Two-level distributed fully-predictive frequency control scheme for inverter-based AC Microgrid considering communication delay

Abstract

Amid the current development of decentralized renewable energy sources, the research of multi-agent-based control for the next generation of the intelligent power system has become one of the most trending and worth-noticing topics in the smart grid research community. Serving as a contribution to the research orientation, this paper proposes a novel two-layer distributed fully-predictive frequency control scheme for an inverter-based three-phase three-wire AC microgrid. The key feature of the suggested methodology is the incorporation of objective functions and dynamic models as constraints in the predictive formulation of both primary and secondary layers in the microgrid control hierarchy. In our study, the controller is designed based on constrained cost function minimization using Quadratic Programming (QP) and the utilization of local measurements and information from surrounding distributed generators to achieve the desired control action. Along with the integration of virtual impedance, second-order generalized integrator phase-lock loop, and harmonic suppression constraint functions, the proposed control scheme shows great response against all variations of load changes, communication delay, communication network failure, and Plug-n-Play capability test. The performance of the proposed controller is verified and evaluated on the 4-DGs microgrid model using MATLAB/Simulink environment.