Process-level dynamics

In order to obtain a description of the dynamics after the fast transient, we first recognize that the equations describing the process evolution in the fast time scale can be replaced, in the time scale t, by the corresponding quasi-steady-state constraints. These constraints are obtained by multiplying Equation (5.10) by i and considering the limit e — 0. Taking into account (5.13), the constraints that must be satisfied in the slow(er) time scale(s) are 

In this limit, the terms (G1(x)u1)/ei become indeterminate. On defining z = lim o(G1(x)u1)/ei, z G IR c m, as these finite but unknown terms, the process model after the fast boundary-layer dynamics takes the form 

Once the flow rates u1 have been specified by appropriate control laws, it is possible to differentiate the constraints in Equation (5.15) to obtain (after differentiating a sufficient number of times) a solution for the algebraic variables z. One differentiation in time will yield 

A minimal-order ODE representation of the system (5.17) can be subsequently obtained by employing a coordinate change of the form 

Specifically, the dynamics after the fast boundary layer will be described by 

---

From a practical perspective, this is the model that should be used to design a (multivariable) controller that manipulates the inputs us to fulfill the control objectives ys. It is important to note that the availability of a low-order ODE model of the process-level dynamics affords significant flexibility in designing the supervisory control system, since any of the available inversion- or optimization-based (e.g., Kravaris and Kantor 1990, Mayne et al. 2000, Zavala... 

Notice that the above model is still stiff, due to the presence of the parameter 2. Considering the limit 2 —> 0, corresponding to the absence of the inert component from the feed and a zero purge flow rate, we obtain the following description of the intermediate (process-level) dynamics ... 

When the grey level dynamic range in the image processed is small, usually because of a poor illumination or a non uniform lighting, it s possible to increase this dynamic range by a histogram transformation. This transformation affect the intensity distributions and increase the contrast. 

The demand for the research will cover the development of the novel algorithms utilizing parallel computation methods. The development of a hierarchical multi-scale paradigm will consolidate theoretical analysis and will lead to large-scale decision-making criteria of the process level design based on the first-principle dynamics. 

From physical considerations, at most C equations (with C being the number of chemical components) are required in order to completely capture the above overall, process-level material balance. Thus, we can expect the dimension of the system of equations describing the slow dynamics of the process to be at most C, and the equilibrium manifold (3.12) of the fast dynamics to be at most C-dimensional. 

Tikhonov s theorem (Theorem 2.1) indicates a further requirement that must be fulfilled by the controllers in the fast time scale in order for the time-scale decomposition developed above to remain valid, these controllers must ensure the exponential stability of the fast dynamics. From a practical point of view, this is an intuitive requirement one cannot expect stability and control performance at the process level if the operation of the process units is not stable. 

Using the methods presented in Chapter 2, the above formulation can be used to derive a state-space realization of the slow dynamics of the type in Equation (2.48). The resulting low-dimensional model should subsequently form the basis for formulating and solving the control problems associated with the slow time scale, i.e., stabilization, output tracking, and disturbance rejection at the process level. 

We will first concentrate on studying the process dynamics, so let us consider a numerical experiment that consists of starting a dynamic simulation of the process from initial conditions that are slightly perturbed from the nominal, steady-state values of the state variables. Although material holdups are stabilized using the proportional controllers in Equation (4.40), in view of the process-level operating objective stated above, this can be considered an open-loop simulation. 

The flow rates us of the streams outside the recycle loop appear as the manipulated inputs available for controlling the overall, process dynamics (5.21) in the intermediate time scale. Control objectives at the process level include the product purity, the stabilization of the total material holdup, and setting the production rate. 

Figure 5.8 Three distinct time horizons in the dynamic response of the reactor-condenser process core. Top fast, unit-level dynamics. Middle the total holdup of component A has an intermediate response time. Bottom the total holdup of impurity evolves in the slowest time scale. The plots depict simulation results with initial conditions slightly perturbed from their steady-state values.

We proposed a method for deriving nonlinear low-dimensional models for the dynamics in each time scale. Subsequently, we proposed a hierarchical controller design framework that takes advantage of the time-scale multiplicity, and relies on a multi-tiered structure of coordinated decentralized and supervisory controllers in order to address distributed and process-level control objectives. 

Subsequently, we considered the control implications of our findings, and showed that control objectives related to the energy dynamics of the individual units (e.g., temperature control) should typically be addressed in the fast time scale. On the other hand, control objectives related to the energy dynamic at the process level (such as managing energy use) should be addressed in the slow time scale. These concepts were illustrated through several examples and a simulation case study. 

Orrillo AG, Escalante AM, Furlan RLE (2008) Covalent double level dynamic combinatorial libraries selectively addressable exchange processes. Chem Commun 2008 5298-5300... 

To have 30% of this compound in its unionised form when the hulk pH is 7, would require a surface pH of about 3.4, which is lower than anticipated. Again, however, we must remember that absorption and ionisation processes are dynamic processes. As the unionised species is absorbed, so the level of [U] in the bulk falls and, because of a shift in equilibrium, more of the unionised species appears in the bulk. In fact, if a pH of 5.3 is taken as the pH of the absorbing surface, the results in Table 9.2 become more explicable, as we discuss in the next section. 

On the instance level, dynamic task nets, products, and resources used within a specific development process are modeled. On the application model side, similar process information is foremost contained in C3 nets. Application modeling experts create process templates in the form of C3 nets to define best-practice work processes. These C3 nets can be transferred structurally into dynamic task nets. Currently, a set of structural restrictions has to apply for the C3 nets used. Dynamic task nets can be generated on the tool side as the surrogates of the process templates modeled as C3 nets. We have realized an integrator for the mapping of C3 nets into dynamic task nets (see Sect. 3.2). 

It is no exaggeration to state that modern analytical methods can detect most of the natural elements in soils at some level of concentration. The specific elemental composition of each particular soil reflects, to a degree modified over time by weathering, the chemical composition of the parent material from which the soil formed. However, knowledge of a soil s composition in terms of total elemental content is usually not very useful when it comes to understanding the processes and dynamics of element availabihty and cycling. Nevertheless, if elemental concentrations are... 

---