Fed-batch processes are of great importance to biochemical industries. Although they typically produce low-volume, high-value products, however, the associated cost is very high. Optimal operation is thus extremely important, since every improvement in the process may result in a significant increase in production yield and saving in production cost. The major objective of the research that is described in this chapter is not to keep the system at a constant set point but to find an optimal control profile to maximize the product of interest at the end of the fed-batch culture. In this work, real-valued GAs are chosen to optimize the high order, dynamic and nonlinear system.
GAs are stochastic global search methods that imitates the principles of natural biological evolution [60,64,65,67]. It evaluates many points in parallel in the parameter space. Hence, it is more likely to converge towards a global solution. It does not assume that the search space is differentiable or continuous and can be also iterated many times on each data received. GAs are a promising and often superior alternative for solving modelling and optimal control problems when conventional search techniques are difficult to use because of severe nonlinearities and discontinuities [76,79]. Some researches on bioprocess optimization using GAs are found in the literature [76,80,81].
GAs operate on populations of strings, which are coded to represent some underlying parameter set. Three operators, selection, crossover and mutation,
L. Z. Chen et al.: Modelling and Optimization of Biotechnological Processes, Studies in Computational Intelligence (SCI) 15, 17-27 (2006)
www.springerlink.com © Springer-Verlag Berlin Heidelberg 2006
are applied to the strings to produce new successive strings, which represent a better solution to the problem. These operators are simple, involving nothing more complex than string copying, partial string exchange and random number generation. GA realize an innovative notion exchange among strings and thus connect to our own ideas of human search or discovery.
The remaining sections of this chapter proceed as follows: in Section 2.2, a seventh order model is introduced and the related practical problems are formulated; Section 2.3 explains the basics of GAs; in Section 2.4, the simulation results are given; conclusions are drawn in Section 2.5.
Was this article helpful?