Learning design rules and concepts for examples-a case study to design an electric power substation design

Abstract

Some principles of machine learning and some links with knowledge base system are described. A domain-independent inductive learning system (ILS) has been developed and implemented. ILS can be attached to any expert system, and will work as a knowledge acquisition module for the expert system. This gives the expert system the ability to update and expand its knowledge base according to the circumstances. ILS is a logic-based, data-driven learning system, focusing on the problem of learning structural descriptions. ILS is tailored to design electrical system components. In the present work, ILS is used for specifying the major components of an electrical substation. The learning system will learn design rules and concepts from positive and negative examples in the form of existing substations. This system will take examples and generate rules and concepts for specifying the major components of an electric substation.<>