28 فروردین 1403

حسن رسائی

مرتبه علمی: استادیار
نشانی:
تحصیلات: دکترای تخصصی / مهندسی صنایع
تلفن: 38305005
دانشکده: دانشکده مدیریت مهندسی

مشخصات پژوهش

عنوان
MAINTENANCE PLANNING FOR A CONTINUOUS MONITORING SYSTEM USING DEEP REINFORCEMENT LEARNING
نوع پژوهش مقاله ارائه شده
کلیدواژه‌ها
Maintenance, Manufacturing Systems, Deep Reinforcement Learning.
پژوهشگران فریبا عزیزی (نفر اول)، حسن رسائی (نفر دوم)، عبدالله صفری (نفر سوم)

چکیده

This paper proposes a maintenance decision-making framework for multi-unit systems using Machine Learning (ML). Speci cally, we propose to use Deep Reinforcement Learning (RL) for a dynamic maintenance model of a multi-unit parallel system that is subject to stochastic degradation and random failures. As each unit deteriorates independently in a three-state ho- mogeneous Markov process, we consider each unit to be in one of three states: healthy, unhealthy, or a failed state. We model the interaction among system states based on the Birth/Birth-Death process. By combining individual com- ponent states, we de ne the overall system state. To minimize costs, we use the Markov Decision Process (MDP) framework to solve the optimal main- tenance policy. We apply the Double Deep Q Networks (DDQN) algorithm to solve the problem, making the proposed RL solution more practical and e ective in terms of time and cost savings than traditional MDP approaches. A numerical example is provided which demonstrates how the RL can be used to nd the optimal maintenance policy for the system under study.