Problems of intellectualizing in SHM systems: estimation, prediction, multi-class recognition https://doi.org/10.33108/visnyk_tntu2017.04.135
Article Sidebar
Main Article Content
Abstract
The paper is devoted to the research of the efficiency of solving the intellectualization problems of multi-channel systems for the structural health monitoring of complex spatial objects with welded joints - tanks with ecologically hazardous substances. Based on the monitoring models of the object a visualization subsystem is developed for the reflection and prediction of the stress-strain state characteristics, spatial position and vibration state. The use of a classifier based on a probabilistic neural network has been developed for the multi-class recognition of structural health of the tank with the multi-site damage. Learning and test sets of the incoming multidimensional vectors of diagnostic features have been formed, classifier training and multi-class recognition in the case of structural degradation have been performed. The dependencies of the efficiency of the classifier on the parameter of the network influence for different values orders of diagnostic features have been found.
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
References
1. Adams D. Health Monitoring of Structural Materials and Components. Methods with Applications, John Wiley & Sons Ltd., 2007, 475 p. https://doi.org/10.1002/9780470511589
2. Nagarajaiah, S. Structural monitoring and identification of civil infrastructure in the United States / S. Nagarajaiah, K. Erazo // Structural Monitoring and Maintenance. 2016. Vol. 3. N. 1. P. 51 - 69. https://doi.org/10.12989/smm.2016.3.1.051
3. Structural Health Monitoring 2003: From Diagnostics & Prognostics to Structural Health Menegment: Proceedings of the 4th International Workshop on Structural Health Monitoring, Stanford University, Stanford, CA, September 15-17, 2003, 1552 p.
4. Speckmann H. Structural Health Monitoring: a contribution to the intelligent aircraft structure. [Електронний ресурс] / H. Speckmann, H. Roesner // Proc. 9th European NDT Confer. (ECNDT), 25–29 Sept., 2006, Berlin, Germany. − Режим доступу: http://www.ndt.net/article/ecndt2006/doc/Tu.1.1.1.pdf.
5. Bouraou N. Structural Functional Synthesis of the Diagnosis Systems of the Construction at the Operation /N. Bouraou, О. Pavlovskyi, D. Shevchuk // Scientific Journal of TNTU. 2013. № 4 (72). P. 77 – 86 [in Ukrainе].
6. Tsybulnik S. The concept of visualization of data in information and diagnostic complexes / S. Tsybulnik, N. Bouraou // Bulletin of Ingineering Academy of Ukraine. Kyiv, 2015. №1. P. 96 – 99 [in Russian].
7. Tsybulnik S. Improvement of means of functional diagnostics and protection of tanks on the basis of simulation modeling: author's abstract. ... candidate techn. sciences: 05.11.13 / S. Tsybulnik. Кyiv. : NТUU «КPІ», 2016. 27 p. [in Ukrainе].
8. Synthesis of neural network for multiclass diagnostics of elements of construction in operation / N. Bouraou, А. Protasov, P. Мyronenko, S. Rupich // Methods and instruments of quality control. Ivano-Frankivsk: ІFNТUNG. 2015. № 2 (35). P. 83 − 93 [in Ukrainе].
9. Duhanov A. Simulation of complex systems: a course of lectures / А. Duhanov, О. Меdvedeva. Vladimir: Publishing of Vladimir. State Univ., 2010. 115 p. [in Russian].
10. The system of registration and visualization of motion parameters [Electronic resource]. Access mode: http://iu2.bmstu.ru/netrad/article08.pdf [in Russian].
11.Bouraou N. Multi-Class Recognition Objects Technical Condition by Classifier based on Probabilistic Neural Network / N. Bouraou, D. Pivtorak, S. Rupich // Eastern-European Journal of Enterprise Technologies. 2017. № 5/4 (89). P. 24 – 31.