INTELLIGENT PRIMARY ROUTER FOR UNDERGROUND RESIDENTIAL DISTRIBUTION DESIGN

This paper presents an intelligent tool for optimizing primary cable r outes in an Underground Residential Distribution system (URD). Althoug h initially developed as one of the modules in the Puget Sound Power & Light Co'.s Automated Electrical Plat Design project, this tool, call ed Automated Primary Router (APR), can also be used on a standalone ba sis. APR, implemented in a Geographic Information System (GIS)-based e nvironment, is capable of efficiently accessing and manipulating spati ally-referenced data from a Facilities Management System database. APR also provides a full-fledged Graphic User Interface, along with on-li ne visualization and accurate cost estimation to facilitate the design process. APR employs a heuristic search algorithm, taking advantage o f the geographical relationship in a residential area, to find the bes t primary cable routes either in new residential developments or in ex isting URD systems, as a part of the ongoing cable replacement program . Based on the test results, APR shows significant stability and effic iency in finding the optimal solution for primary cable routing. With this performance, APR can help distribution engineers improve the qual ity of URD design, thus producing standardized and economically justif iable primary cable routes. Because geographically referenced design d ata are digitally stored in the GIS database. APR further increases th e reusability and accessibility of the URD design information. This pa per also provides some prelimiary results of recent research that comb ined a Genetic Algorithms (GA) based site selection module with APR to fully automate the primary system design for URD systems.