The application of expert systems to fault diagnosis in alternators

摘要

This Paper describes the development, of an expert system demonstrator for fault diagnosis in alternator seal oil, hydrogen cooling and stator-winding water auxiliary systems. It is clear from wide experience that improved plant diagnostic capability can be achieved by the application and use of better condition monitoring tools, which offers the potential benefits of enhanced alternator reliability and lifetime management. Another important potential benefit is the ability of the operator to make the correct diagnosis during operational incidents when under extreme pressure. Existing monitoring systems are not generally extensive enough and do not contain the necessary intelligence to offer these benefits. In particular they do not have the ability to interpret plant sensor information, perform the necessary online analysis or trending and provide the operator with the necessary clear and easily understand guidance. Advice on alternator problems has traditionally been obtained from plant specialists, but their expertise is now becoming very scarce. Expert system technology is seen as one vehicle whereby such experience can be captured and structured into a knowledge base suitable for plant operators. This Paper describes the development of an expert system designed to meet the above functional requirements and concerns. The Paper contains an outline of the principal duties of the alternator support systems, discusses the approach to knowledge acquisition, modelling and analysis, and describes the expert system design. Examples of expert system diagnosis of typical alternator system problems are presented.