Fault detection (FD) and diagnosis in industrial processes is essential to ensure process safety and maintain product quality. Partial least squares (PLS) has been used successfully in process monitoring because it can effectively deal with highly correlated process variables. However, the conventional PLS-based detection metrics, such as the Hotelling's T2 and the Q statistics are ill suited to detect small faults because they only use information from the most recent observations. Other univariate statistical monitoring methods, such as the exponentially weighted moving average (EWMA) control scheme, has shown better abilities to detect small faults. However, EWMA can only be used to monitor single variables. Therefore, the main objective of this paper is to combine the advantages of the univariate EWMA and PLS methods to enhance their performances and widen their applicability in practice. The performance of the proposed PLS-based EWMA FD method was compared with that of the conventional PLS FD method through two simulated examples, one using synthetic data and the other using simulated distillation column data. The simulation results clearly show the effectiveness of the proposed method over the conventional PLS, especially in the presence of faults with small magnitudes. 相似文献
Background: Two of the 3 standardized field sobriety tests that U.S. law enforcement uses at roadside checks have a postural equilibrium component to them. Those tests have been validated to detect impairment caused by blood alcohol concentrations (BACs) of 0.08 g/dL or above. Many medical and traffic safety associations support a lower limit, and one state, Utah, has passed a law to lower the limit to 0.05 g/dL. Many studies have examined the effects of alcohol on postural control (of which postural equilibrium is a component), with a consensus emerging that impairment is usually found at BACs greater than 0.06 g/dL. Most of these studies, however, had a relatively small number of subjects, usually between 10 and 30. The current study collected data from a much larger sample.
Objective: The objective of this study was to provide additional evidence that posture control is negatively affected at BACs greater than 0.06 g/dL or breath alcohol concentrations (BrACs) of 0.06 g/210 L.
Method: This was a between-subjects study, with BrAC group as the independent variable (5 levels: 0.00, 0.04, 0.06, 0.08, and 0.10 g/210 L); 4 measures of postural control as the dependent variables; and age, height, and weight as the covariates. Posture control was measured with a force-sensing platform connected to a computer. The feet's center of pressure (CoP) on the platform was recorded and the corresponding movement of the body in the anterior–posterior and lateral planes was derived. Participants (N = 96) were randomly assigned to one of the BrAC groups. Positive BrAC groups were compared to the zero BrAC group. Data were examined with hierarchical multiple regression.
Results: Adjusted for age, height, and weight, the main effect of lateral CoP with eyes open was not statistically significant. There was a statistically significant main effect of alcohol on anterior–posterior CoP excursion with eyes open and with eyes closed and lateral CoP excursion with eyes closed. For all 3 of those variables, only BrACs of 0.08 and 0.10 g/210 L produced differences against zero BrAC. Although the main effect of alcohol on Lateral CoP Excursion with eyes open was not statistically significant, the contrasts between 0 and 0.08 and 0 and 0.10 g/210L BrAC were in the hypothesized direction.
Conclusion: The current study did not directly address the issue of whether the sobriety tests are sensitive to BrACs of 0.05 g/210 L or above; rather, it provides additional evidence that postural control, one of the components of those tests, is relatively unaffected by BrACs lower than 0.08 g/210 L. Additional research is needed on the diagnostic characteristics of the sobriety tests at BrACs lower than 0.08 g/210 L. 相似文献