Visual Designing Knowledge Bases for Aircraft Diagnostics Using State Transition Diagrams

The development of knowledge bases for classical intelligent systems based on the formalism of logical rules remains an urgent task. One of the ways to improve the efficiency of this task is through visual programming in combination with the use of widely used graphical notations, in particular state transition diagrams. The paper considers the application of these principles in the creation of software components (a virtual assistant knowledge base) to support the diagnostics and maintenance of aircraft. The main steps of the approach are described: analysis of the domain and the allocation of basic concepts and relationships; construction of state transition diagrams using the results of the analysis; creation of elements of logical rules based on the analysis of the diagrams; and debugging and testing of the obtained knowledge base structures. An illustrative example with an indication of the recommended software is given. Based on the results of the approach, it was concluded that state transition diagrams are preferable in describing cyclic processes, as well as that they can be useful for verifying decision tables.