Вікторія Ігорівна Зимовець, Олександр Сергійович Приходченко, Микита Ігорович Мироненко


The study aims to increase the functional efficiency of machine learning of the functional diagnosis system of a multi-rope shaft hoist through cluster analysis of diagnostic features. To achieve the goal, it was necessary to solve the following tasks: formalize the formulation of the task of information synthesis, capable of learning a functional diagnosis system, which operates in the cluster-analysis mode of diagnostic signs; to propose a categorical model and, on its basis, to develop an algorithm for information-extreme cluster analysis of diagnostic signs in the process of information-extreme machine learning of a functional diagnostic system; carry out fuzzification of input fuzzy data by optimizing the geometric parameters of hyperspherical containers of recognition classes that characterize the possible technical conditions of the diagnostic object; to develop an algorithm and implement it on the example of information synthesis of the functional diagnostics system of a multi-rope mine hoisting machine. The object of the study is the processes of information synthesis of a functional diagnostic system capable of learning, integrated into the automated control system of a multi-rope mine hoisting machine. The subject of the study is categorical models, an information-extremal machine learning algorithm of a functional diagnostic system that operates in the cluster analysis model of diagnostic signs and constructs decision rules. The research methods are based on the ideas and methods of information-extreme intellectual data analysis technology, a theoretical-informational approach to assessing the functional effectiveness of machine learning and on the geometric approach of pattern recognition theory. As a result, the following results were obtained: a categorical model was proposed, and on its basis, an algorithm for information-extremal machine learning of the functional diagnostics system for a multi-rope mine hoist was developed and implemented, which allows you to automatically generate an input classified fuzzy training matrix, which significantly reduces time and material costs when creating incoming mathematical description. The obtained result was achieved by cluster analysis of structured vectors of diagnostic signs obtained from archival data for three recognition classes using the k-means procedure. As a criterion for optimizing machine learning parameters, we considered a modified Kullback measure in the form of a functional on the exact characteristics of diagnostic solutions and distance criteria for the proximity of recognition classes. Based on the optimal geometric parameters of the containers of recognition classes obtained during machine learning, decisive rules were constructed that allowed us to classify the vectors of diagnostic features of recognition classes with a rather high total probability of making the correct diagnostic decisions. Conclusions. The scientific novelty of the results obtained consists in the development of a new method for the information synthesis of the functional diagnostics system of a multi-rope mine hoisting machine, which operates in the cluster analysis model, which made it possible to automatically form an input classified fuzzy training matrix with its subsequent dephasification in the process of information-extreme machine learning system.


information-extreme algorithm; training matrix; diagnostics; multichannel shaft lifting machine; clusterization; optimization; forecasting


Popov, Yu. V, Kudryavtsev, S. V., Stepanov, S. V. Modernizatsiya tormoznoy sistemy shakhtnoy pod"emnoy ustanovki [Modernization of the brake system of mine lifting installation] Gornyj informacionno-analiticheskij bjulleten' (nauchno-tehnicheskij zhurnal). – Mining information and analytical bulletin (scientific and technical journal), 2015, vol. 9, pp. 195-197.

Gerike, B. L., Gerike, P. B., Shahmanov, V. N. Dinamicheskaja diagnostika mashinnyh agregatov gornogo oborudovanija [Dynamic diagnostics of machine units of mining equipment]. Gornyj informacionno-analiticheskij bjulleten' (nauchno-tehnicheskij zhurnal) – Mining information and analytical bulletin (scientific and technical journal), 2011, no. 5, pp. 80–89.

Sun, W., Yang, G. A., Chen, Q., Palazoglu, A., Feng, K. Fault diagnosis of rolling bearing based on wavelet transform and envelope spectrum correlation. Journal of Vibration and Control, 2013, vol. 19, no. 6, pp. 924-941. DOI: 10.1177/1077546311435348.

Lytvyn, O. M., Lobanova, L. S., Zaluzhna, H. V. Pro odyn pidkhid do testuvannya novykh metodiv rozv"yazannya nestatsionarnoyi zadachi teploprovidnosti [About one week before testing new methods for developing unsteady tasks and heating problems]. Shtuchnyy intelekt – Artificial intelligence, 2012, no. 1, pp. 219-228.

Dong, S., Xu, X., Liu, J., Gao, Z. Rotating Machine Fault Diagnosis Based on Locality Preserving Projection and Back Propagation Neural Network–Support Vector Machine Model. Measurement and Control, 2015, vol. 48, no. 7, pp. 211–216.

Wang, D., Tse, P. W., Tsui, K. L. An enhanced Kurtogram method for fault diagnosis of rolling element bearings. Mech. Syst. Signal Process, 2013, vol. 35, no. 1, pp. 176-199.

Feng, Z., Chen, X., Liang, M. Joint envelope and frequency order spectrum analysis based on iterative generalized demodulation for planetary gearbox fault diagnosis under nonstationary conditions. Mech. Syst. Signal Process., 2016, vol. 76, pp. 242–264.

Andrzej, O., Zygmunt, S. Modern Methods Control and Diagnostics of Hoisting-Machines with Application of Artificial Intelligence Methods. Archives of Mining Sciences, 2010, no. 1, pp. 217 – 231.

Sidorenko, V. N., Chernyi, A. P. Monitoring of powerful electric drives on the basis of Data Mining technology]. Elektrotehnichni ta komp’yuterni sistemi – Electrical and computer systems, 2011, no. 3, pp. 313 -316.

Jayaswal, P., Verma, S. N., Wadhwani, A. K. Development of EBP-Artificial neural network expert system for rolling element bearing fault diagnosis. Journal of Vibration and Control, 2011, vol. 17, no. 8, pp. 1131–1148.

Samanta, B, Al-Balushi, K. R. Artificial neural network based fault diagnostics of rolling element bearings using time-domain features. Mechanical systems and signal processing, 2003, vol. 17, no. 2, pp. 317-328.

Subbotin, S. The neuro-fuzzy network synthesis and simplification on precedents in problems of diagnosis and pattern recognition. Optical Memory and Neural Networks (Information Optics), 2013, vol. 22, no. 2, pp. 97-103. DOI: 10.3103/s1060992x13020082.

Dovbysh, A. S.,. Budnyk, M. M., Moskalenko, V. V. Information Extreme Algorithm for Optimizing Parameters of Hyperellipsoidal Containers of Recognition Classes. Journal of automation and information sciences, 2012, vol. 44, no. 10, pp. 35 - 44. DOI: 10.1615/JAutomatInfScien.v44.i10.30.

Dovbysh, A., Velikodnyi, D., Shelehov, I., Bibyk, M. Information-extreme machine learning of the control system over the pover unit of a therminal pover main line. Eastern-European Journal of Enterprise Technologies, 2017, vol. 5/4, no. 89, pp. 17-24. DOI: 10.15587/1729-4061.2017.112121.

Dovbysh, A. S., Zimovets, V. I., Kozlov, Z. S. Information extreme learning algorithm system diagnostics about mine lifting machine. Radioelektronika i informatika, 2017, no. 4, pp. 58-63.

Scherbakova, G. Yu. DN-distribution Parameters Evaluation with Noise Stability Clustering for the Automated Systems of Technical Diagnostics. Shtuchnyy intelekt – Artificial Intelligence, 2010, vol. 4, pp. 491-496.

Schonlau M. Visualizing non-hierarchical and hierarchical cluster analyses with clustergrams. Computational Statistics, 2004, vol. 19, no 1, pp 95-111.



  • There are currently no refbacks.