Adaptive Control with Disturbance Modelling for BG Regulation in TIDM Patient
Akshaya Kumar PatraGirija Sankar PanigrahiVijaya Laxmi PatraAlok Kumar MishraNarayan NahakBidyadhar Rout
0
Citation
9
Reference
10
Related Paper
Abstract:
In order to control blood glucose levels in TIDM patients, this paper explains the creation of a Teaching Learning Based Optimization-PID (TLBO-PID) controller that delivers appropriate insulin doses through an artificial pancreas (AP). Using the Teaching Learning Based Optimization (TLBO), that adjusts the controller gains to improve the BG control of the proposed patient model. This classic controller with TLBO is intended to increase the performance and toughness of patient's problems with glycemic management which are resulting from nonlinearities in the patient model. The nonlinearity of patient models can be effectively handled by using an AP-based TLBO, which also helps to keep blood sugar levels in the glycemic range (70–120 mg/dL). The accuracy, robustness, stability, noise reduction, and enhanced capacity to handle uncertainties are examined while using the proposed patient model with TLBO-PID. A comparison of data from different control strategies indicates the reasons for the suggested approach's superior control performance.Keywords:
Robustness
Blood sugar
Artificial Pancreas
Artificial Pancreas
Cite
Citations (41)
An artificial pancreas capable of maintaining blood sugar homeostasis within the physiological range is described in this paper. The blood sugar is continuously monitored and then interpreted by a minicomputer which in turn controls and implements the delivery of insulin (or glucose). The entire system is automatic and by giving insulin according to a projected blood sugar level the pattern of insulin administration is similar to the biphasic response of the normal pancreas. Five parameters for control can be selected and altered at will so that any level of normoglycemia can be maintained. Hypoglycemia is not encountered, and none of the patients experienced any side effects during or after the trials. The clinical trials involved a two-day study. On the first day the blood sugar profiles were monitored throughout the day. The patients were given their usual doses of subcutaneous insulin and ate measured meals and snacks. On the second day, they received no subcutaneous insulin; insulin was administered intravenously in accordance with the moment-to-moment requirements of the patients who were given meals the same as those of the previous day. Graphs plotted on a common time scale compare the blood sugar patterns on the two successive days and show the significant improvement in blood sugar homeostasis achieved by this artificial pancreas.
Artificial Pancreas
Blood sugar
Homeostasis
Blood sugar regulation
Cite
Citations (272)
Abstract A closed‐loop glycemic control system using an artificial pancreas has been applied with many clinical benefits in J apan since 1987. To update this system incorporating user‐friendly features, we developed a novel artificial pancreas ( STG ‐55). The purpose of this study was to evaluate STG ‐55 for device usability, performance of blood glucose measurement, glycemic control characteristics in vivo in animal experiments, and evaluate its clinical feasibility. There are several features for usability improvement based on the design concepts, such as compactness, display monitor, batteries, guidance function, and reduction of the preparation time. All animal study data were compared with a clinically available artificial pancreas system in J apan (control device: STG ‐22). We examined correlations of both blood glucose levels between two groups ( STG ‐55 vs. control) using C larke's error grid analysis, and also compared mean glucose infusion rate ( GIR ) during glucose clamp. The results showed strong correlation in blood glucose concentrations (Pearson's product‐moment correlation coefficient: 0.97; n = 1636). C larke's error grid analysis showed that 98.4% of the data fell in Z ones A and B , which represent clinically accurate or benign errors, respectively. The difference in mean GIRs was less than 0.2 mg/kg/min, which was considered not significant. Clinical feasibility study demonstrated sufficient glycemic control maintaining target glucose range between 80 and 110 (mg/ dL ), and between 140 and 160 without any hypoglycemia. In conclusion, STG ‐55 was a clinically acceptable artificial pancreas with improved interface and usability. A closed‐loop glycemic control system with STG ‐55 would be a useful tool for surgical and critical patients in intensive care units, as well as diabetic patients.
Artificial Pancreas
Cite
Citations (29)
Background: The potential clinical benefits of continuous glucose monitoring (CGM) have been recognized for many years, but CGM is used by a small fraction of patients with diabetes. One obstacle to greater use of the technology is the lack of simplified tools for assessing glycemic control from CGM data without complicated visual displays of data. Methods: We developed a simple new metric, the personal glycemic state (PGS), to assess glycemic control solely from continuous glucose monitoring data. PGS is a composite index that assesses four domains of glycemic control: mean glucose, glycemic variability, time in range and frequency and severity of hypoglycemia. The metric was applied to data from six clinical studies for the G4 Platinum continuous glucose monitoring system (Dexcom, San Diego, CA). The PGS was also applied to data from a study of artificial pancreas comparing results from open loop and closed loop in adolescents and in adults. Results: The new metric for glycemic control, PGS, was able to characterize the quality of glycemic control in a wide range of study subjects with various mean glucose, minimal, moderate, and excessive glycemic variability and subjects on open loop versus closed loop control. Conclusion: A new composite metric for the assessment of glycemic control based on CGM data has been defined for use in assessing glycemic control in clinical practice and research settings. The new metric may help rapidly identify problems in glycemic control and may assist with optimizing diabetes therapy during time-constrained physician office visits.
Artificial Pancreas
Diabetes management
Cite
Citations (38)
During the course of experimental work on the effect of insulin upon the blood sugar content in rabbits, the question arose as to whether insulin and adrenalin were antagonists. It is well known that insulin acts by diminishing the blood sugar, while adrenalin increases it. It was, therefore, thought highly probable that the administration of both drugs would result in a neutralization of these effects so that the blood sugar would remain practically normal. A study of the literature gave support to the assumption. Magenti and Biagotti conclude that adrenalin, when given simultaneously with insulin, acts by strongly disturbing the usual insulin effect. These tests were repeated only for the reason that in the first papers on insulin by Banting and Best, and especially in a recent article by McLeod and Orr, special attention was called to the individual differences in rabbits, for sugar test, after the administration of insulin. Insulin was injected in quantities of 0.5 units per 1 kg. body weight. This amount was chosen because our former work, on the differing effects of insulin in different body tissues, was done with like quantities. These quantities seemed sufficient to lower the blood sugar conient of normal rabbits markedlly, without the danger of an overdosage. Thus convulsions and other by-effects, possibly without determinable manifestation, could be avoided.
Blood sugar
Cite
Citations (1)
Abstract Commercial controllers with a proportional–integral–derivative (PID) control algorithm were introduced back in the 1940s. As it has been widely reported elsewhere, 80 years later PID is still the most common control algorithm used in the processes industry. In this article, fundamentals of PID control are outlined. Starting from the elemental constituent control actions that are at the core of the basic control law, additional considerations, functionalities, and implementation facts are also introduced. Afterward, special attention is placed on the considerations regarding PID‐based feedback control loops. Although the central concept of PID control can be gleaned from an examination of how these three terms are blended to form a control signal, the intelligent application of PID control in any given case requires an understanding of the process dynamics at hand as well of the achievable feedback properties.
Feedback Control
Control signal
Derivative (finance)
Cite
Citations (3)
Phase margin
Proportional control
Derivative (finance)
Cite
Citations (53)
Artificial Pancreas
Insulin delivery
Insulin pump
Cite
Citations (2)
Despite the dramatic advancement of process control in recent decades, the proportional-integral-derivative (PID) controller contin- ues to be the most frequently used feedback controller today. PID control mechanism, the ubiquitous avail ability of reliable and cost effective commercial PID modules, and pervasive operator acceptance are among the reasons for the success of PID controllers. An elegant way of enhancing the performance of PID controllers is to use fractional-order controllers where I and D-actions have, in general, non-integer orders. In a PIλDδ controller, besides the proportional, integral and derivative constants, denoted by Kp, Ti and Td respectively, we have two more adjustable parameters: the powers of s in integral and derivative actions, viz. λ and δ respectively. This paper compares the performance of conventional PID and fractional PID controllers used for bio-reactor control.
Cite
Citations (20)