Hybrid modelling approaches in process intensification: A thorough review

Hybrid modelling has emerged as a powerful approach in process intensification, integrating first-principles models with data-driven models to optimise industrial processes. This review provides a comprehensive analysis of the application of hybrid modelling in process intensification, examining its role in enhancing efficiency, sustainability, and adaptability in chemical and bioprocess industries. The paper discusses various hybrid modelling strategies, including parallel, serial, nested architectures and physics-informed machine learning models, demonstrating their effectiveness in addressing complex engineering challenges. The applications of hybrid modelling are reviewed concerning the four key subgroups of PI: time, energy, structure, and synergy, showcasing their impact in reducing process duration, optimising energy use, integrating unit operations, and enhancing system design. By exploiting hybrid modelling techniques, industries can overcome data limitations, improve predictive accuracy, and accelerate the development of next-generation processes.