Parameter Identification of a Dynamic Model of Cultures of Microalgae Scenedesmus obliquus - An experimental study

This paper discusses the development of a dynamic model of cultures of microalgae
Scenedesmus obliquus in photobioreactors. Unknown parameters are estimated from experimental data collected from cultures operated in batch mode as well as with pulses of inlet substrates, and achieved at
different incident light intensities. The identification procedure involves the minimization of a weighted least squares criterion using a multistart strategy and a combination of Nelder-Mead and Levenberg-Marquardt algorithms. The parameter error covariance matrix is estimated using the Fisher Information Matrix, and confidence intervals are provided both for the parameters and the model prediction. The
latter is achieved through Monte Carlo simulation based on Latin hypercube sampling. The model validation is quite satisfactory, and the resulting model could be used for model-based control.