DYNAMIC PRIORITIES AS A BASIS FOR CONTROL IN NONSEQUENTIAL SYSTEMS

Most programming languages or modeling systems use either sequential c ontrol or, in the other extreme, complete concurrency. We propose here a control structure consisting essentially of dynamic priority relati ons between elementary steps. Elementary steps may be executed if they are ''enabled'' (executable) and if at that time no other elementary step with a higher priority is enabled. This leaves much room for conc urrency. Priorities can be changed after executing a step. Such a cont rol structure can be imposed on Petri nets; this has been investigated more closely for the modeling language MoMo (i.e., for its task layer ), which is based on colored Petri nets. The proposed task layer langu age is presented here in some detail. It has been tried out with sever al medium-size examples selected to find out to which limits this cont rol principle is feasible. Few priorities have been needed and rarely dynamic ones. Dynamic priorities can in principle be modeled in higher Petri nets, but that easily leads to complicated structures. The sepa rate control with priorities is much simpler and more flexible.