1: \begin{abstract}

2:     We propose a primal-dual interior-point method \allowbreak (IPM)  with convergence to second-order stationary points (SOSPs) of nonlinear semidefinite optimization problems, abbreviated as NSDPs.

3:     As far as we know, the current algorithms for NSDPs only ensure convergence to first-order stationary points such as  Karush-Kuhn-Tucker points. 

4:     The proposed method generates a sequence approximating SOSPs while minimizing a primal-dual merit function for NSDPs by using

5:     scaled gradient directions and directions of negative curvature.

6:     Under some assumptions, the generated sequence accumulates at an SOSP with a worst-case iteration complexity.

7:     This result is also obtained for a primal IPM with slight modification.

8:     Finally, our numerical experiments show the benefits of using directions of negative curvature in the proposed method.

9:     \keywords{

10:         Nonlinear semidefinite programming

11:         \and Primal–dual interior-point method

12:         \and Negative curvature direction

13:         \and Second-order stationary points

14:     }

15:     \subclass{90C22 \and 90C26 \and 90C51}

16: \end{abstract}

17: