The article deals with a continuous wide-field hot rolling mill for a foundry and rolling complex (LPC). The goal of automation of a continuous broadband hot rolling mill is to ensure the production of high - quality hot rolled products regardless of the working staff using a self-learning neural network and huge databases in the rolling industry. The collector was designed with an innovative shape and a new cooling liquid spray scheme in the COMPASS-3D CAD. The collector is equipped with infrared sensors that read the surface temperature of the rolling roll and actuators. An open microcontroller platform Arduino UNO was chosen to implement an automatic system for purposeful cooling of rolling mill rolls. The proposed scheme of an automatic system for purposeful cooling of rolls is a classic system with negative feedback. The whole system can be divided into two circuits: the first is to regulate the amount of water supplied based on the readings of temperature sensors, one for regulating the pressure of water supplied based on the pressure sensor. This approach is proposed for the first time to automate cooling of rolling mill rolls. Analysis of the results of modeling the operation of the automated collector showed the following results: 16 transshipments per month, 2 scheduled preventive repairs, and the average roll working rate of 87%. The proposed system allows efficient use of the coolant compared to the old system. Based on the analysis of the operating time of rolling rolls under various cooling modes, the system will allow you to track roll wear and inform the staff about the upcoming transshipment dates. An artificial neural network was developed and implemented in the cooling system of rolling rolls at the 1950 LPC mill in order to increase the stability of the working rolls and improve the quality of the resulting products.
Read full abstract