Key words:ABSTRACT Robust-optimal regulator Control system H∞-norm Weight parameters Technological object Energy savingThe energy efficiency of robust-optimal control systems for continuous-type technological objects in a dynamic mode, operating in conditions of significant uncertainties is investigated in this paper. The energy efficiency of the control system depends directly on the compromise between quality and power consumption, which is currently provided to the designer of the automation system. The logical improvement of such a system is the construction of an optimization problem for the automatic choice of compromise.The purpose of the paper is to develop a methodology for choosing an alternative between energy consumption and the quality of the control system, which regulation device is synthesized by the criterion of the H∞-norm of a closed system in the space of state variables.The general statement for the problem of robust optimum synthesis of the control device in the space of the state variables is presented, and to the object's mathematical model of which weighted parameters are introduced. The latter are given as real numbers, the fixed fractional-rational transfer functions's coefficients or a set of fractional-rational transmitting functions. Three criteria of energy efficiency have been formed on the basis of the introduced weight parameters that determine the compromise between quality and energy saving of the system. Two of them are reduced to a one-criterion optimization problem with constraints and one to a two-criterion problem based on a stable and effective compromise. Approaches to solving such problems are presented.For example, the robust-optimal regulator according to the H∞-criterion for a multidimensional technological object was synthesized. The methods of influence on the energy efficiency of the control system are shown and the weight parameters, which were found on the basis of a compromise between energy consumption and system quality were determined. It is proved that such a system is energy-saving, in addition, when minimizing the established criteria, the alternative to the choice facing the designer of the control system is removed.

No abstract

The article deals with materials concerning estimation of efficiency of intelligent control systems for complex process facilities. Special attention is drawn to the general technicaleconomic indicators, used to assess the degree of reduction in total amount of work expended on the production of a unit of finished product, as well as indicators of energy and resource efficiency, and cost function minimizing. The traditional task of choosing a particular automation system is to compare the expectation effect of its implementation with the cost of equipment and maintenance of the system. In economic effect analysis of modern automation system for complex process facilities, which include АСS, ICS, algorithms of robustness, adaptability, coordination, diagnostics, forecasting, etc. there are distinguished such components as structural effect, technological effect, energy effect, labor effect. There is a number of different approaches, which are offered to form a complex ICS efficiency indicator. It is shown that ICS efficiency is directly related to the time estimations of receipt and realization of making decisions, but main effects of their application are shown in possibility of assured achievement of management aim with the highest quality at the top level and the highest energy and resource efficiency indicators at the lower executive level "object of control + regulator" of hierarchical system. An important requirement in intelligent control systems is ensuring of robustness of traditional regulators P, PI, PD, PID, which are used at the bottom executive level. Different level of intellectuality depending on the completeness and correctness of the knowledge base, can be obtained using hybrid intelligent control systems by means of integration of PID-and unclear regulators. It is analysed, that in ICS more and more tasks for complex process facilities are solved on the basis of the use of neural, adaptive neural and neuron-sensitive methods, sensitivity functions.

The article proposes the implementation of an applied ontology of mathematical models of technological objects for the design of a subsystem of decision-making support, which issues recommendations for the mathematical apparatus in relation to the development goals for the automated management of a food enterprise. It consists of 46 entities with corresponding relationships, attributes, and axioms, and is also implemented in the OWL language by the Protege open platform tools, taking into account existing standards and recommendations. In the structure of the ontology, the mathematical model is represented as branches of subclasses with the corresponding sets of attributes, characterized by their relationship to the higher-level model. Within this ontology, 17 types of relationships are presented. Applied ontology passed two stages of verification: structural - based on generally accepted estimates; logical - by testing queries and manually checking the correctness of the results. In particular, examples of model selection using an ontology for virtual sensors are given. The use of the proposed ontology in the structure of the management decision support subsystem increases the efficiency of these decisions, the validity of management actions and the efficiency of the technological component of the enterprise. Also, the ontology can be integrated into the ontology of industrial enterprises tasks or other domain ontologies.

No abstract