Applying NOSEM® technology, it is possible to predict the behavior of complex and highly evolving systems from experimental measurements or temporal series of data.
Some example of application in environmental and energy sectors are described here below.
Purpose: Development of a model for river flow prediction is a fundamental problem in hydrology to inform decision making process by competent authorities to prevent and manage associated risks. NOSEM® technology was applied to the prediction of Dora Baltea river, for design the model we used the daily data gotten by the river. The prediction results were compared with other methods (neural networks and just in time predictors) normally used in literature for such applications.
Value: Predictions obtained by NOSEM® provided good performances and good robustness versus imprecise knowledge of involved nonlinearities and noise properties. Moreover the applied method compared extremely well with respect to neural networks and just in time predictors.
Purpose: Development of a model for prediction of atmospheric pollution is important for supporting local authorities in pollution control and prevention. NOSEM® technology has been applied to the prediction of the daily concentration peak of OZONE and PM10. Prediction has been performed on experimental time series measured in the city of Brescia. The NOSEM® technology has been compared with other standard methods (neural networks, fuzzy, ARCX).
Value: The NOSEM® technology performed very well in comparison with neural networks, fuzzy, and ARX methods and was adopted by the customer to support decision making process related to pollution management.
Others prediction projects where the NOSEM® technology has been applied are:
- Advanced Driver Assistance System – ADAS (AUTOMOTIVE).
- Electrical energy demand prediction (ENERGY & ENVIRONMENT).
- Prediction of dams crest displacement (ENERGY & ENVIRONMENT).
- Prediction and modeling of sea waves motion (ENERGY & ENVIRONMENT).