A novel strategy for dynamic identification in AC/DC microgrids based on ARX and Petri Nets

Area of Science:

• Electrical Engineering
• Control Systems
• Power Systems

Background:

• AC/DC microgrids (MGs) are increasingly complex, exhibiting non-linear dynamics.
• Accurate dynamic modeling is crucial for stable operation and control of MGs.
• Existing identification methods may struggle with the multi-mode nature of MGs.

Purpose of the Study:

• To develop a novel hybrid strategy for dynamic identification of AC/DC microgrids.
• To obtain a dynamic model applicable to both isolated and connected MG modes.
• To address the challenges posed by the non-linear characteristics of MGs.

Main Methods:

• A hybrid strategy combining Auto-Regressive with exogenous inputs (ARX) and Petri Nets (PN).
• Decomposition of the non-linear system into a bank of linearized models at stable operating points.
• Coordination of linearized models using a Petri Net state machine for dynamic representation.

Main Results:

• Successfully obtained a dynamic model for DC microgrids in isolated and connected modes.
• The hybrid approach effectively captures non-linear dynamic properties of AC/DC MGs.
• Validated the performance of the proposed algorithm via Matlab/Simulink simulations.

Conclusions:

• The proposed hybrid strategy provides an effective method for dynamic AC/DC microgrid identification.
• The use of ARX and PN offers a robust solution for modeling complex, non-linear microgrid systems.
• This approach enhances the understanding and control capabilities of microgrids in diverse operational scenarios.