ABSTRACT
As far as stochastic variable in stochastic programming problem is concerned, because of different manage goal and technology demand, method of adoption is also different, however, the most natural method is that average value of stochastic variable corresponding function is used. In expected value constraint, model which makes expected value of goal function acquire optimum is named Stochastic Expected Value Models (SEVM). The main method of solving SEVM is that stochastic simulation (Monte-Carlo simulation) is combined with intelligent algorithm [5-8], in which GA is excellent so far. However, genetic operation process such as selection, crossover, mutation are not only complex but also slower convergence and worse precision. At present researcher of various countries are going on researching new, more effective algorithm for solving this problem [5-8].
