EBSILON®Professional Online Documentation
This method is already implemented in EBSILONProfessional since the DOS-version and helps in determining the most probable process state when redundant measurement points are present.
This method has the following advantages:
The disadvantages are
However, in this method it is possible to account for the different measurement inaccuracies of the individual measurement points through empirical weighting. These are standardized internally to the mean value of 1. Weights of temperatures are transferred to the related values of specific enthalpy without conversion i.e. the entered weights are not valid for the temperatures, but instead for the values of specific enthalpy resulting from these.
The weighting is included quadratic in the calculation i.e. the contribution of a measurement point with a weight of 2 is as high as the contribution of 4 measurement points having a weight of 1. It must be mentioned here explicitly that this method does not follow the guideline VDI 2048.
This simplified method is enabled by an internal advance-simulation. Thereafter, all the variables can be referred to the results of the simulation, so that all variables take up value by 1. It is now assumed that they show the same standard deviation, if their recording accuracy remains the same. Therefore, only the different relative measurement inaccuracy needs to be noted. The weight values for the individual measurement values must be in the correct ratio with one another.
Correlations between the measurement values cannot be considered here, because here the empirical weight matrix is filled only on the diagonals (namely, with the squares of the standardized weights).