The management and control of material flow forms the core of manufacturing execution systems (MES) in the petrochemical industry. The bottleneck in the application of MES is the ability to match the material-flow model with the production processes. A dynamic material-flow model is proposed in this paper after an analysis of the material-flow characteristics of the production process in a petrochemical industry. The main material-flow events are described, including the movement, storage, shifting, recycling, and elimination of the materials. The spatial and temporal characters of the material-flow events are described, and the material-flow model is constructed. The dynamic material-flow model introduced herein is the basis for other subsystems in the MES. In addition, it is the subsystem with the least scale in MES. The dynamic-modeling method of material flow has been applied in the development of the SinoMES model. It helps the petrochemical plant to manage the entire flow information related to tanks and equipments from the aspects of measurement, storage, movement, and the remaining balance of the material. As a result, it matches the production process by error elimination and data reconciliation. In addition, it facilitates the integration of application modules into the MES and guarantees the potential development of SinoMES in future applications.
Currently, the management and control of material
flow is a topic that challenges the manufacturing
execution system (MES) community [ 1, 21, especially
in the processed product domain such as the petrochemical
industry. A part of this flow may be simply
treated with the method of the static material set, in
which the primordial principle is the enumeration of
the category and quantity of materials, but it is difficult
to adapt the dynamic environment of changing material
flow in this model. A much more smart technology
needs to be discovered and refined to work well [3].
Smart MESs require the services of manufacturing
entities, e.g. resources such as production-cells,
equipments, and workers, so as to achieve their production
needs 14-61. These resources could smartly
organize themselves, based on their own knowledge
and on a flexible cooperation policy, to cany out the
smart-product requests. However, this cooperation is
difficult, mainly because the elements of the factory
are heterogeneous.
The material-flow management and control in
petrochemical industries seems to be a big challenge
because there are physical and chemical changes of
materials during the complicated and long production
processes. Liquids, gases, and solid materials coexist.
Materials flow or stay in the network of devices and
storage areas, entering/leaving factory points, undergoing
the various processes such as mixing, separating,
alternating, recycling, and elimination of materials.
This material path is considered as a microconcept of
material flow. This diversity of modalities could be
homogenized and integrated by their encapsulation, in
an abstract manner, into models constructed through an
adaptive and dynamic procedure. Hence, smart MESsembedded with these models could also serve as
“holons”, able to treat different kinds of dynamic problems
occumng in complicated production processes.
The intention of this study is to propose a conceptual
solution as a meta-model for adaptively-driven
logistics-model construction. This solution is also
used to improve a design and simulation tool for
MESs of the Chinese petrochemical industry, called
the SinoMES V2.0 [3], toward the evolution of a holonic
manufacturing system (HMS).
A future objective is to present t h ~ sm eta-model
and its automatic and adaptive construction method as
an engineering tool that aids in the composition of
such systems, using a set of holons and their relationship
that has been previously developed and tested.
The quality of the meta-model would allow a reduction
in the time for system composition, and address
more accurately real material-flow conditions in the
production process. This is observable in the SinoMES
V2.0 context, some experiments based on it
demonstrating the potential of the solution.
Logistics management and control forms the core
content of MES including periodic service activities
such as planning and scheduling, yield accounting and
performance evaluation. Therefore, establishing a logistics
model that can reflect the logistics conditions
accurately and promptly is one of the key requirements
for the success of MES. The dynamic-logistics-modeling
technique based on incident rules proposed in this paper,
offers a possible method to solve the long-standing
problem of matching the logistics models to the dynamicproduction processes, and in addition, helps in
the construction of the information platform for further
applications of MES in process industries.
