STRUCTURE OF DECISION SUPPORT SYSTEM OF INFORMATION SYSTEM INTELLIGENT CLIMATE CONTROL RESIDENTIAL

Abstract

Context. The ever-growing tendency to rise in price of energy makes it necessary to reduce power consumption, that is, to save energy.In terms of accommodation, the introduction of microclimate necessary for the organization of comfortable conditions for the subjects andeconomical use of energy.Objective. The purpose of work is to solve the actual problem of energy-efficient indoor climate control based on the use of informationintellectual system which takes into account the wishes of the subjects are there, which in turn, helps to ensure effective management ofheating devices by reducing or increasing the ambient temperature.Method. The solution of the problem suggested by the use of expert system structure as a component of the intelligent control system ofindoor climate through the use of neuro-fuzzy inference subsystem. This subsystem allows you to automatically generate control informationfor indoor climate control, depending on the wishes of the subjects, summarizing information on the time and place of their stay in differentperiods of time. As a logical subsystem suggested a five-layer neuro-fuzzy feedforward error system, which implements the fuzzy inferenceSugeno zero order. Scheme of intelligent indoor climate control system is also proposed and the approach to the implementation of the processof identifying the subjects in the room.Results. The experimental results confirmed the efficiency of the proposed expert system structure in systems «Smart House». It was also set parameters affecting the quality and performance of the proposed system. As an energy source, natural gas has been elected, and the average temperature ranges premises.Conclusions. A feature of the proposed system is the versatility of the use of any air conditioning, as well as to automatically adjust theroom climate to meet the wishes of subjects. Also, the main feature of the proposed method is to determine the microclimate settings andmemory behavior of the subjects of the room combined with neural networks makes it possible to predict and detect relevant indoor climatevalues, and as a result, to save energy.