Applied analysis of differences by cross-correlation functions

Abstract

The testing of the statistically significant similarities between two different data series cannot be measured using by Wilcoxon or T paired test, because there is not exists pretest and posttest. For the alternative way of testing the similarities, when data has some boundness, the proposal of using cross-correlation tests is presented in this paper. When necessary, the condition is existing some same characteristics, which can be seen as the connection of two power system nodes in the power transmission system. The cross-correlation analysis can be suitably complemented using by computing the p values between the correlation data themselves, which in addition, the trend of correlation data is measured for expression of the dynamic of cross-correlation achieved data. Application of the proposed method is realized on the real data set of the behavior of power transmission systems in the Moravia-Silesian region.