Modelling of non-conventional sampled data systems

Digital control of continuous time processes is usually designed on the basis of a constant sampling rate. In some industrial environments it is difficult to preserve this simplification, so it is necessary to assume nonconventional sampling patterns. The problem could be much more complex if there were various subprocesses with different performances either in time or frequency domains. A new way of modelling these kind of situations is presented. The approach deals with input increments and exactly matches the plant behaviour in an ideal case. The parameters of this description present a specific physical meaning