ATLASg
- Home
- ATLASg
ATLASg constitutes an intelligent system for short-term gas load forecasting.
The data used by ATLASg are collected from internal and external data sources and processed. The data comprise:
External data providers:
- Gas TSO website: the historical gas consumption per exit point shall be used for the training and the forecasting process of the Artificial Neural Networks.
- Power TSO website: Integrated Scheduling Process (ISP) results, mainly the commitment and scheduling (electricity production) of gas-fired generating units, the reserve awards, the units in AGC mode (providing secondary reserve), and the system/zonal imbalance.
- External data provider(s): wind production and PV production forecasts for days D+2 until D+7; this data are received either through a web service or through the FTP protocol.
Forecasting utilizes:
a) Artificial Neural Networks for gas load forecasting of industrial nodes and Distribution Network nodes
b) ISP and RTBEM simulation for generating units’ nodes
ATLASg has the capability to automatically execute the day-ahead and intra-day gas consumption forecasts on hourly basis (24 x 7 Days), without human intervention. The day-ahead and intra-day gas consumption forecasts are updated automatically every hour according to the new available data from the external sources. The weekly forecasts shall run once per day (every afternoon at about 16:00 – 18:00), based on the data available at that time.
