Applying KBS technology to power transmission network design

Power transmission network design is a promising application area for artificial intelligence techniques. Recently, several attempts have been made to integrate conventional numerical algorithms with more ambitious decision support systems. The authors present SELF, a knowledge-based system supporting a specific task in power transmission network design known as `load-flow analysis´. SELF is the result of long-term cooperation in this field between CISE and the Electric Research Centre of ENEL (the Italian Electricity Board). It integrates a heuristic search approach to the design task with conventional numerical algorithms within the framework of a `propose-critique-modify´ design methodology. The aim of the system is to help the designer interpret numerical results and take design decisions. SELF has been tested on networks of significant size (up to 420 nodes and 600 connections) and is currently used by ENEL researchers