Click thumbnail to view full-size, then press your browser Back button to return to this page.

Many advances in industrial diagnostics have resulted from the substantial growth in measurement technology. This growth has been matched by the availability of sophisticated digital signal processing hardware and computer-based analysis software which now contribute to the enhanced reliability of industrial processes.

The reference text "Signal Processing for Industrial Diagnostics" by T.M. Romberg, J.L. Black & T.J. Ledwidge, originally published by John Wiley & Sons Ltd (1996, ISBN 0-471-96166-3) under the Wiley Series in Measurement Science and Technology, is now out of print.

The book plus the edited case studies below provide industrial diagnostics practitioners, graduates and undergraduates with an overview of the relevant advanced digital signal processing (DSP) techniques.

SIGNAL PROCESSING FOR INDUSTRIAL DIAGNOSTICS

The complete reference text "Signal Processing for Industrial Diagnostics". Chapters include: 1) Overview and Principles; 2) Basic Theoretical Relationships; 3) Sampled Data Processes and Signal Conditioning; 4) Bivariate Linear Process Analysis; 5) Analysis of Closed Loop Processes; 6) Parametric Spectral Analysis; 7) Multivariate Process Analysis; 8) Data Analysis Exercises; 9) Industrial Case Studies.

POWER STATION BOILER DYNAMICS

This case study describes how digital signal processing applied to measurements in coal feed, feedwater flow, drum level, steam flow, TSV pressure error and generator power can be used to identify the dynamic behaviour of a 500 MWe drum boiler during full power under closed loop control.

SMELTING FURNACE DYNAMICS

This case study describes how digital signal processing applied to vibration measurements from piezo-electric sensors connected to selected tuyere flanges of an Imperial Smelting Furnace (ISF) can be used to identify the dynamic thermal-hydraulic behaviour of the furnace.

EXTRUSION DIE DYNAMICS

This case study describes how vibration sensors connected to an extrusion die are used to identify its characteristic frequencies from the 'burst phenomena' using maximum entropy spectral analysis (MESA) techniques.

MINERAL PROCESS DYNAMICS

This case study demonstrates the feasibility of utilising modern signal analysis techniques for identifying and quantifying key process relationships from the random fluctuations in normal operating data of plant variables logged on a lead-zinc mineral processing plant.

Dr Tom Romberg FIEAust CPEng