Optimal Feed Rate Strategy of Biotechnological Process in L-lysine Production Using Neuro-Dynamic Control

In this paper Neuro-dynamic programming (NDP) is pro- posed as an alternative to alleviate the "curse of dimensionality" of the Dynamic programming (DP) for optimal control of a fed-batch fer- mentation process in the L-lysine production. The traditional approach for solving the Bellman equation involves gridding of the state space, solving the optimization for each grid point, as well as performing the stagewise optimization until convergence is reached. The comprehen- sive sampling of the state space can be avoided by identifying the relevant regions of the state space through simulation under judi- ciously chosen suboptimal policies, which is presented using NDP methods. The most effective and cheapest method for the L-lysine biosynthesis (in biological active form) is the microbiological method via direct fermentation. In this paper an optimization method of the L-lysine production from strain Brevibacterium flavum 22LD is used and that is NDP. The results show that the quality of L-lysine enhances at the end of the process. The proposed method is particularly simple to implement and can be applied for on-line optimization.

[1]  Andrew G. Barto,et al.  Learning to Act Using Real-Time Dynamic Programming , 1995, Artif. Intell..

[2]  Abhishek Suraj Soni.  A MULTI-SCALE APPROACH TO FED-BATCH BIOREACTOR CONTROL , 2002 .

[3]  Tatiana Ilkova,et al.  Dynamic and Neuro-Dynamic Optimization of a Fed-Batch Fermentation Process , 2008, AIMSA.

[4]  Niket S. Kaisare,et al.  Hierarchical multiscale model-based design of experiments, catalysts, and reactors for fuel processing , 2006, Comput. Chem. Eng..

[5]  Thidarat Tosukhowong,et al.  Approximate dynamic programming based optimal control applied to an integrated plant with a reactor and a distillation column with recycle , 2009 .

[6]  Jong Min Lee,et al.  An approximate dynamic programming based approach to dual adaptive control , 2009 .

[7]  Saso Dzeroski,et al.  Integrating Guidance into Relational Reinforcement Learning , 2004, Machine Learning.

[8]  Jay H. Lee,et al.  Choice of approximator and design of penalty function for an approximate dynamic programming based control approach , 2006 .

[9]  Niket S. Kaisare,et al.  Simulation based strategy for nonlinear optimal control: application to a microbial cell reactor , 2003 .

[10]  Saso Dzeroski,et al.  Integrating Experimentation and Guidance in Relational Reinforcement Learning , 2002, ICML.

[11]  S. Anastassiadis,et al.  L-lysine fermentation. , 2007, Recent patents on biotechnology.

[12]  Richard S. Sutton,et al.  Learning to predict by the methods of temporal differences , 1988, Machine Learning.

[13]  John N. Tsitsiklis,et al.  Neuro-Dynamic Programming , 1996, Encyclopedia of Machine Learning.