This module is useful when exploring statistical behavior of a variable which is not measured but computed from measured variables or from variables with a known statistical behavior. Using the module, range of the resulting variable, its mean and confidence interval can be estimated. The individual measured variables contributions to the resulting variable variability can be determined as well. The approach is sometimes called uncertainity evaluation or error propagation. Resulting variable properties are evaluated by the Monte Carlo simulation and by the second order Taylor expansion.

Simulation - Pdf manual

AP

3.488

3.328

3.298

3.493

3.353

....

Error Propagation, Uncertainty
Task name :	Sheet1
Function :	(X1-X2)*ln(Y1)
No. of simulations :	1000
Input variables :
	Mean	Std. deviation	95% interval
X1 :	40	4	32.16	47.84
X2 :	50	10	30.4.2007	69.6
Input data :
	Mean	Std. deviation	95% interval
Y1 :	4.851617133	1.406598457	2.094684157	7.608550109
Output value :
Median :	-14.71772521
Mean	Std. deviation	95% interval		+-3sigma
-15.38768532	17.20265039	-49.10488009	18.32950945	-66.9956365
Interval of generated values
-51.98832859	17.66283193
Sensitivity analysis
Absolute sensitivity :
x1:	1.711697033
x2:	1.711697033
y1:	1.805592704
Relative sensitivity :
x1:	26.8396
x2:	67.099
y1:	15.92717428
Taylor expansion approximation
Simple mean :	-15.79312079
Corrected mean :	-15.37284183
Corrected mean (w/covar.) :	-15.37284183
Corrected std. deviation :	17.25502425
Corr. std. dev. (w/covar.) :	17.25502425
95% interval :	-49.19268936	18.44700569
Interval +-3sigma :	-67.13791457	36.3922309