AI for market development
www.snam.itNeural networks for state-of-the-art modeling tools.
In May 2017 we rolled out a groundbreaking project. For the first time in the gas industry and in Europe we wanted to leverage big data and advanced analytics to model and forecast gas delivery. 5 months later, a neural network based predictive tool was operational.
An integrated team has joined together combining diverse expertise in order to incorporate the largest choices of alternate parameterizations and achieve a more predictive model. In fact, the neural network system adopted is an ensemble of 6 pairs of dynamic models: a multi-model forecasting method which significantly reduces errors in model output.
A few indicators on what we achieved
Balancing the system is crucial for the market: not just physically but also from a commercial point of view. Network operators’ improvements in information flow towards market players and regulatory bodies support liquidity and transparency. We applied state-of-the-art machine learning techniques to deliver the most reliable and transparent information to all users.
Perfomance improvement
-
-26% Average error
-
-31% Maximum error
-
-32% Errors > 10%
Evolution of Snam’s Forecasting Models
ARIMA
Linear AutoRegressive Model
CASSANDRA
(Computer Aided Sales Series Analysis Nullifying Daily Residuals Autocorrelation)
NEURAL NETWORK
NEW MODEL
NEURAL NETWORK ENSEMBLE
DAFNE
(Dynamic Adjusted Forecasting Neural Ensemble)
Machine Learning
Machine Learning algorithms
and explanatory variables
The model is trained to forecast target variables and evaluate the impact of each explanatory variable
It performs a “feature selection” on the explanatory variables reducing their number
The objective function is determined by an algorithm only on the basis of the data structure
Relations between variables are determined by non deterministic models
INPUT DATA
WEATHER
Multidimensional segmentation of the country according to temperature, altitude, humidity, precipitations, etc.
IN-OUT POINTS
Daily time series of provisional and definitive data
CALENDAR
Including the ability to capture critical non standard dates such as bank and floating holidays and potential long weekends
SCADA DATA
Hourly time series of SCADA data points
DAY-AHEAD MARKET
Electricity sale and purchase transactions: data related to hourly energy blocks that are traded for the next day.
NOMINATION
Entry points of nomination, which indicates the gas delivery of a specified volume during the gas day at a network exit/entry point.
INTRADAY
METEOROLOGICAL DATA
Multi-parametric weather forecasts with multiple updates per day.
HEATING DATA
Schedule management of heating systems in accordance with the distinction in climatic zones currently in force in Italy.
