Numerical simulation with physical models enables the computing of quantities of interest from input parameters and variables. These data are often uncertain : sensors are never 100% reliable, and the properties of two products from the same product line can differ due to various tolerances. Then, how is it possible to trust our models?

Persalys is an open source GUI allowing to account for uncertainties in simulation. It is jointly developed by EDF, the French national electric utility company, and Phimeca. In this tutorial, we will work together through a case study with a model of a solar collector from the

ThermoSysPro library, also developped by EDF.

First, we will determine the relative influence of the inputs on the model outputs. In our case the inputs are the sun radiation, angle of the mirror, and outside temperature. The outputs are the heating power delivered by the collector and the temperature of the water that is heated. Then we will perform the central tendancy analysis using the Monte Carlo method. Finally, we will calculate a failure probability, which in our case is the probability that the water temperature is too low for the heating to be efficient.

Requirements

Please download the necessary tutorial files here

Password: Phi_Modelica_Conf_15