Artificial neural networks prediction and optimization based on four light regions for light utilization from Synechocystis sp. PCC 6803

Light is crucial in microalgae growth. However, dividing the microalgae growth region into light and dark regions has limitations. In this study, the light response of Synechocystis sp. PCC 6803 was investigated to define four light regions (FLRs): light compensation region, light limitation region, light saturation region, and photoinhibition region. The proportions of cells' residence time in the FLRs and the number of times cells (NTC) passed through the FLRs in photobioreactors were calculated by using MATLAB. Based on the FLRs and NTC passed through the FLRs, a growth model was established by using artificial neural network (ANN). The ANN model had a validation R2 value of 0.97, which was 76.36% higher than the model based on light -dark regions. The high accuracy of the ANN model was further verified through dynamic adjustment of light intensity experiments. This study confirmed the importance of the FLRs for studying microalgae growth dynamics.