Automated Feeder Routing For Planning Of Urban Distribution Network Based On GA

power supply but also decrease frequency of accidents. Based on spatial database, analgorithm for automated feeder routing is presented to optimize the feeders' tendenceof middle voltage distribution network, adopting genetic algorithm that has strongglobal optimization ability. In other words, supposing that load prediction and powersupply region have been determined, automated feeder routing is to design feeders ofhigh-voltage substation separately, and radiate distribution network is generatedaccording to the power supply region and its load.Two steps are taken to settle the problem for the sake of simplifying analyse.Firstly, the primary radiate distribution plan-map of the area is generatedautomatically through the Dijkstra algorithm. This step is called primary feederrouting. Secondly, based on the radiate network gotten in the first step, geneticalgorithm is adopted combining with certain rules to hunt optimal scheme of loadallocation. The key idea of this step is: Applying GA to search for the possiblesequence of the loads, for an arranged load sequence, we put the loads on to the feederone by one from the first according to the known sequence. When the feeder reachesits maximal charge, we then put rest loads on a new feeder. A load allocation schemewill be obtained when all the loads are finished. Finally the scheme will be evaluatedand the optimum planning result is the one with minimal cost.Automated feeder routing software of medium voltage distribution network isdeveloped according to the algorithm mentioned above combining with GIS. In thesoftware, three-layer architecture and COM technology are employed to synthesizethe advantages of several program languages. Each layer is designed to be reusableand flexible component, and the program is organically composed of components,thus the automated feeder routing is implemented. The results of examples aresatisfactory and the software proved effective.