Motor vehicles have been identified as a growing contributor to air pollution, such that analyzing the traffic policies on energy and environment systems (EES) has become a main concern for governments. This study developed a dual robust stochastic fuzzy optimization - energy and environmental systems (DRSFO-EES) model for sustainable planning EES, while considering the traffic sector through integrating two-stage stochastic programming, robust two-stage stochastic optimization, fuzzy possibilistic programming, and robust fuzzy possibilistic programming methods into a framework, which can be used to effectively tackle fuzzy and stochastic uncertainties as well as their combinations, capture the associated risks from fuzzy and stochastic uncertainties, and thoroughly analyze the trade-offs between system costs and reliability. The proposed model can: (i) generate robust optimized solutions for energy allocation, coking processing, oil refining, heat processing, electricity generation, electricity power expansion, electricity importation, energy production, as well as emission mitigation under multiple uncertainties; (ii) explore the impacts of different vehicle policies on vehicular emission mitigation; (iii) identify the study of regional atmospheric pollution contributions of different energy activities. The proposed DRSFO-EES model was applied to the EES of the Beijing-Tianjin-Hebei (BTH) region in China. Results generated from the proposed model disclose that: (i) limitation of the number of light-duty passenger vehicles and heavy-duty trucks can effectively reduce vehicular emissions; (ii) an electric cars’ policy is enhanced by increasing the ratio of its power generated from renewable sources; and (iii) the air-pollutant emissions in the BTH region are expected to peak around 2030, because the energy mix of the study region would be transformed from one dominated by coal to one with a cleaner pattern. The DRSFO-EES model can not only provide scientific support for the sustainable managing of EES by cost-effective ways, but also analyze the desired policies for mitigating pollutant emissions impacts with a risk adverse attitude under multiple uncertainties.
Two-Stage Bayesian Model Averaging in Endogenous Variable Models
