Sign In

SIMPRO

Aalto subproject 2

Hybrid nature-inspired computing methods in optimal electrical machine design

Demand

During recent years, nature has been a rich information resource, inspired by which numerous intelligent computing methodologies have been proposed, developed, and studied. The Nature-Inspired Computing (NIC) methods use nature as a metaphor, inspiration, and enabler.

Solution

What are Nature-Inspired Computing (NIC) methods?

Nature-Inspired Computing (NIC) methods are based on the principles of how the nature works.

 

NIC methods consist of

  • Particle Swarm Optimization (PSO)
  • Ant Colony Optimization (ACO)
  • Harmony Search (HS)
  • Cultural Algorithm (CA)
  • Artificial Immune Systems (AIS)
  • etc.

Hybrid NIC algorithms

Individual NIC methods have their own disadvantages. Hybrid NIC methods can yield superior performances. Some examples of NIC algorithms are:

 

Fusion of HS and CA: HS-CA

 

A hybrid HS method with dual memory: DUAL-HS

NIC Method-based optimal design of a wind generator

 

Optimal wind generator design Design parameters of a wind generator

 

​Parameters​SymbolsRanges​
​Stator outer diameterDse​3.0 – 15.0 m
​Stator core length including the ventilation ductsl0.2 – 3.0 m
​Stator yoke heighthys0.01 – 0.5 m
Rotor yoke height ​hyr0.01 – 0.5 m
​​Stator slot heighthslot0.07 – 0.3 m
​​Stator slot widthb40​.007 – 0.04 m
​Maximum flux density in air gapBmax​​0.4 – 0.9 T
​Number of effective conductors in stator slotzs8 – 26
​Number of poles pairs ​p20 – 100

 

HS-CA-based optimal design of wind generator

 

Wind generator design costs acquired by HS and DUAL-HS

 

​HS​DUAL-HS
​Best Cost (€)​6.5078×105​6.5050×105
Worst Cost (€)​6.7907×105​6.7809×105
Average Cost (€)​6.5844×105​6.5687×105

Hardware implementation of hybrid NIC methods

Implementation of hybrid NIC methods on single computers using CUDA chips

 

Implementation of hybrid NIC methods on networked computers using Cloud Computing (CC)

 

Hardware-based hybrid NIC methods for motor design

 

Conclusions

Fusion/hybridization of different NIC methods is studied to improve the optimization performances of standalone techniques. Hybrid NIC methods are applied to deal with the optimal design problems of electrical machines. CUDA- and CC-based hardware implementation of the hybrid NIC methods is developed. No Free Lunch (NFL) indicates no single optimization method is the best.

Benefits and use cases

The research done in the subproject provides a set of optimization approaches for complex cases.

Subproject deliverables

  1. X. Z. Gao, X. Wang, and K. Zenger
    A modified harmony search method for wind generator design
    International Journal of Bio-inspired Computing, vol. 5, no. 6, 2013
    Link: http://www.inderscience.com/info/inarticle.php?artid=58911
    DOI: 10.1504/IJBIC.2013.058911
  2. X. Z. Gao, X. Wang, T. Jokinen, S. J. Ovaska, A. Arkkio, and K. Zenger
    A hybrid PBIL-based harmony search method
    Neural Computing and Applications, vol. 21, no. 5, 2012
    Link: http://link.springer.com/article/10.1007%2Fs00521-011-0675-6
    DOI: 10.1007/s00521-011-0675-6
  3. X. Z. Gao, X. Wang, S. J. Ovaska, and K. Zenger
    A hybrid optimization method of harmony search and opposition-based learning
    Engineering Optimization, vol. 44, no. 8, 2012
    Link: http://www.tandfonline.com/doi/abs/10.1080/0305215X.2011.628387
    DOI: 10.1080/0305215X.2011.628387
  4. X. Z. Gao, X. Wang, T. Jokinen, S. J. Ovaska, A. Arkkio, and K. Zenger
    A hybrid optimization method for wind generator design
    International Journal of Innovative Computing, Information and Control, vol. 8, no. 6, 2012
    Link: http://www.ijicic.org/ijicic-11-03131.pdf
  5. X. Z. Gao, X. Wang, K. Zenger, and Xiaofeng Wang
    Harmony search method-based multi-modal optimal design of wind generator
    39th Annual Conference of the IEEE Industrial Electronics Society, Vienna, Austria November 2013
    Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6699589
    DOI: 10.1109/IECON.2013.6699589
  6. X. Z. Gao, X. Wang, K. Zenger, and Xiaofeng Wang
    A bee foraging-based memetic harmony search method
    IEEE International Conference on Systems, Man, and Cybernetics, Seoul, Korea, October 2012
    Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6377697
    DOI: 10.1109/ICSMC.2012.6377697
  7. X. Z. Gao, X. Wang, K. Zenger, and Xiaofeng Wang
    A novel harmony search method with dual memory
    IEEE International Conference on Systems, Man, and Cybernetics, Seoul, Korea, October 2012
    Link: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=6377696
    DOI: 10.1109/ICSMC.2012.6377696
  8. X. Wang, X. Z. Gao, and K. Zenger
    An Introduction to Harmony Search Optimization Method
    Springer-Verlag, Springer Briefs in Computational Intelligence, 2014
    Link: http://www.springer.com/us/book/9783319083551
    DOI: 10.1007/978-3-319-08356-8_2
  9. Gao, X.-Z. et al.
    Hybrid nature inspired computing (NIC) methods for electrical machine design
    Poster, SIMPRO-ScarFace Joint Seminar 2, August 25–26, 2015, Espoo

Contacts