Study Of Two-stage Stochastic Programming Based On SQP Algorithm

Stochastic programming is the basic method to solve the optimization problem in uncertain environment.The two-stage stochastic programming problem is based on random variables to be observed before and after the specific values,and meanwhile the decision variables and decision-making process are divided into two stages of mathematical programming to make decisions.In this paper,we study the sequential quadratic programming(SQP)methods and get the two kinds methods of two-stage stochastic programming problems.In this paper,we study the sequential quadratic programming(SQP)methods to solve the two-stage stochastic programming problems,it is mainly used for solving stochastic nonlinear programming and linear partial information(LPI)stochastic nonlinear programming.On the basis of the equivalent nonlinear extensive model,the algorithm based on active set SQP is presented for stochastic nonlinear programming.The search direction is obtained by the active set method which is used to solve the quadratic programming subproblems,and the step length is obtained by the filter method to avoid the difficulty of choosing traditional penalty factor in solving the SQP subproblems,then two kinds of stochastic nonlinear programming algorithm are given.The convergence of the algorithm is proved under certain conditions,and the numerical results show that the algorithm is effective.O n the basis of the stochastic nonlinear programming algorithm skills,LPI stochastic nonlinear programming algorithm is given,and the convergence is proved under certain condition.This paper presents three kinds new algorithms of solving stochastic nonlinear programming.This algorithms form an algorithm system,and they can be given to solve the train of thought for small and medium-size stochastic nonlinear programming.