Improvements on Noninvasive Blood Glucose Biosensors Using Wavelets for Quick Fault Detection

Noninvasive blood glucose sensors are still under developmentstage considering that they are far from being suitable for use inanartificial pancreas. The latter has three main parts: the bloodglucose sensor, the insulin pump and the controller. However, forthe biosensor analyzed here, some common failures such as signalshifts and unreal picks were found. They must be taken intoaccount, for computing the correct insulin dosage for diabeticpersons. Hence, a fault detection system based on discretewavelets transform (DWT) is applied here. The main idea is, whenthe fault occurs, to do a proper measurement compensation forsending the corrected value to the predictive functionalcontroller (PFC) algorithm. The study is done by reproducing thefault on the blood glucose measurements. They are obtained from amathematical model of the endocrine system of an adult diabeticpatient. This model was approved by the FDA in 2008. Then, thesimulation environment includes faulty blood glucose measurementsand a fault diagnosis and identification (FDI) system based onDWT. The FDI system gives to the PFC algorithm the correctinformation to turn it into a fault-tolerant controller (FTC). Themain goal is to deliver the correct insulin dosage to the patient