In this paper, we analyze next generation sequencing (NGS) data of wastewater treatment plant (WWTP) in the North Water facility for revealing the role of 1236 different genera of microorganisms in the aeration basin to the measured process data. Both the time-series data of NGS and process parameters are pre-processed and analyzed using support vector regression technique and is compared with the deep neural network approach. Local sensitivity analysis is performed on the resulting models. Both machine learning analyses show the importance of a subset of genera to the WWTP process and can be used to enrich / to adapt the well-studied activated sludge model (ASM).
|Conference||1st IFAC Workshop on Control Methods for Water Resource Systems (CMWRS19)|
|Period||19/09/2019 → 20/09/2019|