой корреляции Спирмена:

, где d - разности рангов каждой пары сопоставляемых значений.The correlation calculation unit (14) performs the function of calculating the correlation coefficients for a pair of oil production parameters or for one parameter at different wells in order to identify inconsistencies in the behavior of technologically related parameters for a certain period of time. The procedure allows you to calculate the Pearson correlation coefficient:

, Spearman's rank correlation coefficient:

, where d are the rank differences of each pair of matched values.

The graph display unit (15) is a presentation of data obtained during preliminary preparation by the Raw Data Preparation Unit (6), Normalized Data Preparation Unit (7) and Data Preparation Unit in logarithmic coordinates (8).

The mutual influence vector display unit (16) is a form for presenting the calculation results of the Neural Network Analysis Unit (10) and is a topographic map with similarity vectors found as a result of calculations.

The Maryland matrix display unit (17) is a form for presenting calculation results and is a hierarchical matrix of rectangles, the size of which is proportional to the value of the calculated parameters.

The heat map display unit (18) is a form for presenting calculation results. Heat map - display of calculated parameters on a topographic map in the form of color spots proportional to the value of the calculated parameter.

The reporting unit (19) is a server-side procedure for generating a document in .pdf format, which saves the results of calculations to a user's computer.

The user list settings block (20) is a server procedure for managing the list of users and their rights to work with the system. This block allows you to add, delete users of the system and edit their rights to access the system modules.

The block for editing calculation parameters (21) is a server-side procedure for managing the set of parameters used in the module of mathematical calculations. This block allows you to edit, add and delete parameters used for calculations.

The block for editing oil production parameters (22) is a server procedure for managing a set of oil production parameters involved in the work of the procedures of the mathematical calculation module. This block allows you to edit, delete, and add new oil production parameters in order to be able to make calculations for data coming from these parameters.

For convenience of presentation, the following parts and modules are introduced into the description that are of a formal nature:

- Administration module, including the User list settings block (20), Calculation parameters editing block (21), Oil production parameters editing block (22),

- A data preparation module that combines the Raw Data Preparation Unit (6), Normalized Data Preparation Unit (7), Data Preparation Unit in logarithmic coordinates (8), Data Preparation Unit for mathematical calculations (9),

- A module for mathematical calculations, including a Neural Network Analysis Unit (10), a Mathematical Expectation calculation unit (11), a standard deviation calculation unit (12), an asymmetry calculation unit (13), a correlation calculation unit (14),

- the client part, combining the Unit for processing request for parameters (1), the Unit for processing cartographic parameters (2), the Administration module, the Unit for displaying graphs (15), the Unit for displaying the interaction vectors (16), the Unit for displaying the matrix of Maryland (17), the Unit for displaying thermal maps (18), Report building block (19),

- a server part with a telemetry information storage database (3), reference information database (4), a data selection unit (5), a data preparation module and a mathematical calculation module.

The general algorithm of the system can be described as follows.

The user, through interaction with the system through the Parameter Request Processing Unit (1) and the Cartographic Parameters Processing Unit (2), receives reference information from the database (4), a set of possible parameters for calculations and selects the necessary parameters — sets the time interval for analysis, selects from the list (or on a topographic map) a set of wells, oil production parameters, and also indicates the necessary parameters for calculation methods. The data selection unit (5) selects data from the telemetry information storage database (3) in accordance with the parameters entered by the user.

To construct calculation schedules, preliminary data processing is required by the Data Preparation Module. Raw data preparation is carried out by the Raw data preparation unit (6). Data preparation for constructing normalized schedules is carried out by the Normalized Data Preparation Unit (7). Data preparation for constructing logarithmic graphs is carried out by the Data preparation unit in logarithmic coordinates (8). The construction of the described graphs on the user screen is carried out in the Graph display unit (15).

To carry out mathematical calculations, preliminary data processing is required by the Data Preparation Unit for mathematical calculations (9). Further, in accordance with the parameters of the calculation method set by the user, the prepared data is sent to one of the following blocks: Neural network analysis unit (10), Mathematical expectation calculation unit (11), Standard deviation calculation unit (12), Asymmetry calculation unit (13), Correlation calculation block (14).

The neural network analysis unit (10) uses data on the volumes of injection and production of fluid in wells related to a specific area for a certain period of time. At the output, the calculation procedure generates similarity coefficients for each pair of wells based on an analysis of ongoing processes. The results are displayed in the Interaction Influence Vector Display Unit (16) on the topographic map in the form of an overlay of vectors on the basic objects whose thickness is proportional to the value of the calculated coefficient.

The following blocks calculate the corresponding statistics for a group of wells at a given time interval: Mathematical expectation calculation block (11), standard deviation calculation block (12), asymmetry calculation block (13), correlation calculation block (14). The output of the operation of these blocks is carried out by the Maryland matrix display unit (17) and the heat map display unit (18).

The correlation calculation unit (14) uses reference information from the reference information database (4), namely, technological schemes of reservoir pressure maintenance wells located on the same pressure line.

The reporting unit (19) generates reports based on the data obtained from the following units: Chart display unit (15), Interaction vector mapping unit (16), Maryland matrix display unit (17), and Heat map display unit (18). The information obtained using these blocks is stored in the document as a static image. The document also contains all the parameters of the calculation, set by the user of the system.

Management of system operation parameters is carried out through the Administration Module. Only the administrator has access to manage this module. The user list settings block (20) allows you to add, delete users of the system and edit their rights to access the system modules. The block for editing calculation parameters (21) allows you to edit, add and delete parameters used for calculations. The block for editing oil production parameters (22) allows you to edit, delete, and add new oil production parameters in order to be able to make calculations for data coming from these parameters. The data edited in the administration module is stored in the database of reference information (4).

The system provides the user with the following data analysis capabilities:

1. The ability to access the system from any computer on the corporate network without installing additional software on it.

2. The ability to transfer to the client computer only the results of mathematical calculations without transferring the full amount of data.

3. Using a topographic map of NGOs with the ability to select objects on the map for mathematical calculations based on data from these objects.

4. Presentation of aggregated data from telemetry systems in a convenient for perception form, which allows you to quickly monitor changes in the operation of telemetry systems and changes in processes occurring at production facilities.

5. Displaying data on graphs to assess the joint dynamics of the behavior of indicators.

6. Providing information to identify failures in the operation of NGO equipment and failures in the operation of telemetry systems as a result of calculating the pressure correlation for reservoir pressure maintenance wells located on the same pressure line.

7. Providing information to identify failures in the operation of equipment at oilfield facilities of NGOs and failures in the operation of telemetry systems as a result of calculating the correlation for pairs of agreed parameters.

8. Providing information on the interaction between the wells to solve the problems of regulating technological processes, based on a neural network analysis of data from telemetry systems.

Claims (1)

The system of statistical and neural network analysis of telemetry data of oilfield facilities, including a data selection unit, a telemetry information storage database, a user list settings unit, a parameter request processing unit, a reporting unit, a neural network analysis unit, a mathematical expectation calculation unit, a standard deviation calculation unit, a unit asymmetry calculation, correlation calculation unit, graph display unit, mutual influence vector display unit, Meryl matrix display unit NDA, a heat map display unit, characterized in that the system additionally includes a database of reference information, a unit for processing map parameters, a unit for preparing raw data, a unit for preparing normalized data, a unit for preparing data in logarithmic coordinates, a unit for preparing data for mathematical calculations, an editing unit calculation parameters, a block for calculating oil production parameters, and the block for processing parameter requests by one-way communication is connected to the telemeter storage database of information, the block for processing parameter requests with two-way communication is connected to the database of reference information, the block for processing map parameters with one-way communication is connected to the database for storing telemetric information, the block for processing cartographic parameters for two-way communication is connected to the database of reference information, the database for storing telemetric information with one-way communication connected to the data selection unit, the data selection unit is unilaterally connected to the preparation unit data, a unit for preparing normalized data, a unit for preparing data in logarithmic coordinates, a unit for preparing data for mathematical calculations, a unit for preparing raw data with one-way communication is connected to a block for displaying graphs, a unit for preparing normalized data for one-way communication is connected to a block for displaying graphs, a unit for preparing data in logarithmic coordinates one-way communication connected to the graph display unit, the database of information reference two-way communication connected to the user list settings block, calculation parameters editing block, oil production parameters editing block, one-way communication help database connected to the correlation calculation block, data preparation for mathematical calculations block by one-way communication connected to the neural network analysis block, mathematical expectation block, calculation block standard deviation, asymmetry calculation unit, correlation calculation unit, neural network analysis unit, one-way the first connection is connected to the display unit of the mutual interaction vectors, the mathematical expectation calculation unit by one-way connections is connected to the Maryland matrix display unit and the heat map display unit, the standard deviation calculation unit by one-way connections is connected to the Maryland matrix display unit and the heat map display unit, the asymmetry calculation unit is one-way connections connected to the Maryland matrix display unit and the heat map display unit, the correlation calculation unit is one-way connected to a Maryland matrix display unit and a heat map display unit, a one-way graph display unit is connected to a report building unit, a one-way communication vector display unit is connected to a reporting unit, a Maryland matrix display unit is connected to a report building unit, a display unit one-way heat map is connected to the reporting unit.

Images