A Data-Driven Automatic Design Method for Electric Machines Based on Reinforcement Learning and Evolutionary Optimization.
Takahiro SatoMasafumi FujitaPublished in: IEEE Access (2021)
Keyphrases
- data driven
- dynamic programming
- evolutionary optimization
- reinforcement learning
- fully automatic
- preprocessing
- high accuracy
- clustering method
- detection method
- error rate
- significant improvement
- cost function
- case study
- segmentation method
- genetic algorithm
- semi automatic
- decision rules
- similarity measure
- user interface
- pairwise
- multi agent