Evgeniy A. Titenko¹, Tatyana I. Lapina², Oleg G. Dobroserdov³, Alexey N. Schitov4, Evgeny V. Taldykin5

1−5 South-West State University (Kursk, Russia) 1 2 3

 johntit@mail.ru,  lapinati@mail.ru,  serfingk@yandex.ru

The object of the research is informational search processes by a pattern using the crossing operation used in genetic algorithms. The aim of the study is to reduce the search time for a pattern in production systems by introducing a two-point crossing operation in the matching stage.

The work is devoted to considering an important subclass of artificial intelligence systems – production systems. First of all, production computing is applicable in dynamic expert systems such as ECLIPSE. They primarily work with partially defined data. The work considers samples (templates) as discrete structures. These structures contain constant and variable fragments, which determines the ambiguity of their description and the uncertainty of the search. It is shown, that the most critical operation is the pattern search. A comparative analysis of the known pattern search algorithms for production systems has been carried out. The known pattern search algorithms do not correspond to such indicators as hardware orientation, adaptation.

The research method is based on the use of genetic algorithms for the search operation on the sample. The search operation for a given pattern and the operation of crossing chromosomes have the same content. The study is based on identifying matching descriptions of models. The basic operation of genetic algorithms (crossover) allows you to describe the processes of searching for a sample and further modifying a string. For this, a two-point crossover format is selected. In this case, the chromosome, understood as a string, is divided into three fragments. In symbolic processing, they correspond to dividing strings into prefix, body, and suffix. The originality of genetic algorithms for the task of searching by pattern consists in changing the interpretation of the positions (addresses) of occurrences. These positions are calculated by search pattern algorithms and are not known in advance in production systems. They are understood as input that can be arbitrarily set or selected in genetic algorithms. This fundamental point determines the storage and processing of previously calculated items. Such an approach justifies the functional commonality of pattern search and crossover operations. This commonality allows iterations of the production system to use the stored positions of the sample without spending time searching.

The theoretical development of the expert system was carried out taking into account the combination of matching operations and point-to-point crossing. The commonality allowed us to expand the stage of comparison in parallel with the crossing process. This new information from the crossing operation is entered into the matching and conflict resolution blocks. A modified abstract computing machine has been developed. It is expanded by a unit for performing genetic crossing operations. The novelty of such an abstract computing machine is that it calculates additional information. This additional information (index arrays) expands the capabilities of production systems and allows you to remove uncertainty and choose priority products and the appropriate strategy for computing.

The use of a combination of a search algorithm and operations of genetic algorithms has a predominant field of application for two- and four-character samples when processing strings up to 40 characters long, which corresponds to most typical production algorithms.

1. Wichert A. Artificial intelligence and a universal quantum computer. AI Communications. 2016. V. 29. № 4. P. 537−543.
2. Waterman D.A. A Guide to Expert Systems. Reading, MA: Addison–Wesley. 1996. 167 p.
3. Wang J.-J., Qiao Y., Xiong J.-Q., Wang H.-A. Method for Graph-based Real-time Rule Scheduling in Multi-core Environment. Journal of Software. V. 30. № 2. P. 481−494.
4. Wahid F., Ismail L.H., Ghazali R., Aamir M. An efficient artificial intelligence hybrid approach for energy management in intelligent buildings. KSII Transactions on Internet and Information Systems. 2019. V. 13. № 12. P. 5904−5927.
5. Briola D., Mascardi V., Martelli M., Arecco G., Caccia R. A prolog-based MAS for railway signalling monitoring: Implementation and experiments. 9th Workshop «From Objects to Agents». WOA 2008 Evolution of Agent Development: Methodologies, Tools, Platforms and Languages. 2008. P. 11−18.
6. Lyuger Dzh.F. Iskusstvennyi intellekt: strategii i metody resheniya slozhnykh problem. M.: «Vilyams». 2003. 864 s. (In Russian).
7. Titenko E.A., Kripachev A.V., Marukhlenko A.L. Kommutatsionnaya skhema parallelnykh parnykh perestanovok dlya spetsializirovannogo produktsionnogo ustroistva. Izvestiya YuFU. Ser. «Tekhnicheskie nauki». 2018. № 8(203). S. 29−38. (In Russian).
8. Titenko E.A., Emelyanov S.G., Kurochkin A.G. Analiz algoritmov poiska po obraztsu dlya upravleniya gruppoi robotov. Naukoemkie tekhnologii. 2014. T. 15. № 12. S. 4−8. (In Russian).
9. Wu S., Manber U. Fast text searching allowing errors. Communications of the ACM. 1992. V. 35. № 10. P. 83−91.
10. Bioinspirirovannye metody optimizatsii. M.: Fizmatlit. 2009. 384 s. (In Russian).
11. Wahid F., Ismail L.H., Ghazali R., Aamir M. An efficient artificial intelligence hybrid approach for energy management in intelligent buildings. KSII Transactions on Internet and Information Systems. 2019. V. 13. № 12. P. 5904−5927.
12. Xiong Y., Ling Q.-H., Han F., Liu Q.-H. An efficient gene selection method for microarray data based on LASSO and BPSO. BMC Bioinformatics. 2019. V. 20. P. 212003.
13. Lysenko S.N., Makaev V.A., KHalin Yu.A Issledovanie rezultatov meta-poiska na osnove geneticheskogo algoritma. Estestvennye i tekhnicheskie nauki. 2018. № 8(122). S. 171−173. (In Russian).
14. Titenko E.A., Dobroserdov O.G. Metod i modifitsirovannyi algoritm napravlennogo poiska vkhozhdenii. Informatsionno-izmeritelnye i upravlyayushchie sistemy. 2009. T. 7. № 11. S. 98−100. (In Russian).
15. Titenko E.A., Zerin I.S. Ischislitelnaya sistema produktsii i protsedura raspoznavaniya konfliktov dannykh. Vestnik kompyuternykh i informatsionnykh tekhnologii. 2012. № 6(96). S. 50−55. (In Russian).
16. Makkonell Dzh. Osnovy sovremennykh algoritmov. Izd. 2-e, dop.. Per. s angl. pod red. S.K. Lando. M.: Tekhnosfera. 2004. 386 s. (In Russian).
17. Holland J.H. Adaptation in natural and artificial systems. An introductory analysis with application to biology, control, and artificial intelligence. London: Bradford book edition. 1994. 211 p.
18. Titenko E.A., Atakishchev O.I. Metod assotsiativnoi obrabotki strok i apparatno-orientirovannyi algoritm dlya ego realizatsii. Izvestiya Yugo-Zapadnogo gosudarstvennogo universiteta. 2011. № 6(39). Ch. 2. S. 72−77. (In Russian).
19. Titenko E.A., Petrik E.A., Voronin D.A., Atakishcheva I.V. Produktsionnaya model dlya parallelnoi obrabotki znanii. Informatsionnoizmeritelnye i upravlyayushchie sistemy. 2011. T. 9. № 11. S. 81−86. (In Russian).
20. Titenko E.A., Degtyarev S.V. Approximate search in the sample on the basis manber-wu method. Journal of Fundamental and Applied Sciences. 2017. V. 9. № 2. P. 914−918.