Dynamic Optimization of Fed-Batch Bioreactors Using the Ant Algorithm

Abstract

The ant colony algorithm, mimicking the cooperative search behavior of ants in real life, has been employed for the dynamic optimization of fed-batch bioreactors. To test the capability of this new heuristic algorithm, two well-known and extensively studied systems have been chosen. The algorithm rapidly converges to optimal feed rate profiles, which maximize the overall production of the desired product and the profits in a computationally efficient and robust manner. The optimal profiles evolved are easy to implement in plant operation. The algorithm compares favorably with the other known techniques.