Tutorial 7: Real data identification of an actual radiator-room system aimed to virtual sensor design

The tutorial presents both theoretical issues and practical hints on data analysis and processing, in order to identify, by means of techniques implemented by Matlab/Scilab tools, linear and affine mathematical models, which can be used to design a Virtual Sensor. Actual acquired thermal and hydraulic data are used to work out the proposed identification problem and to test the algorithms. The plant, a radiator-room system, is described and its inputs (commands/disturbances) and outputs (measured/estimated) are defined to state the thermal model identification problem. Different model typologies, such as transfer matrix or state space equations and linear or affine models, are described. Some practical suggestions on data preprocessing, as detrend and scaling operations, are also pointed out. To conclude, the obtained models are compared to evaluate their suitability in describing the plant behaviour. Eventually, some notes on Virtual Sensor design are given.