To reduce the computation time in stochastic optimization, one thousand random scenarios adapted from the normal distribution with equal probabilities(0.001), are reduced to fifty scenarios by the forward approach, and the probability of the scenarios occurring (which are no longer equal) is also updated.