In this paper a new robust fractional-order predictive controller is presented and employed to regulate the glucose level in type-1 diabetes. The dynamics of the system is fully unknown an it is online estimated by a fractional-order model using interval Type-2 (T2) fuzzy logic system. The proposed control system is composed of two main controllers which are the predictive General T2 Fuzzy Logic Controller (GT2-FLC) and compensator controller. In this structure, the main controller is the GT2-FLC which is optimized via the Biogeography-based Optimization (BBO) algorithm such that to minimize a cost function in a fixed prediction horizon. The compensator controller is designed to guarantee the closed-loop asymptotic stability. The performance of proposed control strategy is examined on the modified Bergman’s model of some patients with time-varying parameters, external noise perturbation and meal disturbances. The effectiveness of the proposed control scheme is verified and is compared with the other T2 fuzzy and well-known model predictive controllers. The results of the paper clearly show the superiority of the proposed T2 fuzzy logic control system.

本文提出了一种新的鲁棒分数阶预测控制器，并将其用于调节1型糖尿病患者的血糖水平。该系统的动力学是完全未知的，它是使用区间Type-2（T2）模糊逻辑系统由分数阶模型在线估计的。所提出的控制系统由两个主要控制器组成，分别是预测通用T2模糊逻辑控制器（GT2-FLC）和补偿器控制器。在这种结构中，主控制器是GT2-FLC，该控制器通过基于生物地理的优化（BBO）算法进行了优化，以使固定预测范围内的成本函数最小。补偿器控制器旨在确保闭环渐近稳定性。在一些时变参数，外部噪声扰动和进餐干扰的患者中，在改进的Bergman模型上检查了所提出的控制策略的性能。验证了所提出控制方案的有效性，并将其与其他T2模糊和著名的模型预测控制器进行了比较。本文的结果清楚地表明了所提出的T2模糊逻辑控制系统的优越性。