In this research proposal new methodology for supervisory control timed discrete event systems will be elaborated. More specifically, decentralized and modular control of large distributed timed systems will be studied in order to reduce the computational complexity of supervisory control synthesis, which constitutes the major obstacle for application of supervisory control in industry. Weighted automata (e.g. (max,+) automata, interval automata, and other classes of timed automata) model flexible manufacturing systems, computer and information networks with communication protocols, where not only ordering but also timing of discrete events is important. Both cases of local control specifications and more relevant global (indecomposable) specifications will be investigated. Similarly, we will consider the case, where all events are observable as well as the case, where not all events are observable due to the very nature of an event or simply due to economic reasons (too high a cost of a sensor ). Our goal is to find effective methods for supervisory control of this type of systems, in particular decentralized control and coordination control, where local controllers communicate with a coordinator.
The objective of DISC is the design of supervisors and fault detectors exploiting the concurrency and the modularity of the plant model. Coordinated controllers should preferably be designed using only local plant behaviour models, and requiring only limited information exchange between the different local controllers.
We plan to use several techniques to reduce the computational complexity of solving the above mentioned problem for distributed plants: modularity in the modelling and control design phases; decentralized control with communicating controllers;
modular state identification, distributed diagnosis and modular fault detection based on the design of partially decentralized observers;
fluidisation of some discrete event dynamics to reduce state space cardinality.
The expected outcome of this project are: new methodologies for applying the above described techniques for embedded controllers to distributed plants;
new tools for the modelling, simulation and supervisory control design that will be part of an integrated software platform; the application of these methodologies to a few cases of industrial relevance using the developed tools; the dissemination of the results.
This project aims at studying logical and timed discrete event systems using methods from universal coalgebra and idempotent algebra with special focus on supervisory control of large distributed systems. Methods of idempotent algebra enable linear representation on suitable idempotent semirings useful for quantitative (timing) aspect of control, while coalgebra is useful for qualitative aspects of control. A combination of these techniques will be applied to the decentralized and modular supervisory control in order to reduce the computational complexity and make our results applicable to control of large distributed systems.
grants
