A new approach to gross error detection provides unbiased estimates and 100( 1 -a ) % simultaneous confidence intervals of process variables when biased process measurements and process leaks exist. Presented in this article are estimation equations for process variables, as well as equations that help identify biased measurements and process leaks. These equations include the power function for a global test, and two types of a-level component tests and their power functions. Important strengths and weaknesses of this approach are compared to those of the serial compensation strategy, in particular, by varying the significance level ( a ) , the variance-covariance matrix ( C ) , the size of measurement bias (6), the number of biased variables, and the sample size ( N ) . Accuracy of 6 estimation and performance in detecting the presence of process leaks ( y ) are also evaluated and compared. The proposed approach has unique features that can provide a basis for improving the reconciliation of variables in process operations.

Hypoglycemia is a major challenge of artificial pancreas systems and a source of concern for potential users and parents of young children with Type 1 diabetes (T1D). Early alarms to warn the potential of hypoglycemia are essential and should provide enough time to take action to avoid hypoglycemia. Many alarm systems proposed in the literature are based on interpretation of recent trends in glucose values. In the present study, subject-specific recursive linear time series models are introduced as a better alternative to capture glucose variations and predict future blood glucose concentrations. These models are then used in hypoglycemia early alarm systems that notify patients to take action to prevent hypoglycemia before it happens. The models developed and the hypoglycemia alarm system are tested retrospectively using T1D subject data. A Savitzky-Golay filter and a Kalman filter are used to reduce noise in patient data. The hypoglycemia alarm algorithm is developed by using predictions of future glucose concentrations from recursive models. The modeling algorithm enables the dynamic adaptation of models to inter-/intra-subject variation and glycemic disturbances and provides satisfactory glucose concentration prediction with relatively small error. The alarm systems demonstrate good performance in prediction of hypoglycemia and ultimately in prevention of its occurrence.

No abstract

Many patients with diabetes experience high variability in glucose concentrations that includes prolonged hyperglycemia or hypoglycemia. Models predicting a subject’s future glucose concentrations can be used for preventing such conditions by providing early alarms. This paper presents a time-series model that captures dynamical changes in the glucose metabolism. Adaptive system identification is proposed to estimate model parameters which enable the adaptation of the model to inter-/intra-subject variation and glycemic disturbances. It consists of online parameter identification using the weighted recursive least squares method and a change detection strategy that monitors variation in model parameters. Univariate models developed from a subject’s continuous glucose measurements are compared to multivariate models that are enhanced with continuous metabolic, physical activity and lifestyle information from a multi-sensor body monitor. A real life application for the proposed algorithm is demonstrated on early (30 min in advance) hypoglycemia detection.

This paper extends the method introduced by Rollins et al. (ISA Trans. 1998, 36, 293) to multiple-input, multiple-output systems that give an exact closed-form solution to continuous-time Hammerstein processes written in terms of differential equations and nonlinear inputs. This ability is demonstrated on a theoretical nonlinear Hammerstein process of complex dynamics where perfect identification of the closed-form model is assumed. This paper then demonstrates the simplicity of the proposed identification procedure to obtain an accurate estimate of the exact model using a theoretical Hammerstein model. A powerful attribute of this methodology is the ability to make full use of the statistical design of experiments for optimal data collection and accurate parameter estimation. Application of the proposed method is demonstrated on a household clothes dryer with four input and five output variables. Only 27 trials (input changes) of a central composite design were needed for accurate model development of all five outputs over the input space, and the accurate predictive performance is demonstrated. This paper extends the method introduced by Rollins et al. (ISA Trans. 1998, 36, 293) to multipleinput, multiple-output systems that give an exact closed-form solution to continuous-time Hammerstein processes written in terms of differential equations and nonlinear inputs. This ability is demonstrated on a theoretical nonlinear Hammerstein process of complex dynamics where perfect identification of the closed-form model is assumed. This paper then demonstrates the simplicity of the proposed identification procedure to obtain an accurate estimate of the exact model using a theoretical Hammerstein model. A powerful attribute of this methodology is the ability to make full use of the statistical design of experiments for optimal data collection and accurate parameter estimation. Application of the proposed method is demonstrated on a household clothes dryer with four input and five output variables. Only 27 trials (input changes) of a central composite design were needed for accurate model development of all five outputs over the input space, and the accurate predictive performance is demonstrated.