The result of the Sensitivity Analysis for model 'Modelica.Fluid.Examples.AST_BatchPlant.Test.TwoTanks' is that all values of the sensitivity indices are NaN. We found out the cause, which is the calculation of the variance in each analyze function (e.g. Sobol).
The values of the output variables 'pipe.flowModel.states[1].p' and 'pipe.flowModel.states[2].p', which are exponentially increasing, should be different until time about 44 seconds. Instead, y (the list of the samples) contains the same values, which should not be the case. Only the inputs, here the values of x, are changing but just for every second run. (insight through debugging)
The assumption now is that other equations may be important, such as where a density is calculated and depends on the input pipe.flowModel.states[1].T .
The result of the Sensitivity Analysis for model 'Modelica.Fluid.Examples.AST_BatchPlant.Test.TwoTanks' is that all values of the sensitivity indices are NaN. We found out the cause, which is the calculation of the variance in each analyze function (e.g. Sobol).
The values of the output variables 'pipe.flowModel.states[1].p' and 'pipe.flowModel.states[2].p', which are exponentially increasing, should be different until time about 44 seconds. Instead, y (the list of the samples) contains the same values, which should not be the case. Only the inputs, here the values of x, are changing but just for every second run. (insight through debugging)
The assumption now is that other equations may be important, such as where a density is calculated and depends on the input pipe.flowModel.states[1].T .