Genetic Algorithms for the Synthesis and Integrated Design of Processes Using Advanced Control Strategies

This work presents a real-coded genetic algorithm to perform the synthesis and integrated design of an activated sludge process using and advanced Multivariable Model-based Predictive Controller (MPC). The process synthesis and design are carried out simultaneously with the MPC tuning to obtain the most economical plant which satisfies the controllability indices that measure the control performance (H infinity and 11 norms of different sensitivity functions of the system). The mathematical formulation results into a mixed-integer optimization problem with non-linear constraints. The quality of the solutions obtained evidence that real-coded genetic algorithms are a valid and practical alternative to deterministic optimization methods for such complex problems.