A NEW METHODOLOGY FOR SENSOR SELECTION IN PROCESS MONITORING OF RECONFIGURABLE MACHINING SYSTEMS (RmSs)

摘要

This article outlines a new approach for selecting an appropriate monitoring system for any given machining system. Presented are the concepts of fault, sensor, signal processing and computing spaces and their generic characteristics. Study of their fundamental properties reveals that there are certain levels of uncertainty involved in describing them. This has led us to the adoption of fuzzy systems theory as a fundamental tool for development of the "Mapping Theory". It is shown that based on the characteristics of the faults and the type of machining operations, it is possible to rank the sensors based on their characteristics and to come up with an appropriate multi-sensor monitoring system. Examples are provided to show the effectiveness of the approach.