Scheduling Condition-based Maintenance: An Explainable Deep Reinforcement Learning Approach via Reward Decomposition.
Huong N. DangKuo-Chu ChangGenshe ChenHua-mei ChenSimon KhanMilvio FrancoErik BlaschPublished in: FUSION (2023)
Keyphrases
- reinforcement learning
- preventive maintenance
- function approximation
- scheduling problem
- scheduling algorithm
- state space
- reinforcement learning algorithms
- model free
- partially observable environments
- eligibility traces
- learning algorithm
- round robin
- software maintenance
- transfer learning
- multi agent
- machine learning
- markov decision processes
- dynamic programming
- partially observable
- sufficient conditions
- maintenance cost
- action selection
- resource consumption
- decomposition algorithm
- resource constraints
- resource allocation
- reward function
- supervised learning
- decomposition method
- optimal policy
- action space
- policy search
- np hard
- markov decision problems
- flexible manufacturing systems
- learning agent
- deep learning
- temporal difference