An enhanced-interval linear programming (EILP) model and its solution algorithm have been developed that incorporate enhanced-interval uncertainty (e.g., A(+/-), B(+/-) and C(+/-)) in a linear optimization framework. As a new extension of linear programming, the EILP model has the following advantages. Its solution space is absolutely feasible compared to that of interval linear programming (ILP), which helps to achieve insight into the expected-value-oriented trade-off between system benefits and risks of constraint violations. The degree of uncertainty of its enhanced-interval objective function (EIOF) would be lower than that of ILP model when the solution space is absolutely feasible, and the EIOF's expected value could be used as a criterion for generating the appropriate alternatives, which help decision-makers obtain non-extreme decisions. Moreover, because it can be decomposed into two submodels. EILP's computational requirement is lower than that of stochastic and fuzzy LP models. The results of a numeric example further indicated the feasibility and effectiveness of EILP model. In addition, El nonlinear programming models, hybrid stochastic or fuzzy EILP models as well as risk-based trade-off analysis for El uncertainty within decision process can be further developed to improve its applicability. (C) 2008 Elsevier B.V. All rights reserved.
An insulator MnO as an electron injecting and transporting material introduced into organic light-emitting diodes to increase electroluminescence efficiency also can eliminate the problem of the oxidation of reactive dopants to improve stability of devices. (C) 2008 Optical Society of America
The effects of the accuracy of major-point source emissions input data on the predictions of a regional air-quality model (AURAMS) were investigated through a series of scenario simulations. The model domain and time period were chosen to correspond to that of PrAIRie2005, an air-quality field study with airborne and ground-based mobile measurement platforms that took place between August 12th and September 7th, 2005, over the city of Edmonton, Alberta, Canada. The emissions data from standard sources for three coal-fired power-plants located west (typically upwind) of the city were compared to the continuous emissions monitoring system (CEMS) taking place at the time of the study - the latter showed that the original emissions inventory data considerably overestimated NOx, SO2, and primary particulate emissions during the study period. Further field investigation (stack sampling) in the fall of 2006 showed that the measured primary particle size distribution and chemical speciation for the emissions were strikingly different from the distribution and speciation originally used in the model. The measured emissions were used to scale existing emissions data in accord with the CEMS and in-stack measurements. The effects of these improvements to the emissions data were examined sequentially in nested AURAMS simulations (finest horizontal resolution 3 km), and were compared to airborne aerosol mass spectrometer (Aerodyne AMS) measurements of particle sulphate, and particle distributions from an airborne passive cavity aerosol spectrometer probe (PCASP). The emissions of SO2 had the greatest impact on predicted PM, sulphate, while the primary particle size distribution and chemical speciation had a smaller role. The revised emissions data greatly improved the comparisons between observations and model values, though over-predictions of fine-mode sulphate still occur near the power-plants, with the use of the revised emissions data. The modified emissions also had a significant impact on the larger particles of the particulate matter, with more primary PM in sizes greater than 1 mu m diameter than had previously been estimated, and higher large particle concentrations close to the power-plants. Crown Copyright (C) 2008 Published by Elsevier Ltd. All rights reserved.