PSO based stochastic programming model for risk management in Virtual Enterprise

Fuqiang Lu, Min Huang, Xingwei Wang

Abstract: Risk management in a Virtual Enterprise (VE) is an important issue due to its agility and diversity of its members and its distributed characteristics. In this paper, a stochastic programming model of risk management is proposed. More specifically, we consider about the stochastic characters of the risk in VE, and then we build a stochastic programming model to deal with the stochastic characters of the risk. In detail, this is a chance constraint programming model, One of the great advantages of this class of model is that it can exactlly describe the risk preference of the manager. In this model, the risk level of VE is obtained from a composite result of many risk factors. In order to reduce the risk level of VE, the manager has to select effective action for every risk factor. For each risk factor, there are several actions provided. Here we only select one action for a risk factor or do nothing with it. To solve this stochastic programming model, A particle swarm optimization (PSO) algorithm is designed. On the other hand, to deal with those stochastic variables, Monte Carlo simulation is combined with PSO algorithm. Finally, a numerical example is given to illustrate the effectiveness of the PSO algorithm and the result shows that the model is very useful for risk management in VE.