Evolutionary search for processing the results expert compsrisons several homogeneous reliability


  • V. V. Tkahova Department of heat technic and gas supply, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”, Ukraine https://orcid.org/0000-0001-9943-1852
  • V. F. Irodov Department of heat technic and gas supply, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”, Ukraine https://orcid.org/0000-0001-8772-9862


method of expert estimations, the matrix T. Saaty, indice of reliability, failure rates, efficiency, evolutionary search


Purpose. There are regulations requirements for reliability in the projecting and exploitation of new technical elements. The determining task in the parameters of reliability is appeared. It is a requirement in the latest specifications. There are reliability indices among the standardized requirements for the development of new technical elements and systems. There are contradictions for the latest systems. It is necessary to give an estimate of reliability, when there were no long-term experiences of their long exploitation. The traditional method for receiving of reliability indices of equipment or technical systems is the statistical analysis of the results of their long-term exploitation. Statistical method is not a solution of this problem, but we think that using the method of expert estimations can be a better way to solve this problem. In particular, this situation became a real problem for tube gas heaters (TGH). These heaters are widespread for heat supply to industrial plants. To other side, the term of exploitation TGH is counted for only a few years. To get real statistical estimation is not possible on short-term exploitation. In this case for getting estimation of reliability parameters we can use the method of expert estimations. In particular, Thomas Saaty method of pairwise comparison of the factors. Following, T. Saaty method we can create a matrix of pairwise comparisons using approximate scale of importance. Traditionally the T. Saaty matrix is processing by finding the vector of their own values. The matrix must be completely filled. The result of the vector of matrix value depends on possible fluctuations in the values of this matrix. For processing of this matrix we offer to use an evolutionary random search of more preferable (Rs-optimal) solutions. According to formulation the purpose of choosing solutions in the form of rules of preference, the evolutionary search expands capabilities of expert estimations, including incorrect possible of looking for solutions though using of additional criterion (stabilizer).

Methodology. General algorithm of evolutionary random search of more preferable (Rs-optimal) solutions is used. The rule of preference (choice of solutions) is create, which let to get the desired solution in evolutionary random search.

Findings. The estimation of failure rates efficiency for tube gas heaters according to the Thomas Saaty matrix and evolutionary random search of more preferable solutions has been completed. The application of evolutionary random search of more preferable solutions allows to improve the processing of expert estimates using T. Saaty matrix. Such a method provides the ability to process the results of of expert estimations incomplete T. Saaty matrix. And to consider the additional matrix, except this one information on the compared objects at construction function selection decisions.

Originality. The approach to the processing of the results of expert estimation of the matrix T. Saaty using evolutionary random search of more preferable solutions is proposed.

Practical value. Using evolutionary search for the processing of expert estimations, which is based on the Thomas Saaty method. It gives advantages in the flexibilities to formulation requirements for the procedure of processing of expert estimations.

Author Biographies

V. V. Tkahova, Department of heat technic and gas supply, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”

Cand. Sc. (Tech.), Ass. prof.

V. F. Irodov, Department of heat technic and gas supply, State Higher Education Establishment “Pridneprovsk State Academy of Civil Engineering and Architecture”

Dr. Sc. (Tech.), Prof.


Avvakumov V.G., Makurina M.V. Algoritm sinteza matritsy T. Saati dlya sravneniya faktorov nechislovoy prirody [T. Saati matrix synthesis algorithm for comparing factors nonnumerical nature]. Sibirskiy torgovo-ekonomicheskiy zhurnal [Siberian trade and economic journal], 2010, issue 11, pр. 11-16.

Dudkin K.V., Tkacheva V.V., Irodov V.F. Trubchatye gazovye nagrevateli dlya teplosnabzheniya v selskoy mestnosti [Tube gas heaters for heat supply in the countryside]. LAP LAMBERT Academic Publishing, 2014. 156 p.

Ionin A.A. Gazosnabzhenie [Gas supply]. Moscow, Stroyizdat Publ., 1989. 439 p.

Kryvosheiev V.M. Vykorystannia metodiv ekspertnoi otsinky v metodi analizu iierarkhii pry vybori ratsionalnykh variantiv skladnykh tekhnichnykh system [Use of expertise in the analytic hierarchy process in choosing rational options for complex technical systems]. Zbirnyk naukovykh prats VITI NTUU„KPI” [Bulletin of Military Institute of Telecommunications and Information State University of Telecommunications], 2010, issue 2, pp. 42-51.

Orlov A.A. Teoriya prinyatiya resheniy [Adoption Theoryof Solutions ]. Moscow, Ekzamen Publ., 2006. 573 p.

Pavlov A.N. Reshenie mnogokriterialnykh zadach metodom analiza ierarkhiy: uchebnoe posobie [Solution of multi-criteria problems by the Analytic Hierarchy Process]. Moscow, RAGS Publ., 2010. 116 p.

Plotkin V.I., Ogurtsova E.V., Vilkhivskaja O.V. Metodika soglasovaniya ekspertnykh dannykh [Methods of harmonizing the expertise]. Systemy obrobky informatsii [Information processing systems], 2012, issue 8, pp. 146–148.

Ratushniak G.S., Stepankovskyi R.V. Otsiniuvannia nadiinosti funktsionuvannia ventyliatsiinykh system za dopomohoiu neirochitkykh merezh [Evaluation of reliability of the ventilation system using Neural Networks]. Suchasni texnolohii, materialy i konstruktsii v budivnytstvi - Modern technologies, materials and constructions in building, 2013, pp. 12-20.

Saati T. Prinyatie resheniy. Metod analiza ierarkhiy [Making decisions. Analytic hierarchy]. Moscow, Radio i svyaz Publ., 1993. 320 p.

Horn R., Dzhonson Ch. Matrichnyy analiz [Matrix analysis]. Moscow, Mir Publ., 1989. 655 p.

Genest C., Rivest L. A. Statistical look at Saaty`s method of estimating pairwise preferences expressed on a ratio scale. J. M. Psychol, 1994, vol. 38, рр. 477-496.

Pradeep K.S., Ranjan K.P., Om P.S. A critical analysis on software fault prediction techniques. World applied sciences journal, 2015, vol. 33, no. 3, рр. 371-379.

Saati T. L. Decision making with dependence and feedback: the analitic network process. Pittsburgh, RWS Publ., 2001. 376 p.

Siperkovskis V. Using T. Saaty method in transport systems planning. Transport and telecommunication institute, 2009, vol. 10, no. 4, рр. 18-27.

Tomaszewski P, Håkansson J., Lundberg L., Grahn H. The accuracy of fault prediction in modified code –statistical model vs. expert estimation. Proceedings of the 13th Annual IEEE International Conference on the Engineering of Computer Based Systems (ECBS 2006), Potsdam, 2006. рр. 4-14.

Totsenko, V. G. Methods and decision support systems / V. G. Totsenko. – Kiev : Naukova Dumka, 2002. –381 p.Totsenko V.G. Methods and decision support systems.Kiev, Naukova Dumka Publ., 2002. 381 p.



Computer systems and information technologies in education, science and management