Abstract
This paper explores the application of nonlinear autoregressive neural networks with exogenous inputs (NARX) in DC electric drive control systems for hot rolling mills. We consider a NARX network architecture adapted for solving armature current (SEDCM_IA) and angular velocity (SEDCM_WA) problems under significant transient loads and process variables. Mathematical modeling of the electric drive operation is conducted using two separate NARX networks, demonstrating the effectiveness of the proposed approach compared to traditional mathematical modeling methods. It is shown that the use of NARX networks significantly improves control accuracy by taking into account the nonlinear dynamic characteristics of the system and compensating for disturbances in real time. The results of the study are of practical interest for the modernization of rolling mill automation systems and can be used in the design of adaptive control systems for critical electric drives.