Begell House Inc.
Journal of Automation and Information Sciences
JAI(S)
1064-2315
48
9
2016
Valuation of Startups Investment Attractiveness Based on Neuro-Fuzzy Technologies
1-22
10.1615/JAutomatInfScien.v48.i9.10
Elena M.
Kiseleva
Oles Honchar Dnipro National University,
Dnepr
Olga M.
Prytomanova
Oles Honchar Dnipro National University,
Dnepr
Sergey V.
Zhuravel
Oles Honchar Dnepropetrovsk National University
startups investment attractiveness
neuro-fuzzy technologies
undifferential optimization
Shor algorithm
A mathematical model for evaluating startups investment attractiveness based on neuro-fuzzy technologies is proposed. It allows one to take into account not only a statistical uncertainty but also a linguistic one. The parameters optimization of created fuzzy model on the stage of its adjustment is performed using one of the undifferential optimization methods − Shor r- algorithm. The software that implements the proposed approach is developed. The results of modeling to define real startups investment attractiveness are presented
Models of Queueing-Inventory Systems with Randomized Lead Policy
23-35
10.1615/JAutomatInfScien.v48.i9.20
Agasi Zarbali ogly
Melikov
Institute of Control Systems of National
Academy of Sciences of Azerbaijan, Baku
Leonid A.
Ponomarenko
International Research and Training Center of Information Technologies and Systems of National Academy of Sciences of Ukraine and Ministry of Education and Science of Ukraine, Kiev, Ukraine
Sevindzh Alakber kysy
Bagirova
Baku State University, Baku
queueing-inventory system
randomized lead policy
impatient consume customers
finite and infinite queue
The models of the queueing-inventory systems with randomized lead policy and either finite or infinite queue of impatient consume customers are proposed. The exact and approximate methods to calculate the characteristics of the systems under the given lead policies are developed.
Coevolving Feedforward Neural Networks
36-48
10.1615/JAutomatInfScien.v48.i9.30
Oleg G.
Rudenko
Kharkov National University of Radio and
Electronics, Kharkov
Alexander A.
Bezsonov
Kharkov National University of Radio and Electronics, Kharkov
feedforward neural networks
coevolving
evolutionary algorithm
cooperation and competition models
populations totality
networks synthesis
An evolutionary algorithm of determining the architecture of feedforward neural networks and their training is proposed, based on the coevolutionary models of cooperation and competition with using of clustering algorithms for partitioning the main problem of neural network synthesis into subtasks which are to be solved in certain sub-populations. The proposed algorithm implements an environment that is conducive to cooperation and competition of populations in which every individual is a feedforward neural network, and the totality of the populations is responsible for the final solution of the set problem. The simulation results confirm the effectiveness of the proposed method of feedforward neural networks synthesis.
Mathematical Modeling of Direct and Inverse Problems of Dynamics of Thick Elastic Layer. Part II. Control Problems of Field of Transverse Dynamic Displacements of Layer
49-63
10.1615/JAutomatInfScien.v48.i9.40
Vladimir Antonovich
Stoyan
Kyiv National Taras Shevchenko University, Kyiv, Ukraine
Konstantin V.
Dvirnychuk
Kiev National Taras Shevchenko University; Chernovtsy Yuriy Fedkovich National
University, Chernovtsy
direct and inverse problems
transverse dynamic displacement
root mean-square consistency
outward dynamic loads
Control problems of three-dimensional field of transverse dynamic displacements of thick elastic layer of finite thickness by its root mean-square consistency with prescribed continuously or discretely defined required state are formulated and solved. Surface distributed outward dynamic loads and the initial disturbing factors taken individually or together are considered to be controlling factors. The conditions of accuracy and uniqueness of the obtained solutions are studied.
Synthesis Method of Empirical Models Optimal by Complexity under Uncertainty Conditions
64-74
10.1615/JAutomatInfScien.v48.i9.50
Mikhail I.
Gorbiychuk
Ivano-Frankovsk National Technical
University of Oil and Gas, Ivano-Frankovsk
Taras V.
Humenyuk
Ivano-Frankovsk National Technical University of Oil and Gas, Ivano-Frankovsk
synthesis method
empirical models
criteria of regularity
combinatorial method
There was developed the synthesis method of optimal complexity models for conditions when the model variables are fuzzy values. The method is oriented to the class of polynomial models. The best models are selected by using criteria of regularity or displacement. The application of ideas of genetic algorithms gives the opportunity to eliminate the problem of large dimension which is characteristic of combinatorial method. The efficiency of the developed method was verified on industrial data that allowed one to synthesize the empirical model optimal by structure for drilling conditions.
Problems Features of the Robust Control of Process Plants. Part I. Process Plants and their Mathematical Models
75-83
10.1615/JAutomatInfScien.v48.i9.60
Anatoliy P.
Ladanyuk
National University of Food Technologies, Kiev
Natalya N.
Lutskaya
National University of Food Technologies, Kiev, Ukraine
process plants
mathematical models
robust control systems
effectiveness
heat transfer
sugar mill
The process plants are considered in the context of designing robust control systems. By the example of the multibody evaporator of a sugar mill it is shown that the heat transfer coefficient affects not only the heat mode of the process, but also the effectiveness of the whole control system. Here the peculiarities of heat transfer are taken into account, which leads to the need of maintaining the liquid levels in the evaporator at optimal values, which is realized by the robust stabilization system.