Use of optical sensors and artificial neural networks for precision measurement and fault detection in metallurgical environments

Abstract This paper addresses the design and implementation of a measurement system for the steel industry, enabling precise positioning of steel castings before cutting. The focus is on measurement tools and methods, particularly the design of a new measurement chain for distance measurement. It highlights the importance of understanding environmental conditions that affect sensor functionality. The paper discusses proposed non-contact measurement methods for industrial automation and explores various sensor types for system implementation, including their properties and parameters, and the integration of artificial intelligence algorithms in the system design. The key contribution of this work lies in the unique design and innovative implementation of the measurement chain tailored to a specific technology, meeting the required working conditions and system parameters. It also provides a comprehensive analysis of the highly aggressive industrial environment, which cannot be effectively addressed by standard solutions.