Complex Deterministic Models

Other processes that lead to nonlinear compartmental models are processes dealing with transport of materials across cell membranes that represent the transfers between compartments. The amounts of various metabolites in the extracellular and intracellular spaces separated by membranes may be sufficiently distinct kinetically to act like compartments. It should be mentioned here that Michaelis-Menten kinetics also apply to the transfer of many solutes across cell membranes. This transfer is called facilitated diffusion or in some cases active transport (cf. Chapter 2). In facilitated diffusion, the substrate combines with a membrane component called a carrier to form a carrier-substrate complex. The carrier-substrate complex undergoes a change in conformation that allows dissociation and release of the unchanged substrate on the opposite side of the membrane. In active transport processes not only is there a carrier to facilitate crossing of the membrane, but the carrier mechanism is somehow coupled to energy dissipation so as to move the transported material up its concentration gradient. 

The branching pattern of the vascular system and the blood flow through it has continued to be of interest to anatomists, physiologists, and theoreticians [4,317,318]. The studies focusing on the geometric properties such as 

---

Aside from the continuity assumption and the discrete reality discussed above, deterministic models have been used to describe only those processes whose operation is fully understood. This implies a perfect understanding of all direct variables in the process and also, since every process is part of a larger universe, a complete comprehension of how all the other variables of the universe interact with the operation of the particular subprocess under study. Even if one were to find a real-world deterministic process, the number of interrelated variables and the number of unknown parameters are likely to be so large that the complete mathematical analysis would probably be so intractable that one might prefer to use a simpler stochastic representation. A small, simple stochastic model can often be substituted for a large, complex deterministic model since the need for the detailed causal mechanism of the latter is supplanted by the probabilistic variation of the former. In other words, one may deliberately introduce simplifications or errors in the equations to yield an analytically tractable stochastic model from which valid statistical inferences can be made, in principle, on the operation of the complex deterministic process. 

Eyre and Pont, 2003 Jordan et ai, 2003 Nixon et al., 1995 Wollast, 1983 Table Al). In addition, a number of relatively complex, deterministic models of river nitrogen flux have been developed and applied to individual rivers (e.g., GWLF, Haith and Shoemaker, 1987 Lee et al., 1999 HSPF, BickneU et al., 1997 Fdoso et al., 2004 SWAT, Srinivasan et al., 1993). Several of these, as well as other models, have been described recently in Alexander et al. (2002). 

Because corrosion phenomena are complex, deterministic models evolve continually as restrictive hypotheses are eased when additional, empirical knowledge is acquired. In essence, it is the scientific method that nudges a model to reality. Hybrid deterministic models have been developed in fracture and fatigue where a particular property or parameter is considered to be statistically distributed. This statistical distribution is carefully chosen for implementation to selected parameters in the deterministic model (a true deterministic model retains its probabilistic aspect as a placeholder until the statistical scatter can be replaced with true mechanistic understanding). 

Rational air pollution control strategies require the establishment of reliable relationships between air quality and emission (Chapter 5). Diffusion models for inert (nonreacting) agents have long been used in air pollution control and in the study of air pollution effects. Major advances have been made in incorporating the complex chemical reaction schemes of photochemical smog in diffusion models for air basins. In addition to these deterministic models, statistical relationships that are based on aerometric data and that relate oxidant concentrations to emission measurements have been determined. 

Data processing and chemometrics are methods for extracting useful information from the complex instrumental and other data stream(s) (see Chapter 12) for process understanding and the development of deterministic models for process control. The hnal element, the analytical method development life cycle, will be discussed further within this chapter. 

A neural-network-based simulator can overcome the above complications because the network does not rely on exact deterministic models (i.e., based on the physics and chemistry of the system) to describe a process. Rather, artificia] neural networks assimilate operating data from an industrial process and learn about the complex relationships existing within the process, even when the input-output information is noisy and imprecise. This ability makes the neural-network concept well suited for modeling complex refinery operations. For a detailed review and introductory material on artificial neural networks, we refer readers to Himmelblau (2008), Kay and Titterington (2000), Baughman and Liu (1995), and Bulsari (1995). We will consider in this section the modeling of the FCC process to illustrate the modeling of refinery operations via artificial neural networks. 

Determination of optimal conditions in the lab, pilot and full-scale plants is among the most complex problems for a researcher and it belongs to the group of extreme problems. This complexity is caused by the nature of technological processes, which simultaneously include chemical reaction, transfer of mass, heat transfer and momentum transfer. It does not allow to form, by well-known theoretical knowledge, a deterministic model for establishing an optimum analytically. 

The concentrations of heavy metals both in river water and in river sediment are strongly changed by deposition-remobilization processes. The deterministic modeling of the transition between both environmental compartments is severely limited by the complex chemical, physical, and biochemical processes. 

Another situation when the use of the statistical model can be a good choice over the RSM is when the deterministic model is excessively complex. For example, when the process is described by a distributed parameters model, the steady-state mass and energy balances are differential equations. The use of differential equations as constraints in an optimization problem makes its solution difficult and increases the incidence of convergence problems. In this case, solving the optimization problem using the statistical model is much simpler. The statistical model can also be used when the computational effort to solve the optimization problem using the deterministic model is too high, as can be the case for real-time optimization problems. 

Complex specific 3a Deterministic Complex statistical model, partially mechanistic orientation deterministic use... 

Figure 9.28 The exact solution for the complex, c(t), of the Kolmogorov equations associating marginal probabilities with the number of particles. The solid line is the solution of the deterministic model. The areas of disks located at coordinates (t, c) are proportional to pc (t).

The possible states for substrate are 11 and 6 for the complex. R is a 66-dimensional matrix and the initial condition for the master equation is pio,o (0) = 1. Figures 9.27 and 9.28 show the associated probabilities for each state as functions of time for the substrate and the complex, respectively. As previously, the full markers are the expected values and the solid lines the solution of the deterministic model. Notably, the expectation of the stochastic model does not follow the time profile of the deterministic system. This is the main characteristic of nonlinear systems. 

The statistical modelling of a process can be applied in three different situations (i) the information about the investigated process is not complete and it is then not possible to produce a deterministic model (model based on transfer equations) (ii) the investigated process shows multiple and complex states and consequently the derived deterministic or stochastic model will be very complex (iii) the researcher s ability to develop a deterministic or stochastic model is limited. 

Statistical models have built in the same assumptions that are made in the development of deterministic models. The compactness of their formulation is what makes them extremely useful in the analysis of sequence lengths and complex copolymerizations. 

Comprehensive Models. This class of detailed deterministic models for copolymerization are able to describe the MWD and the CCD as functions of the polymerization rate and the relative rate of addition of the monomers to the propagating chain. Simha and Branson (3) published a very extensive and rather complete treatment of the copolymerization reactions under the usual assumptions of free radical polymerization kinetics, namely, ultimate effects SSH, LCA and the absence of gel effect. They did consider, however, the possible variation of the rate constants with respect to composition. Unfortunately, some of their results are stated in such complex formulations that they are difficult to apply directly (10). Stockmeyer (24) simplified the model proposed by Simha and analyzed some limiting cases. More recently, Ray et al (10) completed the work of Simha and Branson by including chain transfer reactions, a correction factor for the gel effect and proposing an algorithm for the numerical calculation of the equations. Such comprehensive models have not been experimentally verified. 

Deterministic simulation will be used to illustrate the effect of fitting a less complex to model to data arising from a more complex model. Concentrationtime data were simulated from a 3-compartment model with the following parameters CL = 1.5 Q2 = 0.15 Q3 = 0.015 VI = 1, V2 = 0.5 V3 = 0.25. A bolus unit dose of 100 was injected into the central compartment (VI). A total of 50 samples were collected equally spaced for 48-time units. The data are shown in Fig. 3.15 and clearly demonstrate triphasic elimination kinetics. Now suppose the LLOQ of the assay was 0.01. Data above this line are consistent with biphasic elimination kinetics. What hap-... 

Mathematical models may be classified into deterministic and stochastic models. For deterministic models, knowledge of the relationship between dependent and independent variables is necessary. Consequently, the complex nature of polymer-based heterogeneous materials is rather incompatible with such requirements. Hence, stochastic models become necessary either when the existing knowledge about the stimulus-response behavior of a system is not enough as to ascertain its behavior or when it is not possible to build an efficient deterministic model able to score the system response. 

The bandwidth of the traditional systems will become more and more of a problem, given the need to pass more and more data and to transfer more complex object models. This will be especially true where control is distributed down to the device level. With the availability of deterministic Ethernet (using protocols such as MMS and running at speeds of 100 Mb or higher), previous objections to using the Ethernet in a real-time control environment will largely disappear. 

Differences such as these are even more pronounced with regard to maintenance. Few facilities have equal levels of maintenance support around the clock—typically it will take longer to find maintenance help at nights, and on the weekends. There is no way that a deterministic model can incorporate complexities and variations of this type. 

Yet, in practice, even greater complexities may exist. For example, the repair of Pump P-IOIA may require a particular spare part if the facility warehouse has only a limited supply of that part, it may run out. Then, if the pump fails, the repair time will include the time it takes to order and ship the new part. Once more, such complexities are quite impossible to handle with deterministic models, yet stochastic models can incorporate them quite readily. 

---