Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.cplire.ru/Lab144/start/e_oil.html
Дата изменения: Mon Sep 24 15:02:27 2007
Дата индексирования: Mon Oct 1 23:24:04 2012
Кодировка:
Поисковые слова: п п п п п п п п п п п п п п п п п п п п п п п р п р п р п р п р п р п р п р п р п р п р п р п р п р п р п
Getting Started in Actor Prolog. Oil Production.

Oil Production

In this chapter we will consider an example of visual expert system written on the Actor Prolog language.

The expert system is intended for solving of the following problem. There are a lot of methods of oil field bed stimulation that increase output of the oil production: thermal, physical-chemical, gas, microbiological, and others. Selection of a method appropriated for given oil field is complex and very important stage of development of oil field. In the course of selection of the method one must account a lot of parameters like rock characteristics, reservoir characteristics, oil characteristics, reservoir water characteristics, native gas characteristics, bedded substances, mineral constituents, and others. A mistake in the selection of bed stimulation method can led to big expenses and material losses even before the beginning of the oil production because modern methods of bed stimulation require expensive reagents and apparatus.

This expert system helps one to estimate the appropriateness of some known methods of bed stimulation on early stages of exploration of oil field, under the conditions of uncertainty and absence of some crucial information.

The expert system implements the following principles of data analysis:

  1. Visualization: The system implements a graphical representation of the structure and parameters of decision making procedure. The visual representation of data domain is implemented by the means of SADT. Actor Prolog implements coloring of the blocks of SADT model. In particular, the blocks that designate methods of bed stimulation appropriated for given oil field parameters are automatically marked by green color, and the methods that are unsuitable are marked by red color. This feature makes it possible to see the decision making process as a whole and to observe how the appropriateness of different methods is changed in the course of modification of the parameters of the oil field.
  2. Interactivity: Any modification of a parameter of the oil field in the dialog box will automatically cause a repeated analysis of the SADT diagrams and revision of the results of the logical inference. The color of blocks of the diagrams changes, and refined list of appropriated bed stimulation methods is printed in the "Display solution" window. This feature of the expert system makes it possible to discover rapidly and easy what bed stimulation methods are appropriated for various values of parameters of the oil field. In addition, a user can take a detailed explanation of what parameters of the oil field exactly are not fit with a certain method.

    Let us start the expert system.



Fig. 1. Start of the expert system.

The Top-Most diagram consists of three blocks: the "Describe parameters of hydrocarbon field" block, the "Check adaptability of existed methods of oil production" block, and the "Display solution" block.



Fig. 2. Selection of bed stimulation method.

Open the "Describe parameters of hydrocarbon field" window. The parameters of the oil field are grouped in separate blocks.



Fig. 3. Describing of the parameters of hydrocarbon field.

Open blocks and enter the parameters of the oil field that you are know. If you do not know some parameters, left corresponding fields blank ones.



Fig. 4. Input of the oil field parameters.

As you will define concretely the parameters of the oil field, the expert system will refine its exploration of the adaptability of the bed stimulation methods.

The following bed stimulation methods are described in the expert system now:



Fig. 5. A taxonomy of the oil field bed stimulation methods.

 



Fig. 6. Thermal methods.

 



Fig. 7. Physical-chemical methods.

 



Fig. 8. The first group of physical-chemical methods.

 



Fig. 9. The second group of physical-chemical methods.

 



Fig. 10. Gas methods.

 



Fig. 11. Nitrogen injection.

 



Fig. 12. Carbon dioxide injection.

 



Fig. 13. Microbiological methods.

 



Fig. 14. Biopolymer waterflooding.

The blocks of the diagrams that correspond to the appropriate bed stimulation methods are automatically colored green. If some doubts are cast upon the appropriateness of a method, corresponding block is marked by red color.

At the same time the expert system outputs a list of appropriated bed stimulation methods to the "Display solution" text window.



Fig. 15. A list of appropriated bed stimulation methods.

At the left mouse click the system outputs an explanation of its solution.



Fig. 16. An explanation of elaborated decision.

In the course of the analysis of the oil field it is recommended to open several diagrams corresponding to the methods that are interesting for you at one time and order them by the "Window | Horizontal Tile" menu command.



Fig. 17. Multiple-windows user interface.

After this one can observe upon the modification of the colors of corresponding diagram blocks in the course of modification of the oil field parameters.



Fig. 18. Visualization of the decision making procedure.

This expert system is a commercial product and is not included into the standard package of Actor Prolog.

Application of Actor Prolog makes it possible to use advantages of the object-oriented logic programming, namely:

  1. Actor Prolog makes it possible to implement principles of abstraction and inheritance in the development of logical models and programs. In addition, it supports component-oriented programming and particularly a reuse of logical descriptions of SADT blocks.
  2. Actor Prolog extensively automates development of logical models and programs with the help of translation of graphical diagrams.
  3. Actor Prolog makes a programmer free from the necessity of manual programming of event handling of the graphical user interface.
All these factors make the Actor Prolog language a powerful tool for rapid prototyping and programming of visual expert systems.

Table of content