An Actor–Critic based controller for glucose regulation in type 1 diabetes

Applicability Results of a Nonlinear Model-Based Robust Blood Glucose Control Algorithm

“…We also plan to investigate model-free approaches 37 by implementing the same controller structure to different models, preferably based on the ones used most frequently in AP research by the international community. 13,14 Finally, the capabilities of these controllers will be tested on either the modified Sorensen model or any model not used for the controller design.…”

Controlling blood glucose variability under uncertainty using reinforcement learning and Gaussian processes

“…(21). To test its robustness, the policy 0.15 G was repeatedly applied in 125 independent simulations of a diabetic patient under meal routines I and II.…”

“…These hybrid methods try to benefit from the repetitive nature of insulin therapy to improve iteratively the efficacy of insulin doses by using run-to-run control algorithms [20,21,42,43]. In a situation with frequent data sampling, iterative learning control (ILC) is the alternative of choice.…”

GoCARB in the Context of an Artificial Pancreas

“…2 The principle of the AC algorithm relies on RL and approximates an optimal control strategy through real-time learning in an iterative manner. It evaluates the current control policy (critic) and updates it (actor), according to the patient-specific characteristics yielding an optimized insulin infusion policy.…”

“…2,6 Although realtime learning permits the continuous adjustment and personalization of insulin therapy that overcomes most disturbances, the announcement of meals may still add robustness against patient variability and uncertainties.…”