0% Complete
صفحه اصلی
/
سی و سومین کنفرانس بین المللی مهندسی برق
Eligibility Trace Algorithm for Fault Tolerant Control of Diabetic Blood Glucose Regulation Systems by Time Window-based
نویسندگان :
Amin Noori
1
Elnaz Kalhor
2
1- دانشگاه صنعتی سجاد مشهد
2- دانشگاه صنعتی سجاد مشهد
کلمات کلیدی :
Reinforcement learning،Fault tolerant control،Diabetes،Eligibility trace،Drug dosage
چکیده :
The current paper investigates the blood glucose level in the presence of a sensor fault. To this aim, the eligibility trace algorithm, as a Reinforcement Learning (RL) method, is employed. Here, a soft and abrupt fault is applied to the system and a time window concept detects the fault type. At each time step, a feedforward neural network monitors the amplitude and type of the applied fault as these can affect the injected dosage. In order to comprehensively investigate the performance of the eligibility trace algorithm on the blood glucose level control in diabetic patients, the present study compares the proposed method with another featuring a combination of the eligibility trace algorithm and a sliding mode control. Finally, it is demonstrated that, with a lower dosage, the eligibility trace algorithm is able to control the blood glucose level in the presence of a sensor fault and so attain the desired level. More importantly, in the proposed method, the controller design is much easier in comparison with the combination of the eligibility trace algorithm and the sliding mode control.
لیست مقالات
لیست مقالات بایگانی شده
Design of a Three-Stage OTA with Wide Capacitive Load Range Using Dual-Path and Q-Factor Compensation
Mohammadreza Abedi Orang
Analysis of an E-core Permanent Magnet Switched Reluctance Motor
Ali Ghaffarpour - Mojtaba Mirsalim
Improving CycleGAN-VC2 Voice Conversion by Learning MCD-Based Evaluation and Optimization
Majid Behdad - Davood Gharavian
Secrecy Sum Rate Analysis and Power Allocation with OSTBC and Artificial Noise for MIMO Systems
Abdolrasoul Sakhaei Gharagezlou - Mahdi Nangir - Nima Imani - Amir Poorfaraj Liqvan
A high speed method for features extraction in face recognition systems
Hosein Khorami - Hadishahriar Shahhoseini
Digitizing Analog ECGs: A Deep Learning Pipeline for Converting Historical Records into High-Quality Digital Signals
Sahar Askari - Somayeh Afrasiabi
Job Title Prediction from Tweets Using Word Embedding and Deep Neural Networks
Shayan Vassef - Ramin Toosi - Mohammad Ali Akhaee
Low-cost Broadband Reflectarray Antenna Using Cross Bow-Tie elements
Mahdieh Bozorgi - Mahmood Rafaei-Booket
Reactive Power Compensation in Distribution Grids: An Application of Trinary Cascaded H-bridge Multilevel Inverter
Yousef Neyshabouri - Mohammad Farhadi-Kangarlu
شناسایی کمپلکس های پروتئینی با رویکرد خوشه بندی EM و با استفاده از داده های زیستی
مریم مولی وردیخانی - سعید جلیلی
بیشتر
ثمین همایش، سامانه مدیریت کنفرانس ها و جشنواره ها - نگارش 43.6.0