Discussing A Simple Method Of Semidefinite Programming

Semidefinite programming is widely used in statistics,structure design and combina-torial optimization lots of fields.Some common problems such as convex quadratic pro-gramming and second-order cone optimization are often related to finding the solution of semidefinite programming problems. This kinds of problems are difficult to solve,especially large-scale semidefinite programming problems. After the inner point method was put for-ward in the eightys, various kinds of algorithm for semidefinite programming problems had sprung up. We are based on the existing algorithm to find a effective algorithm to solve one kind of semidefinite programming problems. This thesis is organized as follows:In the first chapter, semidefinite programming problems are introduced briefly. It also minutely limns the preliminary results and the evolution process of methods for solving semidefinite programming problems. In addition, the key points of the thesis are introduced briefly.The second chapter embarks from the existing algorithm, studies a simple method of large-scale conic semidefinite programming problems, and proves the convergence of the proposed algorithm in detail.The third one briefly expounds the details of Matlab programming, and through numerical results demonstrates the effectiveness of the proposed algorithm.