A meta-reinforcement learning method by incorporating simulation and real data for machining deformation control of finishing process.
Changqing LiuYingguang LiChong HuangYujie ZhaoZhiwei ZhaoPublished in: Int. J. Prod. Res. (2023)
Keyphrases
- synthetic data
- reinforcement learning
- significant improvement
- cost function
- dynamic programming
- detection method
- computational complexity
- objective function
- similarity measure
- control system
- control charts
- finite element method
- mutual information
- numerical simulations
- dynamic characteristics
- control strategy
- high accuracy
- support vector machine
- pairwise