Schiele

testo blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto blablatesto

The soft sensor model-driven determines the estimate of the output variable by a known computational model. That is, it is known how the input variables must be combined to get the output. More in detail, the model used in this type of soft sensor can be defined by the physical laws related to the variables treated by the applying, in the Aristotelian way, the “first principles”. Or else, by adopting a model borrowed from other disciplines. This is the case, for example, when in robotics a bio-inspired model is used to measure quantities that otherwise could not be measured.
This type of soft sensor is characterized by the fact that it is not necessary to have historical series of input-output pairs of the system. Furthermore, a system training period is not required.