To realize droop parameter self-adaptation in islanded operation of DC microgrids according to the impedance and local load distribution of each photovoltaic unit’s output line within the microgrid, a self-adaptive droop control strategy based on the sparrow search algorithm (SSA) is proposed. To begin with, an objective function that incorporates the droop parameters and converter output voltage reference is constituted in the form of function variable. The SSA is then used to search and determine in real-time the local minima that also minimize both the power deviation and bus voltage deviation during operation. In other words, the proposed strategy transforms what should otherwise be a trade-off problem between droop parameter and voltage reference into a local extremum search problem using SSA. It thereby fulfils the objective of droop parameter self-adaptation under variations in photovoltaic output line impedance, local load, and solar irradiance. At the same time, the act of dynamically adjusting the converter’s output voltage reference values also reduces the bus voltage deviation and resolves the inherent contradiction between power allocation accuracy and bus voltage deviation. A simulation run using a system model created with PSCAD/EMTDC proves the validity and effectiveness of the proposed control strategy.