Application of model-predictive control based on artificial neural networks to optimize the fed-batch process for riboflavin production.

The fed-batch process for commercial production of riboflavin (vitamin B2) was optimized on-line using model-predictive control based on artificial neural networks (ANNs). The information required for process models was extracted from both historical data and heuristic rules. After each cultivation the process model was readapted off-line to include the most recent process data. The control signal (feed rate), however, was optimized on-line at each sampling interval. An optimizer simulated variations in the control signal and assessed the forecasted model outputs according to an objective function. The optimum feed profile for increasing the product yield (YB2/S) and the amount of riboflavin at the time of harvesting was adjusted continuously and applied to the process. In contrast to the control by set-point profiles, the novel ANN-control is able to react on-line to variations in the process and also to incorporate the new process information continuously. As a result, both the total amount of riboflavin produced and the product yield increased systematically by more than 10% and the reproducibility of seven subsequently optimized batches was enhanced.

Growth kinetics of suspended microbial cells: from single-substrate-controlled growth to mixed-substrate kinetics.

Growth kinetics, i.e., the relationship between specific growth rate and the concentration of a substrate, is one of the basic tools in microbiology. However, despite more than half a century of research, many fundamental questions about the validity and application of growth kinetics as observed in the laboratory to environmental growth conditions are still unanswered. For pure cultures growing with single substrates, enormous inconsistencies exist in the growth kinetic data reported. The low quality of experimental data has so far hampered the comparison and validation of the different growth models proposed, and only recently have data collected from nutrient-controlled chemostat cultures allowed us to compare different kinetic models on a statistical basis. The problems are mainly due to (i) the analytical difficulty in measuring substrates at growth-controlling concentrations and (ii) the fact that during a kinetic experiment, particularly in batch systems, microorganisms alter their kinetic properties because of adaptation to the changing environment. For example, for Escherichia coli growing with glucose, a physiological long-term adaptation results in a change in KS for glucose from some 5 mg liter-1 to ca. 30 microg liter-1. The data suggest that a dilemma exists, namely, that either "intrinsic" KS (under substrate-controlled conditions in chemostat culture) or micromax (under substrate-excess conditions in batch culture) can be measured but both cannot be determined at the same time. The above-described conventional growth kinetics derived from single-substrate-controlled laboratory experiments have invariably been used for describing both growth and substrate utilization in ecosystems. However, in nature, microbial cells are exposed to a wide spectrum of potential substrates, many of which they utilize simultaneously (in particular carbon sources). The kinetic data available to date for growth of pure cultures in carbon-controlled continuous culture with defined mixtures of two or more carbon sources (including pollutants) clearly demonstrate that simultaneous utilization results in lowered residual steady-state concentrations of all substrates. This should result in a competitive advantage of a cell capable of mixed-substrate growth because it can grow much faster at low substrate concentrations than one would expect from single-substrate kinetics. Additionally, the relevance of the kinetic principles obtained from defined culture systems with single, mixed, or multicomponent substrates to the kinetics of pollutant degradation as it occurs in the presence of alternative carbon sources in complex environmental systems is discussed. The presented overview indicates that many of the environmentally relevant apects in growth kinetics are still waiting to be discovered, established, and exploited.