Empirical Analysis of Control Strategies for DC Microgrid Energy Management in University Campus Applications in Enugu State

University campuses in Enugu State, South East Nigeria, face persistent electricity supply challenges that disrupt academic activities, and the integration of diverse renewable energy sources, battery storage, and controllable loads within campus-scale DC microgrids offers a viable path to energy autonomy. This study presents an empirical analysis of three energy management control strategies for a DC microgrid system at the Enugu State University of Science and Technology campus: droop control, centralized hierarchical control, and a distributed multi-agent system control architecture. A laboratory-scale DC microgrid testbed representing the key campus energy assets is constructed, comprising a photovoltaic emulator, a battery energy storage system, a diesel generator emulator, and programmable DC loads representing campus building demand profiles measured over a four-week field data collection campaign. Each control strategy is implemented on the testbed and evaluated across operating scenarios including PV generation variability, peak demand events, battery low SOC conditions, and simulated communication network failures. Hierarchical centralized control achieves the lowest fuel consumption at 14.3 percent below the droop control baseline, but shows vulnerability to communication network failures. The multi-agent system demonstrates superior resilience to communication faults, maintaining stable operation with only 6.1 percent fuel consumption increase above the centralized optimum. Voltage regulation performance is comparable across all three strategies in normal operation. The study recommends a hybrid hierarchical MAS control architecture for university campus microgrid applications combining the efficiency of centralized optimization with the resilience of distributed control. Keywords: DC microgrid, energy management, droop control, multi-agent system, Enugu State