A simple scheme for separating overlapped peak signals (OPS) is proposed, with the Tsallis distribution as the model for OPS processing and fractional-order differentiation as tool for the signal analysis, the relationships of the maxima and zero-crossing values of the Tsallis peak signal, separately, to the order of differentiation are established, leading to two types of parameter estimators, which are utilized, by adjusting the peak-shape parameter, to calculate the parameters of position, height and width of the Gaussian, Lorentzian and Tsallis distribution. The obtained characteristic parameters can be employed to fit the peak signals to be treated, thereby fulfilling the OPS decomposition.
