Comparison of physics‐based and data‐driven modelling techniques for dynamic optimisation of fed‐batch bioprocesses

Abstract

The development of digital bioprocessing technologies is critical to operate modern industrial bioprocesses. This study conducted the first investigation on the efficiency of using physics‐based and data‐driven models for the dynamic optimisation of long‐term bioprocess. More specifically, this study exploits a predictive kinetic model and a cutting‐edge data‐driven model to compute open‐loop optimisation strategies for the production of microalgal lutein during a fed‐batch operation. Light intensity and nitrate inflow rate are used as control variables given their key impact on biomass growth and lutein synthesis. By employing different optimisation algorithms, several optimal control sequences were computed. Due to the distinct model construction principles and sophisticated process mechanisms, the physics‐based and the data‐driven models yielded contradictory optimisation strategies. The experimental verification confirms that the data‐driven model predicted a closer result to the experiments than the physics‐based model. Both models succeeded in improving lutein intracellular content by over 40% compared to the highest previous record; however, the data‐driven model outperformed the kinetic model when optimising total lutein production and achieved an increase of 40–50%. This indicates the possible advantages of using data‐driven modelling for optimisation and prediction of complex dynamic bioprocesses, and its potential in industrial bio‐manufacturing systems.

A new model to analyze the temperature effect on the microalgae performance at large scale raceway reactors

“…Biological microalgae models can help to estimate and maximize crop productivity Béchet et al (2016), as well as characteristic parameters that can be used in control systems to maximize biomass production (Costache et al, 2013; DelRio‐Chanona, Ahmed, et al, 2019; Pawlowski et al, 2015). However, although there exist some studies combining the microalgae productivity and culture temperature (Bernard & Rémond, 2012; James & Boriah, 2010), most existing biological models do not take the culture temperature into account, what is a limiting factor in the analysis of the microalgae productivity results (Huesemann et al, 2016; Karemore et al, 2020; Ras et al, 2013; Singh & Singh, 2015).…”

Synergising biomass growth kinetics and transport mechanisms to simulate light/dark cycle effects on photo‐production systems

“…Increasing incident light intensity (offsetting light attenuation) and enhancing culture mixing (shortening light/dark cycles) are two strategies to facilitate biomass growth in a PBR (Schulze et al, 2020). Although model‐based optimal control of light intensity has been reported previously (Del Rio‐Chanona et al, 2019; Koller et al, 2017), optimal control of gas inflow rate to enhance culture mixing has never been achieved as such a model was never proposed before.…”

“…Although model-based optimal control of light intensity has been reported previously (Del Rio-Chanona et al, 2019;Koller et al, 2017), optimal control of gas inflow rate to enhance culture mixing has never been achieved as such a model was never proposed before.…”

Machine learning for algal biofuels: a critical review and perspective for the future

“…11 Modeling, optimization, and state estimation of fed-batch systems have been successfully performed using kinetic models. 12 However, the predictive capacities of these models are usually low due to the reduction of a vast number of metabolic reactions into a few parameters and small sets of differential equations; any changes in the base conditions may have significant effects on model accuracy. 10 Data based models have been used as an alternative to kinetic based models in recent years because they can better predict a wide range of untested conditions than kinetic models because they use larger datasets that are valid for a wider spectrum of operating conditions and more variables for better description of the characteristics and behavior of the process.…”