To identify potential risks to the system security presented by time series it is offered to use wavelet analysis, the indicator of time-and-frequency distribution, the correlation analysis of wavelet-spectra for receiving rather complete range of data about the process studied. The indicator of time-and-frequency localization of time series was proposed allowing to estimate the speed of non-stationary changing. The complex approach is proposed to use the wavelet analysis, the time-and-frequency distribution of time series and the wavelet spectra correlation analysis; this approach contributes to obtaining complete information on the studied phenomenon both in numerical terms, and in the form of visualization for identifying and predicting potential system security threats.