Structural Identifiability Analysis of Anaerobic Digestion Model Parameters Using Genetic Algorithm

The anaerobic digestion (AD) process is renowned for its dual role in waste recycling and renewable energy production, and has thus attracted significant attention. This process is characterized by its complexity and high sensitivity, which in turn complicate the models designed to predict its behavior. Models that rely on ordinary differential equations often contain numerous parameters. Identifying these parameters accurately requires the use of measurement-based parameter estimation techniques. The success of such techniques depends on each parameter having a single, unique value that aligns with the observed data. However, this assumption is undermined if the model is not structurally identifiable, based on the selected measurement outputs. This paper is devoted to the application of the Genetic Algorithm (GA) to conduct a comprehensive analysis of the structural identifiability of the Acidogenesis methanogenesis 2 (AM2) model parameters. We introduce a methodology to derive the confidence intervals for the parameter estimates using GA. The validity of our results is further corroborated by data from the Anaerobic Digestion model No. 01 (ADM1).