Abstract
The paper substantiates a method for creating training datasets for fuzzy neural networks, which can be used to promptly obtain probabilistic estimates for the causes of abnormal critical events or incidents in diagnostic systems. The rules for converting the hypotheses on potential incident causes into intervals of defect probability in a process chain at a certain stage of continuous production are considered using belief functions. We propose a procedure for converting these hypotheses into a database of fuzzy production rules automatically, which provides training an adaptive neural network based on the Takagi–Sugeno–Kang fuzzy inference system. This makes it possible to quickly calculate a relatively accurate probabilistic estimate of a malfunction in the process chain without using expensive computing resources.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
REFERENCES
S. Cuervo, TensorFlow implementation of the Adaptive Neuro-Based Fuzzy Inference System (ANFIS). https://github.com/tiagoCuervo/TensorANFIS. Cited January 24, 2023.
S. Ferson and K. Sentz, Combination of Evidence in Dempster-Shafer Theory, Report SAND 2002-0835 (Sandia National Laboratories, 2002).
G. Gregor, An implementation of Adaptive-Network-Based Fuzzy Inference System (ANFIS) based on Keras on top of TensorFlow. https://github.com/gregorLen/AnfisTensorflow2.0. Cited January 24, 2023.
V. K. Ivanov, I. V. Obraztsov, and B. V. Palyukh, “Implementing an expert system to evaluate technical solutions innovativeness,” Software Syst. 32, 696–707 (2019). https://doi.org/10.15827/0236-235X.128.696-70
V. K. Ivanov and B. V. Palyukh, “A software platform demonstrator for configuring ANFIS neural network hyperparameters in fuzzy systems,” Software Syst. 35, 609–617 (2022). https://doi.org/10.15827/0236-235X.140.609-617
V. K. Ivanov, B. V. Palyukh, and A. N. Sotnikov, “Generation of production rules with belief functions to train fuzzy neural network in diagnostic system,” Lobachevskii J. Math. 43, 2853–2862 (2022). https://doi.org/10.1134/s1995080222130169
V. K. Ivanov, B. V. Palyukh, and A. N. Sotnikov, “Evidence theory for complex engineering system analyses,” Adv. Intell. Syst. Comput 874, 70–79 (2019). https://doi.org/10.1007/978-3-030-01818-4_7
J.-S. R. Jang, “ANFIS: Adaptive-network-based fuzzy inference system,” IEEE Trans. Syst., Man, Cybern. 23, 665–685 (1993). https://doi.org/10.1109/21.256541
E. A. Khalov, “The regular review of crisp one-dimensional membership functions of intellectual systems,” Inf. Tekhnol. Vychslitel’nye Sist. 3, 60–74 (2009).
A. E. Lepskiy, “Analysis of information inconsistency in belief function theory. Part I: External conflict,” Probl. Upr. 5, 2–16 (2021). https://doi.org/10.25728/cs.2021.5.1
T. Reineking, A Python library for performing calculations in the Dempster-Shafer theory of evidence. https://github.com/reineking/pyds. Cited January 24, 2023.
Classic Works of the Dempster-Shafer Theory of Belief Functions, Ed. by R. R. Yager and L. Liu, Studies in Fuzziness and Soft Computing, Vol. 219 (Springer, Berlin, 2008). https://doi.org/10.1007/978-3-540-44792-4
Funding
This work was supported by the Russian Foundation for Basic Research (grant no. 20-07-00199).
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
The authors declare that they have no conflicts of interest.
Additional information
Vladimir Konstantinovich Ivanov was born in 1956 and graduated from the Machine-Building Faculty at Kalinin Polytechnic Institute in 1978. He received his Candidate’s degree on methods of adaptive transformation of data storage systems in 1991. Worked in Russian and foreign companies on projects of information systems and multipurpose technologies. Since 1999, he has been an Associate Professor at the Department of Information Systems of the Tver State Technical University. Scientific interests are databases and digital libraries, artificial neural networks in fuzzy expert systems, text data mining, and e-learning. He is the author of more than 150 papers.
Boris Vasil’evich Palyukh was born in 1948 and graduated from the Chemical Technology Faculty at Kalinin Polytechnic Institute in 1971. He received his Candidate’s degree in the area of automation of technological processes and production in 1978 and his Doctoral degree on the bases of construction and development of an automated control system for the operational reliability of chemical production at the Moscow Institute of Chemical Technology in 1991. Honored Worker of the Higher School of the Russian Federation, Professor, expert of the Russian Science Foundation, and member of the Scientific Council of the Russian Association of Artificial Intelligence. He trained and worked as a visiting professor abroad. From 1993 to the present, he has been the head of the Department of Information Systems of the Tver State Technical University; he was the rector of this university in 2007–2013. Scientific interests are control in technical and organizational intelligent systems, reliability and technical diagnostics, expert systems, interval analysis, and decision theory. He is the author of more than 200 scientific publications, including monographs and textbooks.
Translated by I. Tselishcheva
Rights and permissions
About this article
Cite this article
Ivanov, V.K., Palyukh, B.V. Application of Evidence Theory for Training Fuzzy Neural Networks in Diagnostic Systems. Pattern Recognit. Image Anal. 33, 354–359 (2023). https://doi.org/10.1134/S1054661823030197
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1134/S1054661823030197