Abstract--In this paper we propose a new modeling approach.The proposed approach is based on hybrid model which combines Interval Constrained Petri Nets (ICPN) and Timed Automata. These tools allow us to evaluate, respectively, the quality variations and to manage the flow type disturbance . An example analysis illustrates our approach.Index Term--modeling, Petri Nets, Timed Automata, Interval Constraint Petri Nets, Robustness.
I.INTRODUCTION The aim of this paper is to explain the use of hybrid tool which co mbines Interval Constrained Petri Nets (ICPN) and timed Automata. ICPN is a sub-class of High Level Petri Nets with Abstract Marking (AM-HLPN) [2] [3]. ICPN model allows us to model and guarantee a constraint on any parameter of a manufacturing process. In our case it has been used to model the flu x of material flo w for regulating the quality of tobacco. Timed automata is a tool for the modelling and verification of real t ime systems [13] [14]. A timed automata is essentially a fin ite automata (FSM), extended with real-valued variables. Such an automata may be considered as an abstract model of t imed systems. This expressive modelling tool offers the possibilities of model analysis like verification, controller synthesis and also faults detection and isolation to model manufacturing systems whose activity times are included between a minimu m and a maximu m value. We use it for modelling the discrete parts of system co mmand. Both tools are applied to a robustness control for regulation systems and for description of dynamical system. In this paper, first, the process is presented. The following section describes ICPN and gives some basic notions on the timed automaton which is used in the modeling step. Then we build the ICPN mode wh ich describes the material flow. This tool presents a complement to the P-temporal Petri nets [1]. So we prove the robust control laws of this model in order to use production data information of manufacturing production systems. Then we use the timed automata model to describe the control lo w. Finally, when the global model of the process is completely defined, we present an application to tobacco manufacturing to illustrate our approach. Finally, a conclusion is presented with some perspectives.