|
Stochastic programming is of particular importance in mathematical programmingpart and relatively new discipline. With the wide range of mathematical programming,indepth application.We are familiar with the deterministic programming model which can’tsolve practical problems, Therefore, the stochastic programming model is generated.Now,in particular in optimal control, operations research, management science and otherdisciplines, this model more and more demonstrated its practicality.This article focused on the programming function approximation function and theerror in the original problem. The deterministic programming related error problems havebeen discussed in many of references.This article used some conclusions of deter-ministicplanning and further issues of random error.In the analysis of the discussion, Firstly,in the second chapter of Probability andMeasure.The definition of several common models and theorems, and gived detailed proof.Experience approximation is applied widely,which mainly discusses two types ofproblems named wait and see and here and now.The model of wait and see means we canwait for the realization of the variable and make the decision according to all informationof the realized value.However the model of here and now is different,we can’t wait for therealization of the variable,and we need to decide before the realization of the variable.Secondly, the core of this article, I discussed the approximation for stochasticprogramming problems with the original function of the boundary .This would make useasily judge the merits of the chosen approximation function, as much as possible toimprove the model’s degree of approximation.We all know, Stochastic programming has important applications in many fields.However,in the actual, distribution of parameters in the model is not clear, Therefore, weused the methods of approximation to solve problems and then discussed error.
|