Additional constraints

In order to ensure that all the reusable water from storage is completely reused within the chosen time horizon, a constraints for zero storage at the end of the time horizon is formulated as shown in Constraints (5.38). 

Alternatively, constraints (5.39) might be used. This constraints ensures that the overall amount of freshwater used equals the overall amount of effluent produced over the time horizon of interest. This condition ascertains that no water remains in reusable water storage at the end of the time horizon. 

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 solution procedure for the problem stated above involves two optimisation stages in which freshwater and reusable water storage capacity are minimised sequentially, as explained in Section 5.4.1 below. 

1 Two-Stage Optimisation Algorithm for Freshwater and Reusable Water Storage Minimisation 

---

Thus loops, utility paths, and stream splits offer the degrees of freedom for manipulating the network cost. The problem is one of multivariable nonlinear optimization. The constraints are only those of feasible heat transfer positive temperature difference and nonnegative heat duty for each exchanger. Furthermore, if stream splits exist, then positive bremch flow rates are additional constraints. 

Because the gin capacity is usually not sufficient to keep up with the harvesters, the harvested cotton is often stored in a compacted module and ginned at a later date. The type of storage or seed cotton processing may place additional constraints on the harvest process. If the seed cotton is to be placed in module storage, the cotton should not be harvested until the moisture content is 12% or less and the harvested seed cotton should be free of green plant material such as leaves and grass. 

Maximum likelihood methods are commonly used to estimate parameters from noisy data. Such methods can be applied to image restoration, possibly with additional constraints (e.g., positivity). Maximum likelihood methods are however not appropriate for solving ill-conditioned inverse problems as will be shown in this section. 

In the hope that additional constraints such as positivity (which must hold for the restored brightness distribution) may avoid noise amplification, we can seek for the constrained maximum likelihood (CML) solution ... 

The a priori penalty prior(x) oc — log Pr x allows us to account for additional constraints not carried out by the data alone (i.e. by the likelihood term). For instance, the prior can enforce agreement with some preferred (e.g. smoothness) and/or exact (e.g. non-negativity) properties of the solution. At least, the prior penalty is responsible of regularizing the inverse problem. This implies that the prior must provide information where the data alone fail to do so (in particular in regions where the noise dominates the signal or where data are missing). Not all prior constraints have such properties and the enforced a priori must be chosen with care. Taking into account additional a priori constraints has also some drawbacks it must be realized that the solution will be biased toward the prior. 

Equations 22.3-22.14 represent the simplest formulation of filled phantom polymer networks. Clearly, specific features of the fractal filler structures of carbon black, etc., are totally neglected. However, the model uses chain variables R(i) directly. It assumes the chains are Gaussian the cross-links and filler particles are placed in position randomly and instantaneously and are thereafter permanent. Additionally, constraints arising from entanglements and packing effects can be introduced using the mean field approach of harmonic tube constraints [15]. 

Clearly, if under the worst-case scenario no node can be found in the IGs that is labeled with all species present in a mixing constraint, then no potential violation of a mixing constraint exists. If, on the other hand, a potential violation has been detected, then we need to generate additional constraints on the temporal ordering of primitive operations so that we can prevent or negate the preconditions of a mixing constraint. 

The same applies to the other eigenvectors U2 and Vj, etc., with additional constraints of orthonormality of u, U2, etc. and of Vj, Vj, etc. By analogy with eq. (31.5b) it follows that the r eigenvalues in A must satisfy the system of linear homogeneous equations ... 

Although it is clear that there are N(M - N) orthogonality relations between the s and the L s, it is not clear why this is exactly equal to K, unless one has additional knowledge of < s and the ,s, but such knowledge would be incorporated in additional constraints which would have to be counted and would presumably alter the expression for K that was obtained. 

In this case, we have added one column of zeros they are needed to show how b2 is computed. Since b2 = 0 and c, = a0, the Routh criterion adds one additional constraint in the case of a third order polynomial ... 

The two additional constraints from the Routh array are hence... 

Once the optimum profile(s) has been established, its practicality for implementation must be assessed. For a continuous process, the equipment must be able to be designed such that the profile can be followed through space by adjusting rates of reaction, mass transfer, heat transfer, and so on. In a dynamic problem, a control system must be designed that will allow the profile to be followed through time. If the profile is not practical, then the optimization must be repeated with additional constraints added to avoid the impractical features. 

Now consider how to synthesize the reaction and separation system for a batch process. Start by assuming the process to be continuous and then, if choosing to use batch operation, the continuous steps are replaced by batch steps10. It is simpler to start with continuous process operation because the time dependency of batch operation adds additional constraints over and above those for continuous operation. 

Constraints might be applied for the sake of reducing the capital costs (e.g. to avoid long pipe runs). In addition, constraints might be applied to avoid complex heat integration arrangements for the sake of operability and control (e.g. to have heat recovery to a reboiler from a single source of heat, rather than two or three sources of heat). 

The models developed to take the PIS operational philosophy into account are detailed in this chapter. The models are based on the SSN and continuous time model developed by Majozi and Zhu (2001), as such their model is presented in full. Following this the additional constraints required to take the PIS operational philosophy into account are presented, after which, the necessary changes to constraints developed by Majozi and Zhu (2001) are presented. In order to test the scheduling implications of the developed model, two solution algorithms are developed and applied to an illustrative example. The final subsection of the chapter details the use of the PIS operational philosophy as the basis of operation to design batch facilities. This model is then applied to an illustrative example. All models were solved on an Intel Core 2 CPU, T7200 2 GHz processor with 1 GB of RAM, unless specifically stated. 

The model in the above form does not take into account the possibility of using latent storage, i.e. PIS operational philosophy. There are a number of additional constraints needed to fully capture this operational philosophy. 

The mass balance constraints given above would suffice if the process were continuous. However, due to the fact that the processes dealt with are batch processes, additional constraints are required to capture the discontinuous nature of the process. This implies that the time related constraints are necessary. 

---