Data-based modeling (system identification) is the scientific exercise consisting of determining a mathematical model of a system using input-output data collected on that system. Identifying a model of the system can be done using well-established techniques when the data are collected in open loop or in a classical closed-loop setting.
However, we observe that the systems are nowadays becoming increasingly complex and interconnected. These interconnections are much more complex than a classical closed loop. Consequently, the classical closed-loop techniques cannot be directly applied to identify such interconnected systems.
In this project, the objective is to extend the closed-loop identification techniques to be able
- to determine the interconnection structure in a complex network
- and to identify the dynamics between the different signals in such a network.
The identified models of such complex interconnected systems could be then used to control those networks using distributed control. |