Use this URL to cite or link to this record in EThOS:
Title: Integrated decision support for flexible multipurpose plants
Author: Rickard, Julian Graham
ISNI:       0000 0004 2675 2446
Awarding Body: Imperial College London (University of London)
Current Institution: Imperial College London
Date of Award: 2011
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Traditionally, the bulk of chemical manufacturing has taken place in large continuous facilities, since the economies of scale and low labour requirements meant that this offered the lowest production costs. Batch plants were used only for products that were difficult or uneconomic to convert to continuous operation. However, reduced margins in bulk chemicals have led to renewed interest in small flexible batch plants capable of producing a number of products. These plants allow a manufacturer to gain a marketing edge by tailoring products to specific customer requirements. Due to their extremely dynamic nature, these plants are typically difficult to operate close to their maximum capacity, and over the years many attempts have been made to use computer technology to improve the management and control of these plants. Much of this technology is beginning to see industrial use, but less attention has been paid to how various software solutions work together. This thesis focuses on integrating plant scheduling software that decides which products should be made when, with the supervisory software that manages the execution of that schedule in the plant control system. The approach taken is to build a single data model, and then to use the structure of this model to pass information between the software packages. By focusing on a model, this approach is generic to any plant which can be modelled, and is easily adapted as plant and products change. This offers significant advantages over approaches focusing on mapping, which are generally bespoke and inflexible, or to approaches connecting the various systems into a common database. Once information can be freely passed between applications, the triggers for off-line rescheduling, in order to reoptimise the schedule when conditions change from those under which the initial schedule was created, can be studied, and a number of alternative decision primitives are proposed. An implementation of the model and associated algorithms is demonstrated on a number of examples, highlighting the consistency and robustness of the approach.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available