| Model order reduction is becoming increasingly important for the analysis, design, and control of chemical processes. In this work, the general classes of model reduction methods are examined along with new techniques and ideas.; Two methods based on time-scale analysis, each suited to different situations, are presented. When little is known beyond the mathematical description of a reaction system, the fast and slow variables are identified by scaling the rate constants, concentrations, and time variables. When concentration data are available, the Jacobian of the kinetic operator and its eigenvalues are used to identify regions that exhibit different time-scale behavior. The generated reduced models are non- stiff, computationally efficient, and valid over a range of initial conditions. The utility of these methods is illustrated by simplifying a description of cyclohexane oxidation, and incorporating the reduced models into the equations describing dynamics of a CSTR.; Other physical characteristics may be used to separate the correlations between variables in a system. The comprehensive description of a single-file diffusion system is reduced to modeling pairs of overlapping cells in a pore. This reduces the order of the system so previously uninvestigated situations can be simulated. This extension of model reduction ideas shows the wide applicability of the model order reduction, introduces new academic communities to these ideas, and facilitates the simulation and understanding of complex systems in general. |
