An Inexact Copula-Based Stochastic Fractional Programming Model for Planning Emergency Evacuation at Nuclear Power Plant Sites

Abstract

In this study, an inexact copula-based stochastic fractional model (ICSFP) approach is developed for supporting emergency evacuation management in response to nuclear power plant accidents. Based on an integration interval mathematical programming (IPP), fractional programming (FP) and joint chance-constraint programming (JCCP), ICSFP can systematically reflect various complexities in emergency evacuation systems such as multiple uncertainties and controversial targets. Specifically, the copula approach is introduced into ICSFP to reflect nonlinear dependence among random variables and further characterize the interaction among violations of single factors on the desired evacuation schemes and the associated risk levels. To demonstrate the effectiveness of the developed approach, ICSFP is applied to support emergency evacuation management subject to nuclear power plant accidents. The results indicate that the interactions among the traffic flows would pose apparent impacts on the desired evacuation plans, especially for the demand conditions (i.e., lower bound of unit cost). Comparison among the developed ICSFP model and traditional evacuation models (interval evacuation models and interval fractional evacuation model) suggests that ICSFP is advantageous in balancing conflicting objectives and reflecting interactions among dependent random variables to make trade-offs among system risks and unit costs.

Keywords: emergency evacuation management, copula, optimization, uncertainty, interdependence