A new method of extrapolating feeder peak load histories toproduce estimates of future feeder loads is described. The method, animprovement on past multiple regression curve fit methods, uses anassumed geometry based on substation locations and a classification byrecent growth rates to group feeders into six classes, each extrapolatedin a slightly different manner. The new method is simple enough to beapplied in situations where computing resources are limited. A series oftests shows that the new method outperforms other distribution loadextrapolation methods, and that for short range (less than five yearsahead) forecasts, it matches the accuracy of simulation forecastingmethods, which require considerably more data and computer resources
展开▼
