Modeling cognitive network of a physical system using design knowledge base

摘要

In this work, a framework is presented to semi-automate the development of a cognitive network diagram for complex engineering systems. Each node of a cognitive network represents design parameters, environmental factors, and quality characteristics whereas each arc represents causal relationship between two nodes. The framework is proposed to utilize the existing knowledge base, such as, CAD database, expert inputs, and other sources of information to generate the network diagram. Along with these sources of information, the framework also extracts geometric structure, functional relations, and other engineering inputs from CAD system to model the complex physical system into the cognitive network diagram. Semi-automation of the development of a cognitive network diagram will help to reduce the tedious task of manually developing network diagrams. The purpose of developing the cognitive network model is to simulate the behavior of complex physical systems and generate the failure knowledge. The framework is demonstrated by a case study of fuel injection nozzle.