The successful candidate will be the main programmer in the re-implementation and further development of a state of the art design support system; will also provide support to a small research group and sometimes interact with industrial users of the system.
Applicants should be competent in object-oriented programming in a UNIX environment, preferably with previous experience in Common Lisp or another symbolic language. Experience with database management systems (especially object-oriented), development graphical user interfaces, or artificial intelligence techniques will also be an advantage.
The starting date is 1st April 1996, and the starting salary is in the range 14,317 to 15,986.
Further particulars including details of the application procedure should be obtained from the Personnel Office, 1 Roxburgh Street, Edinburgh EH8 9TB or Tel: 0131 650 2511. (24 hour answering service). Please quote reference: 690104.
For informal discussion please contact Rene Banares-Alcantara via electronic mail (rene@chemeng.ed.ac.uk).
Closing date for receipt of applications: 29 February 1996.
Posts A & B - process systems engineering; design synthesis and optimisation;
Post C - optimisation.
Applicants should have a background in chemical engineering with a strong interest in mathematics and computing, or in mathematics, physics or another numerate discipline.
Posts are available for up to 3 years. Salary is on the AR1B scale (14,317 to 15,986). An exceptional candidate may be appointed on the AR1A scale.
Further particulars, including details of the application procedure, should be obtained from the Personnel Office, 1 Roxburgh Street, Edinburgh EH8 9TB. Please quote references 690076 (Posts A and B) and 690079 (Post C). Closing date for receipt of applications is 29 February 1996.
The successful candidate will be a chemical engineer with a degree from a department of recognised excellence, and will have a strong interest in developing computer software. Experience in some aspects of the following will be of advantage:
Candidates must be numerate and experienced in the use of at least one professional programming language. The post is available for up to 3 years, commencing as soon as possible.
Salary is on the AR1B scale (14,317 to 15,986). An exceptional candidate may be appointed on the AR1A scale. The successful candidate will be eligible for membership of the Universities Superannuation Scheme (USS).
The project will be carried out jointly in:
The successful candidate will work as a member of the project team consisting of a Research Associate in each department in addition to the supervisors.
The aim of the project is to develop methods which will allow the efficient solution of very large nonlinear programming (NLP) problems, with 100,000 or more variables, resulting from modelling chemical and similar processes. Such methods are expected to have application in the real-time optimisation of entire processes or process sites, as well as individual units or subsystems of a process.
Models of an integrated process plant are typically composed of sub-models for unit operations (columns, reactors, etc.) connected by streams (material and energy flows). It is often possible to break a large process down into smaller sub-processes by assuming values for a relatively small number of streams. This structure suggests the use of decomposition methods in which the sub-processes are solved by standard techniques such as NLP (nonlinear programming). The overall process is then optimised by a master problem which determines the best values of the linking variables. This is a nonsmooth problem which requires special techniques.
Decomposition introduces various theoretical and organisational issues. One of the main thrusts of the research programme will be to address these issues, leading on to the development of effective algorithms and software.
The main task of the Research Associate in Chemical Engineering will be:
Particular areas of application will include:
plant and site-wide utility systems, using compressor and turbine models; near-ideal rigorous distillation models with interlinked columns and large numbers of components and stages; non-ideal distillation models also with several large columns; and large scale pipe networks with multi-phase flow.
Heat exchanger, pump and valve models will be needed in all applications to demonstrate heat integration and for control. Because these models have some degree of non-linearity the effect of formulation decisions both on convergence of algorithms and convexity of models will be examined.
We intend to develop unit models suitable for on-line use, in which quantities such as exchanger heat transfer coefficients (which may be subject to fouling) are treated as unknown parameters (i.e. variables) to be estimated in a data reconciliation step. In the subsequent optimisation they are regarded as fixed.
The modelling environment will be one in which it is easy to construct models and also to introduce new solvers. The main responsibility for model construction and solver interfacing will be with the Research Associate in Chemical Engineering.
We may use ASCEND or an alternative `Flexible Modelling System', which can be easily interfaced to new solvers and is under development in the Department of Chemical Engineering. This may form the basis for the modelling work in this project.
The candidate should be a graduate in chemical engineering with a strong interest in applied mathematics and computing, or a graduate in mathematics, physics, computer science or another cognate area who is willing to learn those aspects of chemical engineering necessary to develop the models for the project. The candidate will need to work with and be able to discuss ideas and problems with colleagues who are applied mathematicians.
Work will include the following areas, in which proficiency will need to be developed and prior experience would be an advantage:
Salary is on the AR1B scale (14,317 to 15,986). The post is available for up to 3 years, commencing as soon as possible. The successful candidate will be eligible for membership of the Universities Superannuation Scheme (USS).
The project will be carried out jointly in:
The successful candidate will work as a member of the project team consisting of a Research Associate in each department in addition to the supervisors.
The aim of the project is to develop methods which will allow the efficient solution of very large nonlinear programming (NLP) problems, with 100,000 or more variables, resulting from modelling chemical and similar processes. Such methods are expected to have application in the real-time optimisation of entire processes or process sites, as well as individual units or subsystems of a process.
Models of an integrated process plant are typically composed of sub-models for unit operations (columns, reactors, etc.) connected by streams (material and energy flows). It is often possible to break a large process down into smaller sub-processes by assuming values for a relatively small number of streams. This structure suggests the use of decomposition methods in which the sub-processes are solved by standard techniques such as NLP (nonlinear programming). The overall process is then optimised by a master problem which determines the best values of the linking variables. This is a nonsmooth problem which requires special techniques.
Decomposition introduces various theoretical and organisational issues. One of the main thrusts of the research programme will be to address these issues, leading on to the development of effective algorithms and software.
Particular areas of application will include:
The candidate should be a graduate in mathematics, physics, computer science or related dicipline or a gruadate in chemical engineeing with a strong interest in applied mathematics and computing.
Work will include the following areas, in which proficiency will need to be developed and prior experience would be an advantage:
The appointment is expected to be on the RA1B scale, which has a salary in the range 14,317 to 15,986, though post doctoral candidates with suitable experience will also be considered. The post is available for up to 3 years. The successful candidate will be eligible for membership of the Universities Superannuation Scheme (USS).
See our home pages for further information:
http://www.chemeng.ed.ac.uk/ecosse/ and http://www.maths.ed.ac.uk/