Modular adaptive approaches in DES failure diagnosis and control

The increasing complexity of man-made plants, safety requirements and economic constraints all require more autonomous plant operations, quick and efficient failure diagnosis algorithms. The analysis of the properties of large, often heterogeneous, compositional models is difficult due to the size and the complexity of the state space. The computational complexity of the design of supervisory controllers and failure diagnosers also grows very fast with the size of the model. The paper presents a partial solution of these problems by exploiting modular plant and controller concepts of T-time Petri nets and adaptive error recovery approaches to the controllers and supervisors design.