Development of the information technology for decision making support when managing refrigeration units

Abstract

Досліджено особливості процесу прийняття рішень стосовно управлінського впливу з боку оператора холодильної установки, як багатофакторної енергетичної системи із внутрішніми та зовнішніми збурювачами. Запропонована інформаційна технологія може бути використаною для створення комп'ютерних тренажерів для підвищення компетентності та кваліфікації промислово-виробничого персоналу без стажування на об'єктах підвищеної небезпеки. We have studied patterns in the decision-making process related to the managerial influence on the part of the operator of a refrigerating unit as a multifactor energy system with internal and external disturbances. An object such as a refrigeration unit cannot be fully formalized and described using the methods of conventional modeling as it has the properties of partial self-regulation and self-adjustment. Therefore, based on the application of the generalized model of a refrigeration unit, we have improved a decision support system that makes it possible to take into consideration the non-formalized information by means of a neuro-fuzzy component. The information technology has been developed to support decision making when managing different types of refrigeration units. Its implementation would make it possible to reduce the time required for equipment to enter the necessary operation mode and to stabilize a temperature regime at objects. That allows a decrease in the working time ratio of refrigeration equipment and reduces the influence of the human factor, which improves safety of the energy unit operation. Effectiveness of the technology has been experimentally investigated at a single-stage vapor-compression industrial ammonia refrigeration machine, at a single-stage vapor-compression freon refrigeration machine of central air conditioner and at a water- ammonia absorption-diffusion assembly of a household freezer the type of island freezer. The number of disturbing factors and the factors of influence varies in a wide range. The datasets intended to train a neuro-fuzzy system were built based on the results of experiments at actual equipment. The proposed information technology could be used to construct computer simulators to enhance the competence and qualification of industrial-production personnel without training at the facilities of increased danger.