Experimental Modeling

Thanks to the innovative methodology named NOSEM® (NOnlinear SEt membership Modeling), Modelway is able to design and develop mathematical models of complex and non linear systems directly from experimental data. In case of systems where physical laws are unknown or too complex to be modeled, the NOSEM® technology reduces time and costs of design.

Here below you can find some projects examples in which the application of a systematic NOSEM® approach allowed a significant reduction in the model calibration complexity, by having a development time reduction from months to weeks.

ico-automotiveModeling of vehicle with semi-active suspensions


Purpose: Development of algorithms for a mathematical model of vehicle with semi-active suspensions, in order to improve passenger comfort and vehicle handling. Using NOSEM® technology has been developed a procedure for deriving, from experimental data, highly affordable nonlinear models of vehicles with semi-active suspensions. Such models are in form that can be used for the control design technique.

Value: Accuracy of the vertical dynamic model increased of one order magnitude. The use of NOSEM® technology in design and calibration of control permit an important time reduction of development. Control performances are increased by 10-25%.

Milanese, M., Novara, C., and Fortina, A. “Experimental modeling of controlled suspension vehicles from onboard sensors” Vehicle Systems Dynamics(2007), vol. 45, no. 2, pp. 133–148.


icon-mechatronic Modeling of air charging system of diesel engine


Purpose: Development of a nonliner MIMO model of diesel engine with single stage turbo charger. The data used for NOSEM® design are measured from the tests used for evaluating engine dynamics without requiring extensive physical modeling.

Values vs. physical modeling: Simulation error reductions up to 12 times and model building efforts reduced up to 5 times.


icon-mechatronicModeling of high-power laser system


Purpose: Development of a mathematical model for Laser systems with highly nonlinear behaviors. Using NOSEM® technology the model has been identified from I/O measurement on the real system overcoming the high nonlinear system behaviors difficult to model using a classical physical equation approach.

Value: Quite accurate model has been obtained with a reduced design effort.

Others modeling projects where the NOSEM® technology has been applied are:

  • Advanced Driver Assistance System (AUTOMOTIVE).
  • Green naval transportation (ENERGY & ENVIRONMENT).
  • Mandrel vibration suppression system (MECHATRONICS).
  • Methane engine modeling and rpm control (AUTOMOTIVE).
  • Optimization of spaceplane trajectories (ENERGY & ENVIRONMENT).