Multi-objectives optimization and convolution fuzzy C-means: control of diabetic population dynamic

The optimal control models proposed in the literature to a population of diabetics are all single-objective which limits identification alternatives and potential opportunities for different reasons: minimization total does not necessarily imply terms two patients from compartments may support same intensity exercise or severity regime. In this work, we propose multi-objectives model taking into account specificity each compartment such that objective function involves single control. addition, Pontryagin’s maximum principle results expansive devours resources because max-min operators formula is very complex difficult assimilate by diabetologists. our case, use heuristic method, NSGA-II, estimate based on model. Since functions conflicting, obtain Pareto front formed non-dominated solutions fuzzy C-means determine important main strategies typical characterization. To limit human intervention, during period, convolution operator reduce hyper-fluctuations using kernels with size. Several experiments were conducted system highlights four feasible capable mitigating socio-economic damages reasonable budget.