Search academic texts

Information × Registration Number 0215U001334, 0113U000449 , R & D reports Title Development of intelligent decision support systems for the diagnosis, management and optimization of technical and biotechnical objects popup.stage_title Head Dmitrienko Valery, Registration Date 12-02-2015 Organization The Kharkov state polytechnical university popup.description2 The object of study - the functioning of the intellectual processes of decision support systems, trained in the operation, based on stable-plastic neural networks (NN) adaptive resonance theory (ART). Objective: Development of theoretical bases of neural networks adaptive resonance theory with a new architecture that allows for the input of information to determine the appropriate association for the development of intelligent decision support systems (IDSS) that address the problem of recognition, modeling, control and optimization of dynamic processes in a significant a priori uncertainty in creation intelligent decision support systems (IDSS) for the diagnosis of elevators and elevator group supervisory control when creating IDSS to optimize production machining processes in engineering, the development of medical diagnostic and therapeutic systems ISPPR for systems analysis and protection against electromagnetic pulses and natural artificial origin of various objects. Research methods - the theory of neural networks adaptive resonance theory of decision support systems, the theory of electromagnetic processes, the theory of mechanical materials processing blade, the theory of asynchronous traction drive, the theory of operation of of lifting equipment. The results obtained correspond to the world level, since the National Assembly with similar properties do not exist, and the new National Assembly ART open the possibility of developing ISPPR with properties which are unobtainable for NN ART with traditional architecture. First developed NN ART, which allow for the input of information to determine the number of associations, as well as to determine the chain of associations. Also developed a neural network bidirectional associative memory which can input image corresponds to one or more associations. The architecture of universal ISPPR oriented for control system where decision-making procedure should be performed with a minimum of time. Software has been developed to support decision-making of the central dispatching service elevator control of the enterprise "Harkovgorlift." Software is embedded in the enterprise "Harkovgorlift." Completed development of the intellectual structure of decision support systems and medical database structure intellectual decision support system. Developed the core of the system. Further developed discrete neural network based on Hebb's rule and adaptive resonance theory in which, in contrast to the known, introduced K-valued neurons that will automate the process of recognition signals and interference, leading to a risk of disruption in the simulation of digital devices based on K- digit differential calculus. Product Description popup.authors Бречко Вероніка Олександрівна Заковоротний Олександр Юрійович Заполовський Микола Йосипович Ліпчанський Максим Валентинович Леонов Сергій Юрійович Мезенцев Микола Вікторович Носков Валентин Іванович Поворознюк Анатолій Іванович Серков Олександр Анатолієвич Хавіна Інна Петрівна popup.nrat_date 2020-04-02 Close