Online Biomass Estimation in a Raceway Photobioreactor

Microalgae reactors provide an efficient alternative for production of biogas/biofuel, nutrition/pharmaceutics products, etc. Among the many variables involved in the dynamic models, the biomass concentration is a key variable for monitoring and control purposes. Existing sensors are not very suitable for control purposes in large reactors, thus there is a need of alternative soft-sensors (state estimators) providing robust estimates of the concentration. In this work we explore the application of a class of very robust observers, namely sliding-mode observers, to biomass concentration. The novelty of this approach is that it only requires a very reduced dynamic model (a reduced number of states) but retaining the nonlinear and time-varying nature of the dynamics. The designed observer is tested on a large complete model of the bioreactor, and preliminary simulations show promising results in terms of precision and robustness.