ABSTRACT
The chapter presents an overview of possible hybrid modeling application areas in biochemical processes. The applications cover hybrid model-based process monitoring, optimization, and control (including hybrid optimization techniques and hybrid control systems). The main benefit of hybrid modeling arises from the intelligent knowledge integration/fusion from different information sources. It can lead to better-quality bioprocess monitoring (including real-time state estimation), more accurate process control, and a better understanding of the biochemical processes. More-accurate models also allow for the thorough exploration of the state space of the processes, to better predict behavior and make use of these models for more-precise model-based optimization. The application of hybrid models for monitoring can increase biochemical process safety and product quality. Hybrid modeling techniques in biochemical engineering not only cover hybrid models of biochemical processes but also hybrid optimization techniques (combining expert suggestions for optimization, traditional gradient-based optimization methods, and stochastic optimization methods such as simulated annealing as well genetic and evolutionary programming, etc.) and hybrid (fuzzy, neuro-fuzzy, and expert) control systems. The application of such optimization and control techniques can further enhance the model development process and control quality, respectively. Hybrid modeling can be successfully applied for fault detection and analysis, thereby contributing to higher process safety and performance.
