Adaptive control for the 13C isotope separation process

Abstract

In this paper a solution required in order to control the <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sup> C isotope concentration inside and at the output of the separation column is presented. The classical control strategies, namely the PI and the PID controllers tunned using the Ziegler-Nichols method, were approached. The system performances were improved using an adaptive mechanism to compute their optimal parameters and a comparison between the results obtained between those methods is presented. The chemical exchange process between the carbamate and the carbon dioxide taking place, it results the enriched <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sup> C isotope. This process is a distributed parameters one, depending on the time constant and on the position in the column's height, and it can be simulated after determining an analytical solution when the process is working in the transitory regime. The control strategies require the implementation of the process modelling – it is desired to reach the <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sup> C isotope concentration by using an optimal amount of monoethanolamine solution, the input flow. The adaptive mechanism consists of a PI controller which determines the proper time constants of the two controllers so the control of the <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">13</sup> C isotope concentration is possible in any point of height in the separation column, improving the performances of the system by reducing the settling time – if the concentration suffers variations at certain moments of time, the adaptive mechanism is also capable of settling with the new requirements.