Skip to main content

CASE-BASED LEARNING ENVIRONMENT FOR PLANT PROCESS MANAGEMENT

Objective


The CLEAN project developed a system for automatic on-line process optimisation in pulp and paper plants. Optimisation is here defined as producing pulp or paper to a given specification consisting of permitted ranges for several measurable quality and environmental parameters, while minimising the cost per tonne of output subject to these constraints. This was achieved by implementing an adaptive controller using 'Case-Based Learning' techniques, originally developed under research into artificial intelligence, supported by neural nets modelling some of the parameters. Quantified results are presented for one pulp mill and one paper mill. The technique is applicable to other industries where the variable quality of the raw materials causes control problems.
The goal of clean is to create a system for automatic on-line high-level parameter setting in pulp and paper plants.

Automated systems used at present perform the classic automation functions such as measuring, open-loop and closed-loop control. They also perform some of the online functions which optimise the individual plant operations. However, the high-level operations of optimising the set-points and control parameters and checking the model and process constants, assumptions and optimisation criteria are only performed infrequently, offline, manually, and generally by a highly qualified technolgist.

The present project aims at developing an automated online approach based on recent advances in case-based learning methods. The system will make the best use of existing literature and historical data, of applicable analytical approaches, of heuristic rules and deep AI, in assessing the overall performance of the pland and in resetting its high-level parameters. At the same time the system will not be static and circumscribed by the rules and values input: it will have in itself the capability to learn from experience and improve its policies.

It is expected that the utilization of the system will lead to cuts in energy, pollution and the use of auxiliary chemicals, a rise in production yield and turnover, and an increase in profit margins

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

BRITISH MARITIME TECHNOLOGY LTD
Address
Waldegrave Road, 1 Orlando House
TW11 8LZ Teddington
United Kingdom

Participants (4)

COMPANHIA DE CELULOSE DO CAIMA SA
Portugal
Address
Rua Joaquim Antonio Aguira41.3
1000 Lisboa
KUL
Belgium
Address
De Croylaan 46
3001 Heverlee
ROERMOND PAPIER BV
Netherlands
Address
Po Box 1225 Mijnheerkensweg 18
6040 KE Roermond
Siemens AG
Germany
Address
Werner-von-siemens-straße 50
91052 Erlangen