Shaping the Membership Functions in Fuzzy Logic Inference Based Diagnostic Systems
Abstract
The article presents the results from studying the performance quality of a test diagnostic system involving the most characteristic mutual arrangements of the sets containing the pathology sign values and depending on the types of membership functions. A statistical model of the object was used to study the diagnostic results. Systems consisting of two, three, and a group of terms were considered. A procedure for determining the optimal number of terms in the group was elaborated. The study has shown that the group of terms yields a greater number of accurate results; however, there exist errors between the peaks of active terms. To resolve this problem, it was suggested to unite the groups of terms into modified terms that are specific for each problem. Thus, two kinds of uniting were proposed: trapezoidal and triangular. With trapezoid terms, a greater number of correct diagnostic results were obtained; however, a large number of disputable diagnostic results remained. For removing disputable diagnostic results, triangular terms were studied. A system containing an individual triangular term was found to be the best one for the considered object model. On the other hand, the choice of the number and type of terms depends on the initial data, namely, on the size of distribution areas, on their mutual arrangement, version, and range of their intersection. Also, the use of a system involving an individual triangular shape made it possible to shape the terms so that the final term coincided with the lower order system terms.
References
2. Goher K., Tokhi M., Siddique N. Dynamic Modeling and Control of a Two Wheeled Robotic Vehicle with a Virtual Payload // ARPN J. Eng. and Appl. Sci. 2011. No. 3. Pp. 7—41.
3. Zadeh L.A. Fuzzy Logic, Neural Networks, and Soft Computing // Communications of the ACM. 1994. V. 37. No. 3. Pp. 77—84.
4. Асаи К. и др. Прикладные нечеткие системы. М.: Мир, 1993.
5. Колосов О.С., Анисимов Д.Н., Хрипков Д.В. Исследование многоуровневых нечетких диагностических систем с использованием стохастической модели // Мехатроника, автоматизация, управление. 2015. Т. 16. № 4. С. 254—261.
6. Мелихов А.Н., Бернштейн Л.С., Коровин С.Я. Ситуационные советующие системы с нечеткой логикой. М.: Наука, 1990.
7. Анисимов Д.Н. и др. Диагностика текущего состояния динамических объектов и систем сложной структуры методами нечеткой логики с использованием имитационных моделей // Искусственный интеллект и принятие решений. 2012. № 3. С. 39—50.
8. Анисимов Д.Н. Формализация процедуры исключения транзитивно замыкающих дуг при организации иерархической структуры на множестве нечетких ситуаций // Вестник МЭИ. 2011. № 4. С. 34—40.
---
Для цитирования: Хрипков Д.В. Формирование функций принадлежности в системах диагностики на базе нечеткого логического вывода // Вестник МЭИ. 2017. № 4. С. 110—116. DOI: 10.24160/1993-6982-2017-4-110-116.
#
1. Norvich A.M., Tursken I.B. Postroenie Funkcij Prinadlezhnosti. Nechetkie Mnozhestva i Teoriya Vozmozhnostej. M.: Radio i Svyaz', 1986:64—71. (in Russian).
2. Goher K., Tokhi M., Siddique N. Dynamic Modeling and Control of a Two Wheeled Robotic Vehicle with a Virtual Payload. ARPN J. Eng. and Appl. Sci. 2011;3:7—41.
3. Zadeh L.A. Fuzzy Logic, Neural Networks, and Soft Computing. Communications of the ACM. 1994;37;3:77—84.
4. Asai K. i dr. Prikladnye Nechetkie Sistemy. M.: Mir, 1993. (in Russian).
5. Kolosov O.S., Anisimov D.N., Hripkov D.V. Issledovanie Mnogourovnevyh Nechetkih Diagnostiches- kih Sistem s Ispol'zovaniem Stohasticheskoj Modeli. Mekhatronika, Avtomatizaciya, Upravlenie. 2015;16;4:254—261. (in Russian).
6. Melihov A.N., Bernshtejn L.S., Korovin S.Ya. Situacionnye Sovetuyushchie Sistemy s Nechetkoj Logikoj. M.: Nauka, 1990. (in Russian).
7. Anisimov D.N. i dr. Diagnostika Tekushchego Sostoyaniya Dinamicheskih Ob'ektov i Sistem Slozhnoj Struktury Metodami Nechetkoj Logiki s Ispol'zovaniem Imitacionnyh Modelej. Iskusstvennyj Intellekt i Prinyatie Reshenij. 2012;3:39—50. (in Russian).
8. Anisimov D.N. Formalizaciya Procedury Isklyucheniya Tranzitivno Zamykayushchih Dug pri Organizacii Ierarhicheskoj Struktury na Mnozhestve Nechetkih Situacij. MPEI Vestnik. 2011;4:34—40. (in Russian).
---
For citation: Khripkov D.V. Shaping the Membership Functions in Fuzzy Logic Inference Based Diagnostic Systems. MPEI Vestnik.2017; 4: 110—116. (in Russian). DOI: 10.24160/1993-6982-2017-4-110-116.
