Big Chemical Encyclopedia

Chemical substances, components, reactions, process design ...

Articles Figures Tables About

Model predictive control enhancements

Conventional gradient base optimisation techniques are not effective to deal with objective functions with multiple local minima and can be trapped in local minima. Particle swam optimisation (PSO) is a recently developed optimisation technique that can cope with multiple local minima. This paper proposes using PSO and stacked neural networks to find the optimal control policy for batch processes. A standard PSO algorithm and three new PSO algorithms with local search were developed. In order to enhance the reliability of the obtained optimal control policy, an additional term is added to the optimisation objective function to penalise wide model prediction confidence bormds. [Pg.375]

Advanced Process Control (APC) is the term used within Shell for multivariable model-based predictive control . It is used on top of the regulatory control to enhance the stability and operability of the plant and to loeally optimize parts of the plant. Typically advanced control is applied on a single reaetor or a single distillation column, although there is a tendency nowadays to enlarge the scope of model-based controllers to control ever larger parts of a plant with a single APC controller. [Pg.159]

Gradually the steady state model has been enhanced with dynamic capabilities and connected to the mill data historian PI. The model has then been validated against actual mill data, which has made it possible to resolve discrepancies of the model. The actual control strategy has also been modeled in order to facilitate analysis of grade and production rate changes. Some of the applications are minimizing the overflows, water usage and prediction of pulp quality parameters. [Pg.1038]

Rigorous models for stand-alone units also can provide significant benefits. Previously, we reported benefits of US 3,000 per day (US 0.15 per barrel) for the initial optimizer on the hydrocracking complex these benefits were in addition to those provided by model-predictive DMC control. For RWO, a revised model based on Aspen Hydrocracker (AHYC) was developed. It includes a catalyst deactivation block, which enhances maintenance turnaround planning by predicting future catalyst activity, product yields and product properties for a variety of assumed feeds and specified operating conditions. This information also is used to impose constraints on present-day operation. [Pg.259]

Methanol remains the most widely used modifier because it produces highly efficient separations, but it does not always produce the highest selectivity [8]. Recent studies have provided insight into the role of the modifier in enantioselectivity in SFC [69]. Blackwell and Stringham examined a series of phenylalanine analogues on a brush-type CSP and developed a model that allowed prediction of selectivity based on the bulk solvation parameters of various modifiers [70]. Careful choice of modifiers can be used to mask or enhance particular molecular interactions and ultimately provide control of selectivity [71]. [Pg.311]

One of the central problems in air pollution research and control is to determine the quantitative relationship between ambient air quality and emission of pollutants from sources. Effective strategies to control pollutants can not be devised without this information. This question has been mainly addressed in the past with source-oriented techniques such as emission inventories and predictive diffusion models with which one traces pollutants from source to receptor. More recently, much effort has been directed toward developing receptor-oriented models that start with the receptor and reconstruct the source contributions. As is the case with much of air pollutant research, improvements in pollutant chemical analysis techniques have greatly enhanced the results of receptor modeling. [Pg.364]

It was found that the solubility of C02 in component 7 (oil) could be used in much the same way as the mixing parameter to vary the relative rate of movement of C02 and oil, and thus control the extent of enhanced oil recovery in order to match the simulated oil recovery to the laboratory model results. Desirably, the solubility should fall in a narrow range in such matching, so that the value of solubility so obtained could be used for predictive purposes, just as a narrow range of mixing parameter is found to hold in matching laboratory and field miscible floods. [Pg.365]


See other pages where Model predictive control enhancements is mentioned: [Pg.249]    [Pg.132]    [Pg.219]    [Pg.308]    [Pg.525]    [Pg.220]    [Pg.424]    [Pg.486]    [Pg.2493]    [Pg.93]    [Pg.124]    [Pg.626]    [Pg.305]    [Pg.121]    [Pg.1054]    [Pg.228]    [Pg.610]    [Pg.455]    [Pg.527]    [Pg.1305]    [Pg.347]    [Pg.142]    [Pg.672]    [Pg.310]    [Pg.11]    [Pg.14]    [Pg.356]    [Pg.104]    [Pg.56]    [Pg.141]    [Pg.202]    [Pg.41]    [Pg.622]    [Pg.232]    [Pg.29]    [Pg.310]    [Pg.581]    [Pg.1603]    [Pg.237]    [Pg.2007]    [Pg.3638]    [Pg.121]   
See also in sourсe #XX -- [ Pg.198 ]




SEARCH



Control models

Model predictive control

Modeling Predictions

Modelling predictive

Prediction model

Predictive models

© 2024 chempedia.info