Approximate compilation for embedded model-based reasoning

摘要: The use of embedded technology has become widespread. Many complex engineered systems comprise features to perform self-diagnosis or self-reconfiguration. These require fast response times in order be useful domains where are typically deployed. Researchers often advocate the compilation-based approaches store set environments (resp. solutions) a diagnosis reconfiguration) problem, some compact representation. However, size compiled representation may exponential treewidth problem. In this paper we propose novel method for compiling most preferred reduce large space requirements our We show that approximate compilation is an effective means generating highest-valued environments, while obtaining whose can tailored any application. also provides graceful way tradeoff with completeness coverage environment space.