Solving geometry conflicts in GA Optimizations with large numbers of geometric parameters
- Resource Type
- Conference
- Authors
- Schneider, N.; Kanamaru, M.; Sano, H.; Yamada, T.
- Source
- 2022 International Conference on Electrical Machines (ICEM) Electrical Machines (ICEM), 2022 International Conference on. :1034-1040 Sep, 2022
- Subject
- Engineering Profession
Power, Energy and Industry Applications
Signal Processing and Analysis
Geometry
Torque
Sociology
Machine learning
Predictive models
Torque measurement
Statistics
Genetic algorithm
Optimization
Geometry conflict problems
Electric machines
Machine Learning
Finite element analysis
- Language
Genetic Algorithm (GA) based optimization with finite element analysis is expected to become a solution for the high performance motor designs which need to satisfy many strict requirements. However, as its use goes towards complex design which having many geometric design parameters, the technique becomes difficult to use. The problem is the fact that the many design parameters lead to geometry conflicts and make the optimization malfunction. The main cause of this is that the geometry conflicts prevent generating an appropriate population at the initial stage of the GA optimization. To overcome this, a technique which consists of an optimization of parameter ranges and a predicted model using machine learning was developed. The proposed technique was applied to an optimization problem of a PMSM with over 30 geometric design parameters confirming that the initial population was appropriately generated and the optimization can be performed successfully.