Distributed Continuous Time-Varying Optimization for Microgrids with Heterogeneous Renewable Energy Systems

Microgrid control systems are implemented using a hierarchical framework comprising primary, secondary and tertiary control levels. This paper proposes a distributed, time varying optimization method for the secondary control level of heterogeneous energy sources (photovoltaics and batteries) in a microgrid. The method addresses time-varying objectives and constraints while achieving frequency and voltage regulation and optimal power sharing. Its main advantages are: (i) provision of continuous-time optimal control inputs that adapt to fast changes in generation and load; (ii) seamless interaction with grid forming and grid following converters; (iii) adaptive real/reactive power sharing and battery state-of-charge balancing based on net power and state of charge; and (iv) implementation over a sparse neighbor-to-neighbor communication graph. Real-time validation on a modified IEEE 13- and 37- test feeder using RTDS with a server-side solver, hardware-in-the-loop, confirms feasibility. Compared with established distributed secondary control schemes, the proposed controller reduces frequency and voltage root-mean-square regulation errors, lowers the maximum observed voltage deviation under load or renewable disturbances, and achieves a lower cumulative operating cost while maintaining all battery state-of-charge trajectories balanced. Scalability and heterogeneity are further demonstrated on an islanded IEEE 37 bus case with mixed grid forming and grid following resources.