Inexact two-stage stochastic partial programming: application to water resources management under uncertainty

摘要

In this study, an inexact two-stage stochastic partial programming (ITSPP) method is developed for tackling uncertainties presented as intervals and partial probability distributions. A scenario-based interactive algorithm is proposed to solve the ITSPP model. This algorithm is implemented through: (ⅰ) obtaining extreme points of the linear partial information (LPI); (ⅱ) generating an inexact two-stage stochastic programming (ITSP) model under each extreme point; (ⅲ) solving ITSP models through interactive algorithm proposed by Huang and Loucks (Civil Eng Environ Syst 17:95-118, 2000); (ⅳ) acquiring the interval solutions under each extreme point and the final optimal interval for the objective function. The developed method is applied to a case study for water-resources planning. The modelling results can generate a series of decision alternatives under various system conditions, and thus help decision makers identify the desired water-resources management policies under uncertainty.