Cognitive Model Based Prediction of the Object’s Response to External Inputs
DOI:
https://doi.org/10.24160/1993-6982-2018-5-89-95Keywords:
cognitive model, strength of inter-factor link, linear regression, correlation, data analysis, process prediction, controlled object stabilityationAbstract
Matters concerned with the use of cognitive models in control systems and in decision support systems are considered. Prerequisites and constraints for the problem of quantitatively predicting the evolvement of processes in the controlled object are formulated for the case when one of the factors in the model changes for some reason. The case when the cognitive model is constructed using the knowledge of experts about the controlled object in the absence of observational data made at this object is considered. It is assumed that the factors included in the model are only measured by means of quantitative scales and provide a sufficiently good reflection of the processes occurring in the modeled object. A statistical representation of the cognitive model, in which links between the factors are described by linear regression dependences, is used. The change of factors is predicted using the method that was previously applied to produce simulation data for the cognitive model. The method uses statistical characteristics of factors and estimated strength of links between the factors that are considered known to experts. Estimates of regression dependence parameters are calculated. It is assumed that the changes in the influencing factors of each link are transferred to neighboring dependent factors within a single clock cycle. As a result, changes in the factors propagate over the cognitive model graph in a wave-like manner. The predicted changes of factors for three cognitive models are presented. The first of them demonstrates the developed method for calculating the prediction for one link of the "many to one" type. The second model describes in a simplified form the relationship between research activities and educational process in universities of Russia. This model shows how the variation of factors evolves with time and demonstrates the possibility of keeping the controlled object stability even if its cognitive model contains positive cycles. The third model illustrates, taking a hypothetical object as an example, that instability may occur in the controlled object operation if its cognitive model contains certain link configurations which form cycles.
References
2. Полотнов М.М., Фомин Г.А. Методы построения и использования когнитивных моделей объектов принятия решений. М.: Изд-во МЭИ, 2015.
3. Анадиотис Дж. Путь от данных к аналитике и ИИ: от наглядного отображения к прогнозам // PC Week. 2016. № 21 (920) [Электрон. ресурс] https://www pcweek.ru/ai/article/detail.php?ID=190406 ¶m=mail (дата обращения 06.12.2016).
4. Кульба В.В., Миронов П.Б., Назаретов В.М. Анализ устойчивости социально-экономических систем с использованием знаковых орграфов // Автоматика и телемеханика. 1993. № 7. С. 130—137.
5. Радченко С.А. Когнитивное моделирование как средство поддержки принятия решений при управлении социально-экономической системой // Проблемы регионального управления, экономики и права и инновационных процессов в образовании: Труды III Междунар. науч.-практ. конф. Таганрог: Изд-во ТИУиЭ, 2003. С. 298—300.
6. Федулов Ю.Г., Юсов А.Б., Матвеев А.А. Исследование социально-экономических и политических процессов с помощью когнитивных моделей. М: Изд-во РАГС, 2004.
7. Федулов А.С., Борисов В.В. Анализ нечетких реляционных когнитивных карт // Нейрокомпьютеры: разработка, применение. 2016. № 7. С. 7—14.
8. Борисов В.В., Денисенков М.А., Федулов А.С. Нечеткие ситуационные сети на основе когнитивных карт // Естественные и технические науки. 2016. № 2 (92). С. 118—123.
9. Борисов В.В., Федулов А.С., Федулов Я.А. «Совместимые» нечеткие когнитивные модели: прямые и обратные задачи // Нечеткие системы и мягкие вычисления. 2016. № 2 (11). С. 103—114.
10. Фомин Г.А., Фомина Е.С. Метод формирования имитационных данных по когнитивной модели объекта управления // Вестник МЭИ. 2018. № 1. С. 106—111.
---
Для цитирования: Фомин Г.А. Прогнозирование по когнитивной модели реакции объекта на внешние воздействия // Вестник МЭИ. 2018. № 5. С. 89—95. DOI: 10.24160/1993-6982-2018-5-89
#
1. Trakhtengerts E.A. Komp'yuternaya Podderzhka Prinyatiya Resheniy. M.: SINTEG, 1998. (in Russian).
2. Polotnov M.M., Fomin G.A. Metody Postroeniya i Ispol'zovaniya Kognitivnykh Modeley Ob′ektov Prinyatiya Resheniy. M.: Izd-vo MPEI, 2015. (in Russian).
3. Anadiotis Dzh. Put' ot Dannykh k Analitike i II: ot Naglyadnogo Otobrazheniya k Prognozam. PC Week. 2016;21 (920) [Elektron. Resurs] https://www.pcweek. ru/ai/article/detail.php?ID=190406 ¶m=mail (Data Obrashcheniya 06.12.2016). (in Russian).
4. Kul'ba V.V., Mironov P.B., Nazaretov V.M. Analiz Ustoychivosti Sotsial'no-ekonomicheskikh Sistem s Ispol'zovaniem Znakovykh Orgrafov. Avtomatika i Telemekhanika. 1993;7:130—137. (in Russian).
5. Radchenko S.A. Kognitivnoe Modelirovanie kak Sredstvo Podderzhki Prinyatiya Resheniy pri Upravlenii Sotsial'no-ekonomicheskoy Sistemoy. Problemy Regional'nogo Upravleniya, Ekonomiki i Prava i Innovatsionnykh Protsessov v Obrazovanii: Trudy III Mezhdunar. Nauch.-prakt. Konf. Taganrog: Izd-vo TIUiE, 2003:298—300. (in Russian).
6. Fedulov Yu.G., Yusov A.B., Matveev A.A. Issledovanie Sotsial'no-ekonomicheskikh i Politicheskikh Protsessov s Pomoshch'yu Kognitivnykh Modeley. M: Izd- vo RAGS, 2004. (in Russian).
7. Fedulov A.S., Borisov V.V. Analiz Nechetkikh Relyatsionnykh Kognitivnykh Kart. Neyrokomp'yutery: Razrabotka, Primenenie. 2016;7:7—14. (in Russian).
8. Borisov V.V., Denisenkov M.A., Fedulov A.S. Nechetkie Situatsionnye Seti na Osnove Kognitivnykh Kart. Estestvennye i Tekhnicheskie Nauki. 2016;2 (92): 118—123. (in Russian).
9. Borisov V.V., Fedulov A.S., Fedulov Ya.A. «Sovmestimye» Nechetkie Kognitivnye Modeli: Pryamye i Vychisleniya. 2016;2 (11):103—114. (in Russian).
10. Fomin G.A., Fomina E.S. A Method for Shaping Simulation Data from a Controlled Plant Cognitive Model. Vestnik MPEI. 2018;1:106—111. (in Russian).
---
For citation: Fomin G.A. Cognitive Model Based Prediction of the Object’s Response to External Inputs. MPEI Vestnik. 2018;5:89—95. (in Russian). DOI: 10.24160/1993-6982-2018-5-89-95.
