|LAURICELLA MARCO||Cycle: XXXII |
Section: Systems and Control
Tutor: PIRODDI LUIGI
Advisor: FAGIANO LORENZO MARIO Major Research topic
:Methodologies for the Estimation and Monitoring from Data of Electric SystemsAbstract:
Electricity distribution systems are seeing a constant increase of devices and services able to collect, elaborate and transmit data – typically current and voltage – pertaining to many measurement points and with a relatively fast sampling frequency. The data can be employed to achieve advanced or “smart” functionalities to improve the reliability, safety, and efficiency of the network. Crucial ingredients required to realize these advanced functionalities are methodologies and algorithms, able to exploit the available data for system estimation and monitoring purposes. The main goal of the research is to investigate such methodologies, from theory to real-world demonstration. From the theoretical side, the aim of the reaserch project is to develop identification approaches having guaranteed accuracy, resorting to the Set Membership framework of unknown but bounded uncertaintes. Here, new theoretical results are derived for the case of linear systems, which allow one to estimate the noise bound, the model order and the system decay rate prior to perform the identification of the system model parameters. Moreover, guaranteed error bounds are derived for the identified model, along with guarantees on its asymptotic stability. Finally, the developed identification methods are validated on experimental and real-world data pertaining to electric systems or electricity related devices, such as smart circuit breakers. The developed methodology can then be used for monitoring electric networks and their devices, or for forecasting of quantities that are of interest, e.g. building energy consumption.