Mathematical model constraints

Development of Process (Mathematical) Models Constraints in optimization problems arise from physical bounds on the variables, empirical relations, physical laws, and so on. The mathematical relations describing the process also comprise constraints. Two general categories of models exist ... 

In this section we consider the boundary value problem for model equations of a thermoelastic plate with a vertical crack (see Khludnev, 1996d). The unknown functions in the mathematical model under consideration are such quantities as the temperature 9 and the horizontal and vertical displacements W = (w, w ), w of the mid-surface points of the plate. We use the so-called coupled model of thermoelasticity, which implies in particular that we need to solve simultaneously the equations that describe heat conduction and the deformation of the plate. The presence of the crack leads to the fact that the domain of a solution has a nonsmooth boundary. As before, the main feature of the problem as a whole is the existence of a constraint in the form of an inequality imposed on the crack faces. This constraint provides a mutual nonpenetration of the crack faces ... 

Once the objective and the constraints have been set, a mathematical model of the process can be subjected to a search strategy to find the optimum. Simple calculus is adequate for some problems, or Lagrange multipliers can be used for constrained extrema. When a Rill plant simulation can be made, various alternatives can be put through the computer. Such an operation is called jlowsheeting. A chapter is devoted to this topic by Edgar and Himmelblau Optimization of Chemical Processes, McGraw-HiU, 1988) where they list a number of commercially available software packages for this purpose, one of the first of which was Flowtran. 

No single method or algorithm of optimization exists that can be apphed efficiently to all problems. The method chosen for any particular case will depend primarily on (I) the character of the objective function, (2) the nature of the constraints, and (3) the number of independent and dependent variables. Table 8-6 summarizes the six general steps for the analysis and solution of optimization problems (Edgar and Himmelblau, Optimization of Chemical Processes, McGraw-HiU, New York, 1988). You do not have to follow the cited order exac tly, but vou should cover all of the steps eventually. Shortcuts in the procedure are allowable, and the easy steps can be performed first. Steps I, 2, and 3 deal with the mathematical definition of the problem ideutificatiou of variables and specification of the objective function and statement of the constraints. If the process to be optimized is very complex, it may be necessaiy to reformulate the problem so that it can be solved with reasonable effort. Later in this section, we discuss the development of mathematical models for the process and the objec tive function (the economic model). 

The scope of a study required to satisfy these goals will be dependent upon the extent of the risk, the depth of the study required, and the level of resources available (mathematical models and tools and skilled people to perform the study and any internal or external constraints). 

The natural frequency, co associated with the mode shape that exhibits a large displacement of the pump is compared with the fundamental frequency, of the wall. If co is much less than ru, then the dynamic interaction between the wall and the loop may be neglected, but the kinematic constraint on the pump imposed by the lateral bracing is retained. If nearly equals nr , the wall and steam supply systems are dynamically coupled. In which case it may be sufficient to model the wall as a one-mass system such that the fundamental frequency, Wo is retained. The mathematical model of the piping systems should be capable of revealing the response to the anticipated ground motion (dominantly translational). The mathematics necessary to analyze the damped spring mass. system become quite formidable, and the reader is referred to Berkowitz (1969),... 

Mathematical Models. As noted previously, a mathematical model must be fitted to the predicted results shown In each factorial table generated by each scientist. Ideally, each scientist selects and fits an appropriate model based upon theoretical constraints and physical principles. In some cases, however, appropriate models are unknown to the scientists. This Is likely to occur for experiments Involving multifactor, multidisciplinary systems. When this occurs, various standard models have been used to describe the predicted results shown In the factorial tables. For example, for effects associated with lognormal distributions a multiplicative model has been found useful. As a default model, the team statistician can fit a polynomial model using standard least square techniques. Although of limited use for Interpolation or extrapolation, a polynomial model can serve to Identify certain problems Involving the relationships among the factors as Implied by the values shown In the factorial tables. 

In parameter estimation we are occasionally faced with an additional complication. Besides the minimization of the objective function (a weighted sum of errors) the mathematical model of the physical process includes a set of constrains that must also be satisfied. In general these are either equality or inequality constraints. In order to avoid unnecessary complications in the presentation of the material, constrained parameter estimation is presented exclusively in Chapter 9. 

In some cases besides the governing algebraic or differential equations, the mathematical model that describes the physical system under investigation is accompanied with a set of constraints. These are either equality or inequality constraints that must be satisfied when the parameters converge to their best values. The constraints may be simply on the parameter values, e.g., a reaction rate constant must be positive, or on the response variables. The latter are often encountered in thermodynamic problems where the parameters should be such that the calculated thermophysical properties satisfy all constraints imposed by thermodynamic laws. We shall first consider equality constraints and subsequently inequality constraints. 

Given a set of data points (x y,), i=l,...,N and a mathematical model of the form, y = f(x,k), the objective is to determine the unknown parameter vector k by minimizing the least squares objective function subject to the equality constraint, namely... 

Most of the constrained parameter estimation problems belong to this case. Based on scientific considerations, we arrive quite often at constraints that the parameters of the mathematical model should satisfy. Most of the time these are of the form,... 

Using the above transformation, we are able to perform an unconstrained search over k,. For any value of k the original parameter k, remains within its limits. When k, approaches very large values (tends to infinity), k, approaches its lower limit, kmm j whereas when k, approaches very large negative values (tends to minus infinity), k, approaches its upper limit, kmax>1. Obviously, the above transformation increases the complexity of the mathematical model however, there are no constraints on the parameters. 

In this section, the above mathematical model is applied to a literature example shown in Fig. 2.2 (Ierapetritou and Floudas, 1998). The SSN representation is given in Fig. 2.3b. Table 2.1 gives data for this example. 5 time points and a 12-h time horizon were used. Using less time points leads to a suboptimal solution with an objective value of 50, and using more time points than 5 did not improve the solution. It is worthy of note that, in this particular example, constraint (2.13) is redundant as mentioned earlier, since each unit is only performing one task. 

These correspond to constraints (2.14) and (2.15) given in the mathematical model. 

Table 4.4 is the summary of the mathematical model and the results obtained for the case study. The model for scenario 1 involves 637 constraints, 245 continuous and 42 binary variables. Seventy nodes were explored in the branch and bound algorithm. The model was solved in 1.61 CPU seconds, yielding an objective value (profit) of 1.61 million over the time horizon of interest, i.e. 6 h. This objective is concomitant with the production of 850 t of product and utilization of 210 t of freshwater. Ignoring any possibility for water reuse/recycle, whilst targeting the same product quantity would result in 390 t of freshwater utilization. Therefore, exploitation of water reuse/recycle opportunities results in more than 46% savings in freshwater utilization, in the absence of central reusable water storage. The water network to achieve the target is shown in Fig. 4.14. 

This mathematical model is made up of two sets of constraints that are built within the same framework. One set of constraints focuses on the exploration of water reuse/recycle opportunities and the other on proper sequencing to capture the time dimension. Although this model has been presented in detail in Chapter 4, it is presented here in sufficient detail to facilitate understanding. 

The sequencing set of constraints focuses on capturing the time dimension, which is intrinsic in batch operations. The following constraints, which apply irrespective of the chosen scenario (scenario 1 or scenario 2), constitute the scheduling set of constraints for the proposed mathematical model. 

The two sets of constraints presented in Sections 3.1 and 3.2 constitute an overall mathematical model, which is used in the proposed two-stage solution algorithm. 

The optimisation procedure presented in this chapter entails two stages as summarized in Fig. 5.3. In the first stage, a mathematical model for minimisation of freshwater requirement is solved based on maximum potential reusable water storage, gf. For clarity, this model will be referred to as model Ml in this chapter. In the second stage, the minimum freshwater requirement obtained from model Ml is used as an input parameter in another mathematical model for which the objective function is the minimisation of reusable water storage. This model will be referred to as model M2 in this chapter. Since different amounts of reusable water will be stored at various intervals within the time horizon of interest, the minimum reusable water storage capacity will correspond to the maximum amount of reusable water stored at any point within the time horizon of interest as obtained from model M2 (Constraints (5.40)). 

The mathematical formulation comprises of a number of mass balances and scheduling constraints. Due to the nature of the processes involved, the time aspect is prevalent in all the constraints in some form or another. A superstructure is used in the derivation of the mathematical model, as discussed in the following section. A description of the sets, variables and parameters can be found in the nomenclature list. 

As with the multiple contaminant wastewater minimisation model, the mathematical model for multiple storage vessels comprises of two modules, namely, a mass balance module and a scheduling module. The constraints that comprise the mass balance module are described first. 

Constraints (11.17) shows how the variation in duration due to the heat integration mode is accounted for in the mathematical model. It is very likely that the duration times will be affected by the mode of operation and this should not be ignored in the formulation. 

The design q>roblem can be approached at various levels of sophistication using different mathematical models of the packed bed. In cases of industrial interest, it is not possible to obtain closed form analytical solutions for any but the simplest of models under isothermal operating conditions. However, numerical procedures can be employed to predict effluent compositions on the basis of the various models. In the subsections that follow, we shall consider first the fundamental equations that must be obeyed by all packed bed reactors under various energy transfer constraints, and then discuss some of the simplest models of reactor behavior. These discussions are limited to pseudo steady-state operating conditions (i.e., the catalyst activity is presumed to be essentially constant for times that are long compared to the fluid residence time in the reactor). 

Equations 12.7.48 and 12.7.39 provide the simplest one-dimensional mathematical model of tubular fixed bed reactor behavior. They neglect longitudinal dispersion of both matter and energy and, in essence, are completely equivalent to the plug flow model for homogeneous reactors that was examined in some detail in Chapters 8 to 10. Various simplifications in these equations will occur for different constraints on the energy transfer to or from the reactor. Normally, equations 12.7.48 and 12.7.39... 

The ingredients of formulating optimization problems include a mathematical model of the system, an objective function that quantifies a criterion to be extremized, variables that can serve as decisions, and, optionally, inequality constraints on the system. When represented in algebraic form, the general formulation of discrete/continu-ous optimization problems can be written as the following mixed integer optimization problem ... 

The main tools used to provide global projections of future climate are general circulation models (GCMs). These are mathematical models based on fundamental physical laws and thus constitute dynamical representations of the climate system. Computational constraints impose a limitation on the resolution that it is possible to realise with such models, and so some unresolved processes are parameterised within the models. This includes many key processes that control climate sensitivity such as clouds, vegetation and oceanic convection [19] of which scientific understanding is still incomplete. 

Geochemists, following early theoretical work in other fields, have long considered the multicomponent equilibrium problem (as defined in Chapter 3) to be mathematically unique. In fact, however, this assumption is not correct. Although relatively uncommon, there are examples of geochemical models in which more than one root of the governing equations satisfy the modeling constraints equally... 

The SNP optimizer is based on (mixed-integer) linear programming (MILP) techniques. For a general introduction into MILP we refer to [11], An SAP APO user has no access to the mathematical MILP model. Instead, the modeling is done in notions of master data of example products, recipes, resources and transportation lanes. Each master data object corresponds to a set of constraints in the mathematical model used in the optimizer. For example, the definition of a location-product in combination with the bucket definition is translated into inventory balance constraints for describing the development of the stock level over time. Additional location-product properties have further influence on the mathematical model, e.g., whether there is a maximum stock-level for a product or whether it has a finite shelf-life. For further information on the master data expressiveness of SAP SCM we refer to [9],... 

---