Advanced control of the 13C isotopic separation process using AI

In this paper is introduced an approach for modeling, implementing, and simulating the production of the
¹³
C isotope by using the simplified model of the chemical process. To address the challenges of implementing and simulating the production of the enriched isotope, this paper approaches the approximation of the fractional-order system considering that the system's differentiation order varies, leveraging artificial intelligence methods. Given the process's slow nature, strong nonlinearity, time constants of increased values, and extended time required for settling, this paper also presents a comparison between the performances of the plant obtained by implementing three control structures: a PID controller, an IMC structure with a PD controller and an IMC structure with and adaptive PD controller. This paper also emphasizes the performances obtained by the adaptive strategy, which is operating with by variation of the sampling time, having an ability to reject the impact of disturbances in the system. All coefficients for the adaptative process are learnt through artificial intelligence methods, specifically utilizing neural networks.