No abstract

A normalized performance index, η(b), is introduced to characterize the performance of feedback control schemes. η(b) is attractive because it provides a measure of the proximity of control to minimum variance control, which is the optimal feedback control provided that the process can be described by a linear transfer function with additive disturbance. Both time domain and spectral interpretations of this performance index are discussed. A fast, simple, on‐line method for estimating η(b) is given, along with some of the statistical properties of the estimator. Simulation and industrial examples demonstrate the utility of η(b/)

An analysis of variance for univariate feedforwardlfeedback systems has been derived which allows the variance contributions due to both disturbances and controllers to be established. A performance index which assesses the performance of the overall control scheme has been developed. Estimates of the analysis of variance and performance index can be obtained using ordinary closed loop data and least squares regression. These quantities can be used on-line or off-line for either the design of new controllers or as a means of assessing the performance of existing feedforwardlfeedback control schemes. The practicality of the results in the paper are demonstrated using simulated and industrial data.On a Ctabli une analyse de variance de systbmes actionlrktroaction univariCs qui permet l'ktablissement des contributions de la variance dues aux perturbations et aux contrbleurs. Un indice de performance Cvaluant la performance du schtma de contr6le global a Ct C calculk. I1 est possible d'obtenir des estimations de l'analyse de la variance et de I'indice de performance a l'aide des donntes en boucle fermCe habituelles et de regression des moindres carrCs. Ces quantitks peuvent Ctre utilisCes en ligne ou de manikre indbpendante soit pour la conception de nouveaux contr6leurs, soit comme moyen d'evaluer la performance des schkmas de contrble post-actionirCtroaction existants. L'applicabilitk de ces rCsultats est demontrCe 3 I'aide de donnkes simulCes et industrielles.Keywords: control performance assessment, feedforwardlfeedback control, analysis of variance, multi-input single-output.here are many applications where multivariate control

Aim:To describe the clinical profile, maternal and fetal outcomes, and the conversion rates to diabetes in women with gestational diabetes mellitus (GDM) seen at a tertiary care diabetes center in urban south India.Materials and Methods:Clinical case records of 898 women with GDM seen between 1991 and 2011 were extracted from the Diabetes Electronic Medical Records (DEMR) of a tertiary care diabetes center in Chennai, south India and their clinical profile was analyzed. Follow-up data of 174 GDM women was available. To determine the conversion rates to diabetes, oral glucose tolerance test (OGTT) was done in these women. Glucose tolerance status postpartum was classified based on World Health Organization (WHO) 2006 criteria.Results:The mean maternal age of the women was 29 ± 4 years and mean age of gestation at first visit were 24 ± 8.4 weeks. Seventy percent of the women had a family history of diabetes. Seventy-eight percent of the women delivered full-term babies and 65% underwent a cesarean section. The average weight gain during pregnancy was 10.0 ± 4.2 kg. Macrosomia was present in 17.9% of the babies, hypoglycemia in 10.4%, congenital anomalies in 4.3%, and the neonatal mortality rate was 1.9%. Mean follow-up duration of the 174 women of whom outcome data was available was 4.5 years. Out of the 174, 101 women who were followed-up developed diabetes, of whom half developed diabetes within 5 years and over 90%, within 10 years of the delivery.Conclusions:Progression to type 2 diabetes mellitus (T2DM) in Indian women with GDM is rapid. There is an urgent need to develop standardized protocols for GDM care in India that can improve the maternal and fetal outcomes and help prevent future diabetes in women with GDM.