Our algorithm is an ensemble model in which a genetic algorithm optimizes the calibration parameters of a Physical Informed Graph Neural Network, which consists of both fluid dynamics and deep learning models, to achieve the best accuracy of results and predictive capability.
SERVICES
We support the Utilities starting from traditional activities, integrating and refining them with artificial intelligence.
We redefine the value of data: from a generator of insights to a generator of decisions.
DESIGN OF MEASUREMENT CAMPAIGNS AND DISTRICTUALIZATION
HYDRAULIC MODELLING
DATA QUALITY ANALYSIS, DATA CLEANSING AND MEASURES RECONSTRUCTION
Redundancy analysis and measurement points virtualization
PREDICTION OF CHEMICAL AND PHYSICAL PARAMETERS A FEW DAYS IN ADVANCE
ENSEMBLE MODELS WITH THE EXISTING FLUID-DYNAMICS MODELS