On-Line Condition Monitoring Technology and Applications

Condition monitoring is applicable in a wide range of industrial sectors.

By monitoring the condition of plant and machinery, maintenance can be scheduled based on its condition rather than on a time-based regime.

Production downtime and plant failure are minimized, maintenance costs are reduced, and the life of expensive components and plant are maximized.

Despite this the uptake of the technology in some areas has been slow.

This report reviews existing off-line and on-line condition-monitoring technology and its application in a number of key industrial sectors.

The relative merits and limitations of various technologies and techniques are presented, and the results of a survey are used to examine the current extent of applications.

The main condition-monitoring techniques, the plant that can be monitored, the types of equipment currently available and approximate equipment costs are summarized in a table.

The report then discusses the potential for more advanced on-line condition monitoring using recent developments in the field of neural networks. These new technologies offer substantial benefits over current condition-monitoring equipment, and could change the way in which condition monitoring is used in the future.

The development by ERA of a generic approach to condition-monitoring based on neural-network technology is presented, followed by a practical evaluation of the technology applied to monitoring and industrial process plant.

The results show how the neural-network-based approach can be used to give early warning of rapidly developing impending failures that current technology would miss.

The trial system uses multiple-parameter information, and makes use of the neural network's ability to learn and adapt solutions to complex data-analysis problems.

The advanced technology presented in this report should, therefore, find application in a number of diverse industrial sectors, including paper, food, metals, building services, electricity supply and most other process industries.

The nature of the technology allows easy transfer of the techniques to different plant and reduces the amount of effort required to interpret condition-monitoring information.