The DC distribution network adopting droop control is chosen as the research object

of which the small signal state space expression is established. The possibility of instability in the traditional economic optimization model of droop parameters is analyzed. Consequently

the economic optimization model of droop parameters with the assurance of small-signal stability is put forward

which optimizes the network loss with constraints on the eigenvalues of small-signal state space equations under the prediction scenarios and extreme scenarios. To solve the optimization model

the improved strategies based on nonlinear programming algorithm and genetic algorithm are put forward to directly against the defects in the traditional algorithm of slow

easily getting trapped in local optimum and having an accurate solution because of the error. Numerical results show that this optimization model can improve the performance of the economic operation of the system on the premise of ensuring the small-signal stability and improve the robustness of the system when facing uncertain power fluctuation. The strategies can improve the optimization ability and computational efficiency while the results are ensured to be reliable and effective.