Modelling of crystallizer wear

Purpose: In this paper will be described an analysis of control problems and technical lifetime modeling of continuous casting device crystallizers. A full exploitation of continuous casting equipment (CCE) advantages can only be achieved through a control system that minimizes all undesirable effects on the technological process. Some of the undesirable effects influencing the CCE process effectiveness are the failures and service interruptions. The failures and service interruptions are caused by a number of factors, impacts and processes that effect and run directly on the equipment in its individual parts during its operation. Design/methodology/approach: This problem was solved by connection of dependability theory and artificial neural networks. Findings: A prediction of crystallizer’s desk’s wear model was created on the basis of artificial neural networks and analytics diagnostics. Research limitations/implications: The limitations are given by operational data quantity. These limitations are for learning process and model adaptability. Practical implications: These problems are solved with cooperation with regional metallurgical companies. Gained results will be applied into the operational conditions. Originality/value: Signification consists of dependability theory and artificial neural networks, when solving a prediction model of crystallizers wear.