Skip to main content

Self learning model for intelligent predictive control system for crystallization processes

Objective

Due to unpredictable crystallisation mechanisms, industrial crystallisation (up to 70% of all chemicals) is seldom operated under automatic control of product size distribution. This frequently results in non-reproducibility, unacceptable product qualities and excess energy consumption. Objective of SINC-PRO is to increase efficiency (reproducibility) and effectiveness (-20% energy, -5% cost, reduced time-to-market) by developing advanced techniques for on-line measurement and control. Results: flexible process modelling tool, observer/feedback system integrated with on-line measurement techniques, toolbox comprising of both Model Predictive Control and self-learning neural network type control. Exploitation: modelling and control tool box (two software developers), integrated modelling and control system with new measurement techniques (5 end-user industries), improved insight in crystallisation mechanisms and control (2 RTD institutes).

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

DSM RESEARCH BV
Address
Koestraat 3
6160 MD Geleen
Netherlands

Participants (11)

DANISCO SUGAR OY
Finland
Address
Sokeritethaantie 20
02460 Kantvik
DELFT UNIVERSITY OF TECHNOLOGY
Netherlands
Address
Leeghwaterstraat 44
2628 CA Delft
INNOVA SPA
Italy
Address
Via Della Scrofa 117
00186 Roma
INTELLIGENT SYSTEM MODELING AND CONTROL N.V.
Belgium
Address
Technologielaan 11 / 0101
3001 Leuven
IPCOS TECHNOLOGY BV
Netherlands
Address
Bosscheweg 145 A
5282 WV Boxtel
KEMIRA OYJ
Finland
Address
Porkkalankatu 3
00101 Helsinki
LABOR S.R.L.
Italy
Address
Via Della Scrofa 117
Roma
PROCESS SYSTEMS ENTERPRISE LTD.
United Kingdom
Address
Blackfriars Lane 20
EC4V 6HD London
PURAC BIOCHEM BV
Netherlands
Address
Arkelsedijk 46
4200 AA Gorinchem
ROQUETTE ITALIA SPA
Italy
Address
Via Serravalle 26
15063 Cassano Spinola
UNIVERSITY OF ROME "LA SAPIENZA"
Italy
Address
Via Eudossiana 8
00184 Roma