Application of an ANN-based Controller for the Control of Pneumatic Conveyed Polypropylene Pellet Flow

P.V.S. Ponnapalli, D. Benchebra, R. Deloughry, and I. Ibrahim

Department of Engineering and Technology, Manchester Metropolitan University, Manchester M15GD, UK, Email: p.ponnapalli@mmu.ac.uk

ABSTRACT

Electrical Capacitance Tomography (ECT) has been used for real-time imaging of material flow without interfering with the behaviour of the process itself. This feature made ECT an attractive tool to image and measure the flow of material (e.g. flow of polypropylene pellets in this application). This paper presents continuing work in applying ECT for closed-loop control.

The nature of pellet flow is highly nonlinear which tends to lead to the formation of dunes, necessitating an increase of air velocity to the maximum to clear the dunes, and hence requires large control energy. If the air velocity can be controlled such that the pellet flow is maintained at a constant rate without the build up of dune level, energy usage associated with such processes can be considerably reduced.

One of the main problems in the control of the pneumatic pellet flow system is the difficulty in building good models of the nonlinear dynamics of the system. In this work, Artificial Neural Networks (ANNs) are used to initially build and validate a forward model of the system and then to develop an inverse model of the plant. This inverse model is implemented as an Inverse Controller to maintain constant pellet flow and to clear dunes as quickly as possible. Results obtained from a laboratory flow rig interfaced to a National Instruments PXI ECT imaging system and controlled using dedicated hardware and software implemented in LabView and Matlab are presented.

Results indicate that the NN-based Inverse Controller can successfully control the flow rig system.

Keywords Process Tomography, ECT, LabView, PXI, Neural Network, Mass Flow