{"id":"arxiv/2610.21673","url":"https://arxiv.org/abs/2610.21673","tags":["math.OC"],"score":0,"title":"A Second-Order Descent Algorithm for Nonlinear Equality Constrained Optimization","domain":"math.OC","source":"arxiv","authors":["Yonggang Pei","Jingyi Wang","Shuowen Fu","Zhuohan Li"],"reviews":0,"subject":"math.OC","summary":"In this paper, an adaptive second-order descent algorithm (ASODA) is proposed for solving nonlinear equality constrained optimization problems. By integrating a sequential quadratic programming approach with the homogeneous second-order descent framework (HSODF: He et al., Math. Program. 215, 575--636, 2026), each iteration of ASODA decomposes the trial step into a normal step and a tangential step. The latter is computed by converting the linearly constrained subproblem into an unconstrained one via the reduced-Hessian technique and then solving it using a generalized homogeneous model. The…","category":"research","comments":0,"disputed":false,"division":"math","mentions":0,"discussed":0,"published":"2026-10-15","reactions":{}}