Baker's yeast fermentation parameters estimation : an evolutionary approach

An initial understanding of the application of an evolutionary approach to baker’s yeast fermentation parameters estimation is described. This problem has a nonlinear and multimodal nature. Thus, a global optimization technique, such the evolution strategies, must be used as an approximation to the global optimum. The process is modelled by a set of differential-algebraic equations to obtain the six state variables concentration profiles, in a well-mixed reactor. The state variables profiles are obtained using parameters usually referred in literature: seven kinetic parameters and nine yield coefficients. A new problem of optimization can be formulated with the objective of minimizing the overall sum of the squared residuals that adjust the simulation results to the experimental data. With the evolutionary approach not only the parameters can be estimated but also the most significant model parameters can be identified