A λ-cut approximate algorithm for goal-based bilevel risk management systems

Abstract

Bilevel programming techniques a re developed for decentralised decision problems with decision makers located in two levels. Both upper and lower decision makers, termed as leader and follower, try to optimize their own objectives in solution procedure but are affected by those of the other levels. When a bilevel decision model is built with fuzzy codlicients and the leader and/or follower have goals for their objectives, we call it fuzzy goal bilevel (FGBL) decision problem. This paper first proposes a A-cut set based FGBL model. A programmable A-CUt approximate algorithm is then presented in detail. Based on this algorithm, a FCBL software system is developed to reach solutions for FGBL decision problems. Finally, two examples are given to illustrate the application of the proposed algorithm.