CONICET | Buscador de Institutos y Recursos Humanos

This work proposes an alternative methodology for designing multi-loop control structures based on steady-state indexes and multi-objective combinatorial optimization problems. This unified approach avoids any dynamic simulation to decide about the final optimal control structure and systematizes the overall procedure minimizing the heuristic knowledge requirements. Furthermore, this methodology allows to incorporate any potential modifications of the problem context due to structural faults in the process such as sensor and/or actuator malfunctions. The case study selected here is the well-known Tennessee Eastman process.