Introducing the effect of aging into the Honey Bee Mating Optimization to determine parameters for blood glucose level calculation

Diabetes mellitus is one of the most frequent civilization diseases. It is characterized with elevated blood glucose level. Elevated blood glucose level damages multiple organs with an increased risk of death, mainly due to hypo- or hyperglycemic shock. Eventually, diabetic patient becomes dependent on insulin to maintain normal blood glucose level. Over-dosage of insulin can lead to death in a short period. Therefore, patient and his physician must be aware of patient's blood glucose level. Patient does sporadic measuring of blood glucose level using a finger-stick. From time to time, the patient additionally wears a continuous glucose monitoring system. Such a system measures glucose level in subcutaneous tissue - it does not measure blood glucose level. As these two glucose levels are not necessarily the same, a model of glucose dynamics is needed to calculate blood glucose level from subcutaneous tissue glucose level. In this paper, we present a modification of the Honey Bee Mating Optimization algorithm to determine parameters of glucose-dynamics model. The proposed modification introduces effect of aging into the mating process. We demonstrate that aging improves parameter determination, especially when escaping local extreme of a fitness function.