Inexact probabilistic optimization model and its application to flood diversion planning in a dynamic and uncertain environment

Flood management systems involve a variety of complexities, such as multiple uncertainties and their interdependences, as well as multiregion and dynamic features. This paper thus presents an inexact two-stage mixed-integer programming with random coefficients (ITMP-RC) model for flood management in a dynamic and uncertain environment. ITMP-RC is capable of addressing dual uncertainties expressed as random boundary intervals that exist in the coefficients of the objective function. A case study of flood diversion planning is used to demonstrate the applicability of the proposed methodology. Results indicate that total system costs would be rising gradually with increased probabilities of occurrence, implying a trade-off between economic objective and system safety. A variety of decision alternatives can be obtained under different policy scenarios, which are useful for decision makers to formulate appropriate flood management policies according to practical situations. The performance of ITMP-RC is analyzed and compared with an inexact two-stage stochastic programming model.