Data-Driven Multivariable Control of a Microalgae Growth Process

Abstract

Data driven control assembles new techniques that can use data from the plant and find the parameters of the controllers that achieve the desired behavior. The paper presents a simplified virtual reference feedback tuning method applied on the photosynthetic growth process of microalgae in photo bioreactor. The complex nonlinear nature of the process requires multiple control channels to establish the proper operating conditions for high growth yields. Two control channels were selected, namely dilution- biomass and CO2 gas flow-pH, on which independent controllers were designed. The data used for training was obtained with a benchmark model that was validated on different reactor types and microalgae strains. The obtained controllers were validated in simulation and proved to attenuate the interferences between the control channels.