ABSTRACT
The bioprocess model used is nonlinear with constraints. A genetic algorithm (GA) has been used for dynamic optimization of the process. The results serve as optimal time profiles of process variables and act as reference time profiles for the controller. GA has been applied to find the optimum control inputs under different operating modes. Results showing the time profiles of different bioprocess variables and the corresponding control variables obtained from the simulation experiment are presented. Structured and segregated models have not been used in the present case, as they are far more complicated with a large number of parameters and states, many of which are unknown and unmeasurable. The nonlinear bioprocess model, along with the imposed state and control constraints, makes the solution of the dynamic optimization problem a difficult one by existing classical techniques.
