Abbasi Zohreh, Shafieirad Mohsen, Amiri Mehra Amir Hossein, Zamani Iman
Department of Electrical and Computer Engineering, University of Kashan, Kashan, Iran.
Electrical and Electronic Engineering Department, Shahed University, Tehran, Iran.
Evol Syst (Berl). 2023;14(3):413-435. doi: 10.1007/s12530-022-09459-9. Epub 2022 Sep 15.
The study of the COVID-19 pandemic is of pivotal importance due to its tremendous global impacts. This paper aims to control this disease using an optimal strategy comprising two methods: isolation and vaccination. In this regard, an optimized Adaptive Neuro-Fuzzy Inference System (ANFIS) is developed using the Genetic Algorithm (GA) to control the dynamic model of the COVID-19 termed SIDARTHE (Susceptible, Infected, Diagnosed, Ailing, Recognized, Threatened, Healed, and Extinct). The number of diagnosed and recognized people is reduced by isolation, and the number of susceptible people is reduced by vaccination. The GA generates optimal control efforts related to the random initial number of each chosen group as the input data for ANFIS to train Takagi-Sugeno (T-S) fuzzy structure coefficients. Also, three theorems are presented to indicate the positivity, boundedness, and existence of the solutions in the presence of the controller. The performance of the proposed system is evaluated through the mean squared error (MSE) and the root-mean-square error (RMSE). The simulation results show a significant decrease in the number of diagnosed, recognized, and susceptible individuals by employing the proposed controller, even with a 70% increase in transmissibility caused by various variants.
由于新冠疫情对全球产生了巨大影响，对其进行研究至关重要。本文旨在采用包括隔离和疫苗接种两种方法的最优策略来控制这种疾病。在这方面，利用遗传算法（GA）开发了一种优化的自适应神经模糊推理系统（ANFIS），以控制被称为SIDARTHE（易感、感染、诊断、患病、识别、受威胁、治愈和灭绝）的新冠疫情动态模型。通过隔离减少诊断和识别出的人数，通过疫苗接种减少易感人数。遗传算法生成与每个选定群体的随机初始数量相关的最优控制措施，作为自适应神经模糊推理系统训练高木-关野（T-S）模糊结构系数的输入数据。此外，还提出了三个定理，以表明在存在控制器的情况下解的正性、有界性和存在性。通过均方误差（MSE）和均方根误差（RMSE）对所提出系统的性能进行评估。仿真结果表明，即使在各种变体导致传播率增加70%的情况下，采用所提出的控制器也能使诊断、识别和易感个体的数量显著减少。
