Developing a diagnostic system through integration of ant colony optimization systems and case-based reasoning

摘要

This study is dedicated to integrating both the clustering method and case-based reasoning (CBR) for developing a diagnostic system in maintenance. The reason for this is that searching similar cases for CBR is time consuming if the case base is fairly large. It is necessary to cluster the cases into some groups, and then perform the search for the most appropriate possible group. A novel approach, the ant colony system clustering algorithm (ASCA), is employed for this purpose. The main advantage of this technique is the reduction in the amount of time used in comparison. A real-life problem for car maintenance has shown evidence of this advantage as well as its precision ability.