Calculating Restart States for Systems Modeled by Operations Using Supervisory Control Theory

Abstract

This paper presents a supervisory control theory based offline method for calculating restart states in a manufacturing control system. Given these precalculated restart states, an operator can be given correct instructions for how to resynchronize the control system and the manufacturing resources during the online restart process. The proposed method enables restart after unforeseen errors. It is assumed that the control system is modeled by operations and that possible operation sequences emerge through dependencies between the operations. The paper shows how reexecution requirements may be included in the calculation to obtain a correct behavior for the restarted system. In addition, it is shown how to filter out restart states that require less effort for the operator during the online restart, and how to adapt the nominal production to always enable restart in desired restart states.

Keywords

• discrete event systems
• restart
• system recovery