Search

Soft sensors have been widely used in chemical plants to estimate process variables that are difficult to measure online. One of the crucial difficulties of soft sensors is that predictive accuracy drops due to changes in state of chemical plants. Characteristics of adaptive soft sensor models such as moving window models, just-in-time models and time difference models were previously discussed. The predictive accuracy of any traditional models decreases when sudden changes in processes occur. Therefore, a new soft sensor method based on online support vector regression (SVR) and the time variable was developed for constructing soft sensor models adaptive to rapid changes of relationships among process variables. A nonlinear SVR model with the time variable is updated with the most recent data. The proposed method was applied to simulation data and real industrial data, and achieved higher predictive accuracy than traditional ones even when time-varying changes in process characteristics happen. © 2013 American Institute of Chemical Engineers AIChE J 60: 600-612, 2014 [ABSTRACT FROM AUTHOR]

Recent laboratory investigations into the bioaccumulation and toxicity of polycyclic aromatic hydrocarbons (PAH) have focused on low-level, time-variable exposures to early life-stage fish. Polycyclic aromatic hydrocarbon body-burden residues reported in these studies were lower than critical body-burden residues predicted by the target lipid model (TLM). To understand this discrepancy, a time-variable uptake and depuration model of PAH bioaccumulation was developed. Kinetic constants were fit using measured exposure and tissue concentrations. The resulting lipid-water partition coefficients (KLW) were uncorrelated with the octanol-water partition coefficient (KOW)-a qualitatively unrealistic finding considering that numerous studies have reported a positive correlation between the two. Because PAHs are known to be metabolized, the comparison of KLW with KOW suggests that metabolism may be occurring in early life-stage fish. Therefore, the uptake and depuration model was modified to include metabolism while assuming linearity of KLW with KOW. Calculated metabolism rates were positively correlated with KOW-a finding qualitatively similar to those of other studies. The present study provides a reasonable explanation for the discrepancy between the TLM predictions and the measured toxic effect levels. Given the time-variable exposure concentrations, the maximum measured body burdens used to relate to toxic effects may be underestimated. In addition, the maximum body burden of parent PAH plus metabolites may be a better measure in relating tissue concentrations to toxic effects. Incorporating these refinements in relating body burdens to toxic effects may result in a better comparison between TLM predictions and measured effect levels. [ABSTRACT FROM AUTHOR]

Bryant, Hartley & Jessen (1960) presented a two-way stratification sampling design when the sample size is less than the number of strata. Their design was extended to a three-way stratification case by Chaudhary & Kumar (1988) , but this design does not take into account serial correlation, which might be present as a result of the presence of a time variable. In this paper, a new sampling procedure is presented for three-way stratification when one of the stratifying variables is time. The purpose of such a design is to take into account serial correlation. The variance of the unweighted estimator of the population mean with respect to a super population model is used as the basis for comparison. Simulation results show that the suggested design is more efficient than the Chaudhary & Kumar (1988) design. [ABSTRACT FROM AUTHOR]