How to optimize control strategy and parameters for Distributed Generation (DG) units in off-grid microgrids under resistive feeder impedance condition is great of practical significance. For this purpose, a modified droop controller design method for the DG units based on Adaptive Population-based Extremal Optimization (APEO) is designed. The content of the proposed method is that firstly through introducing the resistive droop control and the sum of phase translation based on the fundamental harmonic droop control. Then the design problem of modified droop controller for the DG units is formulated as a typical constrained optimization problem, and an APEO is designed to solve this formulated problem. Finally, the problem of the coupling between voltage stability and power shearing by DG units is solved; the transient impact of DG units on switching between grid-connected and islanded modes is reduced; and hot swap can be realized without DG changing work modes. At last, the superiority in power control performance of the proposed method is demonstrated by experimental results on a microgrid demonstration which is consist of two 10 kW three-phase inverters. This work is supported by National Natural Science Foundation of China (No. 61972288), Welfare Technology Applied Research Project of Zhejiang Province (No. LGG20F030010), and Natural Science Foundation of Zhejiang Province (No. LZ16E050002).