Search academic texts

Information × Registration Number 0213U000765, 0110U006214 , R & D reports Title Modern information technologies at identification and management of difficult technical systems in conditions of uncertainty popup.stage_title Head Lyevi Leonid, Registration Date 18-03-2013 Organization Lugansk national agrarian university popup.description2 A research object is the difficult technical system of class of the hydromeliorative systems. A research purpose is development of theoretical aspects of problem of authentication and management the difficult technical systems in the conditions of vagueness by unclear models and adaptive algorithms. Theoretical and practical results. The first section gives the basic results of the first stage of theme "Analysis of the state of problem of automation control processes of the complex technical systems in the uncertainty conditions": 1) a technology of modelling dependences a "multidimensional input - output" by experimental data in the complex technical systems is described. The integration of two methodologies, such as soft computing and regression analysis is used; 2) an optimum location of sprinklers on the field by genetic algorithms is realized; 3) an identification of infiltration process and soil moisture distribution with taking into account the structural changes in soil is presented. The section two provides the basic results of second stage of theme "Fuzzy approaches, neural networks and genetic algorithms in identification and control of the complex technical systems": 1) the problems of micro-irrigation systems are certain. The necessity of development of fuzzy systems of decision making and fuzzy comptrollers for the operative planning of micro-irrigation is justified; 2) a problem of heterogeneity of soils is considered. The new construction of irrigation machine which allows to take into account heterogeneity is offered; 3) a metamodel of plant growth is designed. It is based on neural network and using for the control systems. The realization of algorithm is carried out in language of C++; 4) the questions of analysis and selection of adaptive control of complex objects in the uncertainty conditions are considered. The section three contains basic results of the third (final) stage of theme "algorithm and program realization of automatic control processes of the difficult technical systems": 1) an analysis of different neural networks is done. The task of selection of the most effective topology for the modelling of evapotranspiration and control irrigation is solved; 2) an algorithm for time series analysis with interval uncertainty which based on singular spectrum analysis (SSA) is described. 3) the task of management of the angle slope of sprinkler in the conditions of presence of side wind has been solved. Product Description popup.authors Дубовіков Артем Юрійович Правдюк Дмитро Михайлович Тащілін Максим Васильович Тащіліна Анастасія Віталійовна Чичикалов Олександр Володимирович popup.nrat_date 2020-04-02 Close