|
Honeywell case studies
These seven case studies explain how Honeywell
have used Xpress-MP.
Flowsheet Decomposition Heuristic for Scheduling:
A Relax & Fix Method
Jeff Kelly & John Mann, Honeywell Hi-Spec
Solutions
January 2004
This note describes a relax-and-fix heuristic
that has been found useful when solving large production network
planning and scheduling problems. It uses Xpress-Mosel to make the
heuristic easily implementable. Three examples help illustrate the
effectiveness of this heuristic.
[pdf download]
A Staged Branch-and-Bound Acceleration Heuristic
for Production Scheduling
Jeff Kelly, Honeywell Hi-Spec Solutions
Submitted to Computers and Chemical Engineering,
December 2003
A short paper on a very simple technique deployed
in Xpress-Mosel to help find quickly integer-feasible solutions
to production scheduling problems in the process industries A classic
batch example is detailed. It shows how easy it is to model and
implement ad hoc B&B searches using the Xpress-Optimizer within
Mosel as well as providing an effective technique to help find integer-feasible
solutions quicker when the B&B is taking too long.
[pdf download]
Next-Generation Refinery Scheduling Technology
Jeff Kelly, Honeywell Hi-Spec Solutions
Presented at the NPRA Plant Automation and Decision
Support Conference, San Antonio, Texas, September 2003
This paper uses crude-oil blend scheduling as
a specific example to highlight how new optimization technology
can be utilized to identify multiple schedules that are logistically
feasible and then multiple variants of these feasible schedules
that optimize quality parameters. It provides an application of
Xpress-SLP, implemented through the Xpress-Mosel modeling language.
[pdf download]
Chronological Decomposition Heuristic for Scheduling:
A Divide & Conquer Method
Jeff Kelly, Honeywell Hi-Spec Solutions
Published in the December 2002 issue of the AIChE
Journal Vol. 48, No. 12, pages 2995-2999.
This article describes the chronological
decomposition heuristic (CDH), which has been developed using Xpress-Mosel
and the Xpress-Optimizer. The heuristic is a simple time-based divide-and-conquer
strategy intended to find rapidly, integer-feasible solutions to
production scheduling optimization problems of practical scale.
It was specifically designed for production scheduling optimization
problems found in the manufacturing of petroleum distillates, petrochemicals,
chemicals and pharmaceuticals which are formulated by discretizing
the temporal dimension using a pre-specified time grid with fixed
time-period spacing.
[pdf download]
Smooth-and-Dive Accelerator: A Pre-MILP Primal
Heuristic applied to Scheduling
Jeff Kelly, Honeywell Hi-Spec Solutions
Accepted (December 2002) for publication by the
journal 'Computers and Chemical Engineering'.
This article describes an effective and
simple primal heuristic to greedily encourage a reduction in the
number of binary or 0-1 logic variables before an implicit enumerative-type
search heuristic is deployed to find integer-feasible solutions
to "hard" production scheduling problems. The heuristic
has been developed using Xpress-Mosel and the Xpress-Optimizer.
The effectiveness of the heuristic is illustrated by its application
to an oil-refinery’s crude-oil blendshop scheduling problem.
[pdf download]
On the Formulation of Petroleum
and Petrochemical Planning Optimization Models
Jeff Kelly, Honeywell Hi-Spec Solutions
Chemical Engineering Progress magazine in January
2004, Vol. 100, No.1.
This article describes some of the non-linear
formulations used to describe petroleum and petrochemical planning
models. The introduction also provides a brief overview of the technology
available to solve these problems and the types of problems that
are solved. It provides insight into the complexities of these types
of planning applications.
[pdf download]
Reconcile Quantity and Time to Diagnose Production
Defects in the Process Industries
Jeff Kelly, John Mann and Douglas Moffat, Honeywell
Hi-Spec Solutions
Submitted to the 'Journal of Process Control'.
This article illustrates one way in which the
Xpress QP Optimizer is used in the process industries. It describes
a method to help identify production anomalies in the data, reconciling
both quantity and time of the production execution data. It uses
two quadratic programming (QP) problems to minimize the sum of squares
of adjustments to measured flows, inventories and start and end-times.
The technique is demonstrated using the Xpress QP Optimizer and
the Xpress-Mosel modeling system to solve a representative example.
[pdf download]
|
