Journal "Software Engineering"
a journal on theoretical and applied science and technology
ISSN 2220-3397

Issue N2 2026 year

DOI: 10.17587/prin.17.83-94
Synthesis of Iterative Plans for Solving Declarative Set Problems using the Planning Artificial Neural Network
M. F. Stepanov, Professor, mfstepanov@mail.ru, A. M. Stepanov, Senior Lecturer, amstepanov@mail.ru, O. M. Stepanova, Postgraduate Student, omstepanova@mail.ru, Yuri Gagarin State Technical University of Saratov, Saratov, 410054, Russian Federation
Corresponding author: Mikhail F. Stepanov, Professor, Yuri Gagarin State Technical University of Saratov, Saratov, 410054, Russian Federation, E-mail: mfstepanov@mail.ru
Received on July 30, 2025
Accepted on September 09, 2025

An approach to modifying an intelligent solver based on planning artificial neural networks is proposed to enable action planning for declaratively defined tasks solved by iterative methods. Planning actions to solve a declaratively defined task is based on a knowledge model of the problem domain, which is used in this article as a knowledge model of the methods of the theory of automatic control. Solving declaratively defined tasks using a planning artificial neural network involves two stages: selecting relevant axioms (operations) of the knowledge model and determining the order of their execution. The issues related to solving the problem of building action plans, including cycles, necessary for the implementation of iterative methods for solving problems in the problem domain, are considered. Since the problem of synthesizing cyclic programs is generally unsolvable, the introduction of two auxiliary structures is proposed to develop a large-block approach to program synthesis: a selector and an iterator. The selector is designed to select operations to be cyclically repeated using a predefined iterative process scheme (iterator) at the stage of executing the generated action plan. The basis of the proposed approach to the construction and organization of the execution of action plans for solving problems by iterative methods is the introduction of auxiliary properties and/or relationships in the knowledge model, describing the conditions for the repeated execution of certain operations of the generated action plan to solve the problem. The specified properties (relations) act as additional conditions for the applicability of the operations of the generated action plan. They are used in the algorithmic schemes of the executive subsystem of the problem solver to organize the cyclic execution of the operations provided with them in the action plan for the implementation of an iterative solution to the problem. For the generated action plan, presentation schemes have been developed in the form of products, as well as in the form of a program record according to A. A. Lyapunov. An illustrative example of constructing a cyclically executed action plan for solving a subtask within the framework of a control task under conditions of uncertainty is considered.

Keywords: action planning, declaratively defined tasks, synthesis of iterative plans of problem solving, neural network intelligent solver, planning artificial neural networks, knowledge representation
pp. 83—94
For citation:
Stepanov M. F., Stepanov A. M., Stepanova O. M. Synthesis of Iterative Plans for Solving Declarative Set Problems using the Planning Artificial Neural Network, Programmnaya Ingeneria, 2026, vol. 17, no. 2, pp. 83—94. DOI: 10.17587/prin.17.83-94.
The study was supported by the Russian Science Foundation grant No. 24-21-00488, https://rscf.ru/project/24-21-00488/
References:
  1. Potemkin V. G. System of engineering and scientific calculations MATLAB 5.x, Moscow, DIALOG-MIFI, 1999 (in Russian).
  2. Markov A. A. The impossibility of certain algorithms in the theory of associative systems, Dokl. AN SSSR, 1947, vol. 55, pp. 587—590 (in Russian).
  3. Nepejvoda N. N. Relationship between natural inference rules and operators of high-level algorithmic languages, Dokl. AN SSSR, 1978, vol. 239, no. 3, pp. 526—529 (in Russian).
  4. Lavrov S. S., Zalogova L. A., Petrushina T. I. Principles of planning of problem solving in the system of automatic program synthesis, Programmirovanie. 1982, no. 3, pp. 35—43 (in Russian).
  5. Tyugu Je. H. Conceptual programming, Moscow, Nauka, 1984, 256 p. (in Russian).
  6. La Valle S. M. Planning Algorithms, Cambridge, Cambridge University Press, 2006, 844 p.
  7. Ramirez-Rueda R., Benitez-Guerrero E., Mezura-Godoy C., Barcenas E. A. A Decade of Advancements in Program Synthesis from Natural Language: A Systematic Literature Review, Proceedings of the Institute for System Programming of the RAS, 2024, vol. 36, no. 6, pp. 59—82. DOI: 10.15514/ISPRAS-2024-36(6)-4.
  8. Zuenko A. A., Fridman O. V. The development of a logical approach to solving the problems of intelligent planning: an analytical review, Vestnik Voronezhskogo gosudarstvennogo universiteta. Seriya: Sistemny'j analiz i informacionnye texnologii, 2023, vol. 4, pp. 104—127. DOI: 10.17308/sait/1995-5499/2023/4/104-127.
  9. Zajcev A. F. A method of searching and logical deduction of expert information in an oriented cyclic multigraph of knowledge, Informacionnye i matematicheskie texnologii v nauke i upravlenii, 2022, vol. 4 (28), pp. 213—222. DOI: 10.38028/ESI.2022.28.4.017.
  10. Joudakizadeh M., Beltiukov A. P. Adaptive human-machine theorem proving system, Proceedings of the Institute of Mathematics and Informatics at Udmurt State University, 2024, vol. 64, pp. 17—33.DOI: 10.35634/2226-3594-2024-64-02.
  11. Joudakizadeh M., Beltiukov A. P. Two-level realization of logical formulas for deductive program synthesis, Bulletin of Udmurt University. Mathematics. Mechanics. Computer Science, 2024, vol. 34, no. 4, pp. 469—485. DOI: 10.35634/vm240401.
  12. Sidorenko O. I. On direct deductive conclusions in traditional integral syllogistics from 50 basic propositions, Matematicheskie metody v tekhnike i tekhnologiyakh, 2020, vol. 7, pp. 39—45 (in Russian).
  13. Reference book on automatic control theory / Ed. A. A. Krasovskij. Moscow, Nauka, 1987, 712 p. (in Russian).
  14. Methods of classical and modern automatic control theory / Eds K. A. Pupkov, N. D. Egupov, Moscow, Izd-vo MGTU im. N. Je.Baumana, 2004, 656 p. (in Russian).
  15. Aleksandrov A. G. Synthesis of regulators of multidimensional systems, Moscow, Mashinostroenie, 1986, 272 p. (in Russian).
  16. Poljak B. T., Shherbakov P. S. Robust stability and control, Moscow, Nauka, 2002, 303 p. (in Russian).
  17. Poljak B. T., Hlebnikov M. V., Rapoport L. B. Mathematical theory of automatic control, Moscow, LENAND, 2019, 500 p. (in Russian).
  18. Marchuk G. I. Selected works, in 5 vol., Moscow, Rossijskaya akademiya nauk, 2018. (in Russian).
  19. Novikov A. I. Elements of discrete mathematics, Moscow, Dashkov i Ko, 2019. 209 p. (in Russian).
  20. Kameneczkaya N. V. Computational mathematics, Saint-Peterburg, SPb University GPS MChS, 2018, 116 p. (in Russian).
  21. Novikov F. A. Discrete mathematics for programmers, Saint-Peterburg, Piter, 2007, 364 p. (in Russian).
  22. Stepanov M. F. Automatic solution of automatic control theory problems on the basis of planning artificial neural networks, Nejrokomp'jutery: razrabotka iprimenenie, 2003, no. 3, 4, pp. 27—44 (in Russian).
  23. Stepanov M. F. Neural networks for planning the solution of automatic control theory problems, Problemy upravlenija, 2004, no. 2, pp. 66—71 (in Russian).
  24. Stepanov M. F., Stepanov A. M. Mathematical modelling of intellectual self-organizing automatic control system: Action planning research, Procedia Engineering, 2017, vol. 201, pp. 617—622. DOI: 10.1016/j.proeng.2017.09.657.
  25. Efimov E. I. Intellectual problem solvers, Moscow, Nauka,1982, 316 p. (in Russian).
  26. Gencen G. Research on logical inference, Matematicheskaja teorija logicheskogo vyvoda, Moscow, Nauka, 1967, pp. 9—76 (in Russian).
  27. Ljapunov A. A. About program logic diagrams, Problemy kibernetiki, no. 1. Moscow, Fizmatgiz, 1958, pp. 46—74 (in Russian).
  28. Aleksandrov A. G. Automatic control system design methods, Moscow, Izd-vo fis.-mat. liter., 2008, 232 p. (in Russian).
  29. Stepanov M. F., Stepanov A. M., Stepanova T. V., Stepanova O. M. Features of Knowledge Representation of Automatic Control Systems Design Methods, Artificial Intelligence in Models, Methods and Applications. Studies in Systems, Decision and Control, Springer Nature Switzerland AG, 2023, vol. 457, pp. 231—243. DOI: 10.1007/978-3-031-22938-1_16.
  30. Stepanov M. F., Stepanov A. M., Stepanova O. M. On the approach to synthesizing plans for solving declaratively defined tasks solved by iterative methods, Matematicheskie metody v tekhnologiyakh i tekhnike, 2025, no. 8, pp. 93—98 (in Russian)