| 摘要 |
<p>A health monitoring (HM) system for complex networked apparatus includes a number of neuro-fuzzy inference (NFI) apparatuses (102) feeding inference results (104) into a data fusion hierarchy (106). At each level in the hierarchy, fuzzy inference is applied to generate a desired output signal by processing selected input signals in accordance with a knowledge base (126) defining fuzzy membership functions and fuzzy inference rules defined in advance. The knowledge base comprises alternative definitions (A, B, C) of membership functions and/or inference rules. The apparatus selects which definition to use according to environmental or other conditions, and predetermined selection criteria. Neuro-fuzzy inference can thus be optimised according to subsets of the full range of operating conditions, leading to a simplification of the real-time inference computation.</p> |
