We study some properties of a general class of probabilistic normed space and prove approximations theorems for functions with values into such a space. By taking into account that time series and random signals are, in fact, such particular functions, these results give some tools for the analysis of time series and random signals.