A knowledge-based expert system for gearing design application using Prolog and C

This paper describes the development of an expert system in gearing design application. The knowledge in this design field is vast and involves selection of appropriate data from large amounts of engineering standards data available in engineering catalogues and design handbooks. Experts very oftens have different opinions in determining the design solution. The theme of the work is to organise, systemise, conserve and spread this knowledge relating to gearing design and configuration. There is a large number of design parameters involved in gearing design. The design of gears also requires an iterative approach to optimize these design parameters that govern the kinematic, as well as the strength performance. Owing to the complex combinations of these aspects, conventional design office practice tends to become complicated and time consuming. Hence, in this piece of work, attempts are made to formulate a design system capable of designing a required type of gear and its configuration under a specified set of working conditions. The main features of the expert gearing design system are the use of semantic nets in representing the objects, a frame-based data acquisition and representation system, and production rules in representing the domain-specific knowledge base. The system provides a very user-friendly interface through the use of an expert system development tool Prolog. While Prolog is used to set up the various databases and the inference engine, the detail gear design calculations are performed by C. It is found that the communication between these two language media is excellent in terms of performance and efficiency.