Data-Driven System Identification and 3DoF-KF Model Predictive Control For Integrating Processes: Application to a Microalgae Raceway Reactor
Year: 2025
Type: preprint
Abstract: The production of microalgae offers significant societal benefits to sustainability, among these, industrial-scale wastewater treatment and biomass production. Commercial raceway bioreactors exhibit complex dynamics featuring nonlinearities, delays, and integrating characteristics, further complicated by variations in ambient conditions (e.g. temperature, radiation) affecting reactor pH, whose accurate control is critical for optimal operations. To address the need for effective pH control, a datadriven system identification and control framework is presented, leading to high-performance closed-loop control meeting user requirements. This work introduces control-relevant input signal design guidelines for integrating systems, allocating signal power to prioritize closed-loop performance within a frequency band defined via Internal Model Control design. The input spectrum is realized in the time domain using plant-friendly multisine signals, with parsimonious AutoRegressive with exogenous inputs (ARX) models for data-driven estimation. A Three-Degree-of-Freedom Kalman Filter-based Model Predictive Control (3DoF-KF MPC) framework is presented, enabling independent tuning of control parameters for setpoint tracking and measured and unmeasured disturbance rejection. The end-to-end input signal design, experimental execution, ARX model estimation, and model predictive closed-loop control methodologies are experimentally validated through the pH control of a semi-industrial-scale microalgae photobioreactor facility in the presence of varying environmental conditions. These results display excellent setpoint tracking, active disturbance rejection, and robustness to uncertainty from an experimental design for integrating systems that can be effectively performed in the open-loop, all underscoring their importance in scalable and sustainable microalgae generation.
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4413966808