作者: Alexander Feldman , Gregory Provan , Rui Abreu , Johan de Kleer
DOI: 10.1016/J.IFACOL.2015.09.564
关键词:
摘要: Abstract System models that are used in model-based diagnosis often composed of components drawn from component libraries. In these libraries, there may be multiple systems equations per (component implementations). For example, a modeled as non-linear system (high-fidelity model), linear system, and qualitative (low-fidelity model). Choosing the right model for is difficult task requires search space all possible type combinations. this paper we propose method automates computes optimizes set diagnostic metrics scenarios. Initial experimental results show having some preserves accuracy isolation time while, at same time, improves computational complexity numerical stability.