Monte-Carlo parameter uncertainty analysis under dynamical and operational measurement conditions

For controlling, observing and optimizing engineering processes one needs often dedicated experiments. Unfortunately no measurement is exact such that deriving conclusions from a measurement campaign requires some caution. Hence, in order to control or optimize a certain parameter of interest, uncertainty of the parameter needs to be the measurement quantified. In the literature two methods are proposed to perform this task: analysis of the noise propagation or Bootstrap Monte-Carlo (BMC) methods. The first one is inaccessible for the layman user. The BMC is difficult to perform if noise sources are mutually correlated since all correlations need to be taken into account. We present a new direct measurement for parameter uncertainty which can be operated under correlated noise sources without the need of explicit knowledge or description of the correlation at hand.