A mixed-integer two-stage interval stochastic programming model for regional air quality management

This paper proposes a hybrid mixed-integer two-stage interval stochastic (MITIS) programming Model for regional air quality management planning. This proposed model is a combination of the existing two-stage stochastic programming, mixed integer programming and interval programming methods. The parameters in the MITIS model can be expressed as probability density functions (PDFs) and discrete interval parameters. The model is designed to take both integer and discrete parameters. The objective of this research is to minimize the cost of treating SO2-emission, while maintaining the standard emission allowance. The application of the MITIS model to regional air quality management is novel. The model can also deal with penalty costs for the excess SO2-emission, and obtain optimal decisions of capacity expansion for new treatment measure. The method has been applied to a case study of SO2-emission treatment strategy and facility-capacity expansion for regional air quality management system. The results indicate that reasonable solutions for binary and continuous variables have been obtained. Therefore this model can be used for planning regional air quality management. The results cover all of the cost aspects, and the decision for applying a new treatment measure for each sources.