Determination of storage tanks location for optimal short-term scheduling in multipurpose/multiproduct batch-continuous plants under uncertainties

Abstract

A multipurpose/multiproduct plant has to deal with many combinations of tasks sequences and operation rates that lead to accumulation problems. These problems can be handled using storage tanks, but usually their location within the flowsheet is predetermined and not subject to optimization, missing the opportunity to better satisfy the customers. In this work we will determinate the optimal location of storage tanks for the short-term scheduling under uncertainty. A hybrid simulation-based optimization (SBO) strategy was developed and implemented to solve the problems combining stochastic and deterministic solution algorithms.

Joint optimization of pricing and planning decisions in divergent supply chain

“…The recipe of a product p needs of units of the basic component b . The use of such recipe parameters are common in multiproduct optimization models for representing the amount of each component needed to produce a final product (Gallego and van Ryzin, ; Kong, ) or the amount of each resource needed to perform a task in the production process (Durand et al., ; Floudas and Lin, ; Fumero et al., ).…”

Comparing the Performances of Two Techniques for the Optimization Under Parametric Uncertainty of the Simultaneous Design and Planning of a Multiproduct Batch Plant

“…The simulation-based optimization (SbO) algorithm proposed for the comparison is a variation of the one presented in Durand, Mele and Bandoni (2011) and Durand et al (2012) can be followed from Usually, the expected value is utilized and then is returned to the GA outer loop, which uses it to search for the optimum combination of first-stage variables' values. A filter is utilized to avoid the use of the inner loop when an already-tried combination of first-stage variables is chosen again (the filter returns the same expected value of the objective function).…”

“…46 The variation introduced in this work with respect to the SbO strategy described in Durand, Mele and Bandoni (2011) and Durand et al, (2012) is that the simulation step in the inner loop is replaced with an LP optimization for each sample of the uncertain parameters.…”