Characterizing bubbles in bioreactors using light or ultrasound probes: data analysis by classical means and by neural networks

The applicability of neural networks to a typical biochem. engineering problem was investigated. The data treatment required for a novel technique for measuring the mean bubble size and the sp. interfacial area in aerobic bioreactors was chosen as an appropriate example in case. The new measuring technique is based on either light or ultrasound transmission measurements through the dispersion. The function relating the results of such measurements to bubble size cannot be explicitly modelled. It was therefore approximated by a purely math. function fitted to numerically simulated examples. By training neural networks, the same unknown relation could be mapped with similar or better precision without any prior knowledge of its structure or form and without any phys. understanding of the problem. [on SciFinder (R)]