دانلود رایگان مقاله لاتین تجزیه مقیاس در سیستم واکنش پیچیده از سایت الزویر
عنوان فارسی مقاله:
زمان تجزیه مقیاس در سیستم واکنش پیچیده: تجزیه و تحلیل نظری نمودار
عنوان انگلیسی مقاله:
Time scale decomposition in complex reaction systems: A graph theoretic analysis
سال انتشار : 2016
برای دانلود رایگان مقاله تجزیه مقیاس در سیستم واکنش پیچیده اینجا کلیک نمایید.
مقدمه انگلیسی مقاله:
1. Introduction
Complex reaction networks are present in numerous chemical and biochemical systems, such as combustion, pyrolysis, nanoparticle synthesis, catalytic conversion of hydrocarbons, and cell metabolism. These reaction networks are of particular recentinterest because of the emergence of new feedstocks and chemistries, e.g. for biomass and methane processing. Microkinetic modeling is an essential step towards rigorous design, optimization, and control of these reaction systems; however, the development of microkinetic models, with the underlying parameter estimation problem, is computationally challenging, with two key challenges being model stiffness and size. Stiffness arises from the difference in the order of magnitude of reaction rate constants, while the large model size is due to the large number of species and reactions typically present in such networks. Although there exist numerical methods for simulation of large scale, stiff models, the use of such models in optimization-based tasks (e.g. parameter estimation, control) results in ill-conditioning of the corresponding optimization task. Model reduction methods for kinetic simpli- fication involving lumping, sensitivity analysis, and time-scale analysis are generally used to address these challenges (Okino andModel reduction methods based on time-scale analysis include numerical approaches like the computational singular perturbation method (Lam and Goussis, 1989, 1994; Massias et al., 1999) where the eigenvalues of the Jacobian of the kinetic system of differential equations are used to identify the slow invariant manifold (Fenichel, 1979); the intrinsic low-dimensional manifold method (Maas and Pope, 1992) where an eigenvalue-eigenvector decomposition of the Jacobian matrix is performed with the assumption that the fast subspace vanishes quickly (Pope, 1997; Yang and Pope, 1998); geometric-based analysis (Fraser, 1988; Roussel and Pope, 1991) where a comprehensive investigation of the features of trajectories in the concentration phase space starting from many different initial conditions is used; and analytical, projection-based methods (Vora and Daoutidis, 2001; Gerdtzen et al., 2004; Adomaitis, 2016; Remmers et al., 2015; Lee and Othmer, 2010; Prescott and Papachristodoulou, 2014). All of these methods, however, require considerable computational effort in practical applications to complex, large scale systems (Lebiedz, 2004) and have been mostly applied to homogeneous reaction systems. Alternatively, mechanism reduction methods based on reaction rate evaluation (Lu and Law, 2006; Lu et al., 2009; PepiotDesjardins and Pitsch, 2008; Susnow et al., 1997) allow elimination of unimportant species and reactions from the reaction network, thereby, reducing the computational complexity of the system. The reaction rate evaluation is possible when the system has well-defined kinetics like in the case of gas-phase chemistry, due to the existence of a vast kinetic database for gas-phasechemical reactions (NIST Chemical Kinetics Database). For systems where there is significant uncertainty in the kinetic parameters, eliminating unimportant species and reactions based on an approximate set of kinetic constants may lead to erroneous results. In this research, a graph-theoretic framework is proposed for generation of non-stiff reduced models of isothermal reaction systems with fast and slow reactions. A directed bi-partite graph is used to represent the reaction network and the reactions are characterized as fast or slow using a kinetic threshold and an equilibrium tolerance. Cycles that correspond to closed walks are then used to identify interactions between species participating in fast/equilibrated reactions. Subsequently, an algorithm which connects these cycles to generate pseudo-species that evolve in the slow time scale alone is presented. The result is an automated, generic procedure for generating non-stiff reduced models in terms of these pseudo-species, while enforcing typical quasi-equilibrium or complete conversion constraints for fast reactions. The efficacy of the developed framework is illustrated through its application on two chemical reaction systems: 1-butene cracking and carbon metabolism in erythrocytes.
برای دانلود رایگان مقاله تجزیه مقیاس در سیستم واکنش پیچیده اینجا کلیک نمایید.
کلمات کلیدی:
22222
