Birandom variables and birandom programming

摘要

The emphasis of this paper is to introduce a novel concept of birandom variable and to exhibit the framework of birandom programming. The so-called birandom variable is a measurable mapping from a probability space to a collection of random variables. Based on this definition, the expected value operator of birandom variable and chance measures of birandom event are further introduced. As a generalized scenario of stochastic programming, a spectrum of birandom programming models are developed to deal with birandom systems. To solve the proposed models, birandom simulations are presented and then a hybrid intelligent algorithm is designed by embedding neural networks into genetic algorithm. Finally, some numerical experiments are provided to illustrate the effectiveness of the algorithm.