Neural network based optimal control of a biosynthesis process

Abstract

Knowing information about the physiologic state of microorganisms, an optimal control for biosynthesis processes can be implemented. Information regarding the physiologic state of microorganisms is given by their mean age, which can be estimated based on the measurable state variables of the process. In this paper, a neural network based estimator for the microorganisms' mean age is developed and used for implementing an optimal control law for the biosynthesis process. The data taken from such biosynthesis processes are affected by considerable measurement noise. The estimator in this paper was developed using noisy real data, the whole control structure for the process shows good control performances.