ANNs Combined with Genetic Algorithm Optimization for Symbiotic Medium of Two Oil-Degrading Bacteria Cycloclasticus Sp. and Alcanivorax Sp.

In order to study the symbiotic medium of two petroleum degrading bacteria, Cycloclasticus sp. and Alcanivorax sp. Single factor experiment, uniform design and neural network combined with genetic algorithm were used to optimize medium components. The concentration of carbon source (diesel) is 0.5%, the concentration of nitrogen source (NH4)2SO4 is 8 g/L, phosphorus source is KH2PO4:Na2HPO4 mol ratio 3:1, yeast powder concentration is 0.03 g/L, culture time is 3D, culture temperature is 20, pH = 7.6, loading amount is 30 mL, inoculation amount is 1.5%, and rocking bed speed is 150 rpm. The test results were consistent with expectations. The degradation rate reached 89.80% after optimization and 14.4% higher than the original 78.50%. Through the optimization of the symbiotic medium, the culture medium suitable for the symbiosis of bacteria group is obtained, which shows that the neural network and genetic algorithm have remarkable superiority in the optimization of culture medium.