TECHNICAL MECHANICS
|
|||||||
|
|||||||
Home
>
Journal Issues
>
No 1 (2022) Technical mechanics
>
8
___________________________________________________ UDC 519.816 Technical mechanics, 2022, 1, 77 - 90 Construction of a multiplicative-additive criteria convolution in the space of quantitative and qualitative indices to determine the priority of projects DOI: https://doi.org/10.15407/itm2022.01.077 Mamchuk V. M.
Mamchuk V. M.
One of the main problems in scientific activity organization on a competitive basis is to improve methods of
R&D project evaluation and priority determination. The project priority level may be determined using
approaches based on the multi-attribute utility theory (MAUT), whose development is the subject matter of
many studies and publications. Despite of the large number of publications on the subject, the development
of a scientifically substantiated mathematical apparatus for multicriteria project evaluation is still a
topical and challenging problem. The complexity of the development of project priority determination methods
is due to difficulties in the construction of a unified rating scale that would allow one to measure the
value of project indices differing in physical content and dimension. That is, what is difficult is the
structurization of a decision-making person (DMP)’s preferences and the formalization of preference
evaluation. It is also difficult to construct a criterion-target model that would adequately represent the
system of DMP preferences in the form of a scalar value function, which is termed a criteria convolution,
an integral criterion, or an integral value function (IVF). MAUT-based computational algorithms widely use
procedures of common criteria scaling, in which one quality index is replaced with another. Such algorithms
have a resolution equal to one; however, they operate with quantitative criteria alone, thus significantly
narrowing their application area. Another drawback of theirs is the lack of simple methods to determine the
value function at indifference (DMP preference equality) points. The aim of this work is to eliminate these
drawbacks in a multiplicative-additive IVF model. To do this, the following was done. Functional
relationships between DMP preferences and alternative quality indices were established to give analytical
expressions for evaluating local value functions at indifference points. A method was developed for
constructing a multiplicative-additive criteria convolution to evaluate and rank alternatives in the space
of quantitative and qualitative indices. An algorithm was developed to determine the priority of projects;
the algorithm allows one to rank alternatives with a resolution equal to one. In this work, decision theory,
multicriteria utility theory, and verbal decision analysis methods were used. The results obtained may be
used in R&D efficiency evaluation, competitive project selection, and space program formation in the rocket
space industry.
quantitative and qualitative criteria, value function, criteria convolution, alternative ranking
1. Pylypenko O. V., Pereverzev E. S., Marchenko V. T., Khorolskyi P. P., Pechenevskaya O. K. Effectiveness of R&D Projects and Programs. Dnipropetrovsk: Porogi, 2008. 509 pp. (in Russian).
Copyright (©) 2022 Mamchuk V. M. Copyright © 2014-2022 Technical mechanics ____________________________________________________________________________________________________________________________ |
GUIDE FOR AUTHORS ==================== Open Access Policy ==================== REGULATIONS on the ethics of publications ==================== |