Stability monitoring of fuzzy systems using trend analysis

Research in stability analysis of fuzzy logic controlled systems have intensified in the last decade without producing an adequate stability theory for the design of such systems. The paper presents an on-line stability indicator that identifies the stability state of the control system by monitoring system response. Unlike most stability analyses, the indicator does not depend on the development of a mathematical process model. A quantitative stability index is computed and stability defined from a set of stability criteria. This combination is based on the principle that system knowledge can be utilised to determine the stability state of a system. A graphical representation is also presented. Viability of this approach is demonstrated by applying the stability indicator to a simulated process with varying dynamics