A High Performance Online Data Processing EIT System

Y Ma1, N Holliday1, Y Dai1, M Wang1, R A Williams1 and G Lucas2

1Institute of Particle Science and Engineering, Leeds University, Leeds LS2 9JT, UK, m.wang@leeds.ac.uk

2School of Engineering and Computing, University of Huddersfield, Huddersfield HD1 3DH

ABSTRACT

This paper presents the development of a new Electrical Impedance Tomography (EIT) system for online data acquisition and processing of measurements from two-phase flows, whose axial velocities can be up to 10ms-1, with ±5% precision. To be able to measure two-phase flows at these velocities requires data acquisition speeds (and processing) of 1000 frames/s per plane in order to generate the flow velocity maps. The system has been carefully designed in a modular fashion and can consist of several data acquisition modules and computing modules. The former incorporates a voltage controlled current source with a DC-restore circuit, an equal-width pulse synthesizer unit to produce synthetic waveforms for electric field excitation and demodulation signals (for synchronization and sampling) and a synchronized digital demodulation unit with sixteen parallel measurement channels. The computing module incorporates a powerful Digital Signal Processor with an IEEE1394 interface for communication. These modules form a data processing pipeline for a series of tasks ranging from flow measurement to image reconstruction to determining flow velocities. The DSP system can implement three image reconstruction methods, these being the SCG method, the single-step SCG Method and the Back-Projection method. An online-updating cross-correlation algorithm has been developed and is to be implemented on the system. This particularly suited to flow velocity calculations, as this technique is less time and memory consuming compared to direct cross- correlation techniques. The system architecture, special design techniques, and data processing algorithms will be introduced in the paper along with the current progress in the system’s development. Potential applications will be also be discussed.

Keywords EIT system architecture, Multiphase flow, Multi-DSP