Electric Load Forecasting with Geographic Hierarchy
Duration: 50 minutes
Electric load forecasting, a well-established problem in the utility industry, has received a lot of attention from the forecasting community due to some of the key features of the electricity demand series, such as hourly or sub hourly interval, multiple seasonal patterns and high dependency on explanatory variables. Prior to the massive deployment of smart grid technologies, most scientific efforts on electric load forecasting were devoted to forecasting one series at the corporate level. In today’s world, many utilities are collecting hourly data from millions of smart meters, which introduce geographic hierarchy as an additional dimension to the problem. This presentation will answer a challenging question, how can we take advantage of geographic hierarchy to improve the load forecast?
This talk has been presented at the following event:
SAS Analytics Conference 2012, Las Vegas, NV, Oct 8-9, 2012. download presentation slides (PDF)