Register or Log In

International Society for Industrial Process Tomography

1st World Congress on Industrial Process Tomography

Control of Pneumatic Conveying Using ECT


D. Neuffer1, A. Alvarez1, D.H. Owens1, K.L. Ostrowski2, S.P. Luke2 & R.A. Williams2


1Centre for Systems and Control Engineering, School of Engineering, University of Exeter, North Park Road, Exeter, Devon, EX4 4QF, UK,

email : D.Neuffer@ex.ac.uk

2Camborne School Of Mines, University of Exeter, Redruth, Cornwall, TR15 3SE,UK, email : S.Luke@csm.ex.ac.uk


Abstract - The control of dense-phase pneumatic conveying systems is notoriously difficult. Specifically, achieving sufficiently low air velocity to ensure efficient power utilisation, low product degradation and plant wear, whilst ensuring that blockage of the pipeline does not occur, is the greatest challenge. ECT could potentially offer a cheap, durable and potentially effective method of visualising solids movement within conveying pipelines and ultimately act as a primary sensory input to an opti- mal control system.


A pneumatic conveying rig has been designed and constructed in order to simulate the slug flow of plastic pellets when using air as a transportation media. Twelve electrode ECT sensors have been placed on a test section of the rig to measure variations of slug flow and to further investigate any potential for control. A pre- requisite paper [1] details the successful time-based ?animation? of solids flow through the pipeline, the analysis of these flow patterns and the potential of this work for the control of pneumatic conveyors. This ?bolt-on? paper will describe two parallel control strategies, which may be employed in association with an ECT system for the optimisation of pneumatic conveyors. One technique will utilise parametric modelling of conveying systems, while the other will be based on a fuzzy logic approach, where the membership functions will be optimised by the use of genetic algorithms.


Keywords: ECT, pneumatic conveying, control, parametric modelling, fuzzy logic.


Sign-in to access the full text