TY - JOUR
T1 - Performance Improvement of Element Description Method Using Artificial Bee Colony Algorithm
AU - Takeuchi, Issei
AU - Katsura, Seiichiro
N1 - Funding Information:
to compare with the GA of the conventional method. Experiments confirmed that the ABC algorithm improves the search performance of the EDM. Particularly, the avoidance performance of the local solution of the ABC algorithm is significantly higher than that of the GA. The avoidance performance of the local solution appears in the mean and standard deviation of the training data. In this study, the best performance was made when N is 8 in the ABC algorithm. By using the ABC algorithm, more accurate modeling can be expected by the EDM. In future work, we will consider shortening the search time by improving the search efficiency by improving the search behavior of bees. Then, we plan to apply the proposed method to a larger search space and confirm the modeling performance. Acknowledgment This research was partially supported by the Ministry of Internal Affairs and Communications, Strategic Information and Communications R&D Promotion Programme (SCOPE), 201603011, 2022.
Publisher Copyright:
© 2022 The Institute of Electrical Engineers of Japan.
PY - 2022
Y1 - 2022
N2 - In order to improve control performance in various control fields, it is important to model the controlled object accurately. In this case, the quality of the model is considerably influenced by the structure of the model determined by the engineer. An element description method is a method that can optimize not only parameters but also the structure of the model. Therefore, it is possible to search over a wide range without being restricted by human design. However, this considerably increases the search space, and it is easy to fall into a local solution. In this study, the artificial bee colony algorithm is combined with the element description method to improve its search ability. The artificial bee colony algorithm is known to be effective for high-dimensional and multimodal problems. The performance of the proposed method is validated using a heat sealing system in packaging machinery. The proposed method is evaluated in comparison with the genetic algorithm, which is a conventional method. Experiments confirm that the local solution avoidance performance of the artificial bee colony algorithm is significantly better than that of the genetic algorithm.
AB - In order to improve control performance in various control fields, it is important to model the controlled object accurately. In this case, the quality of the model is considerably influenced by the structure of the model determined by the engineer. An element description method is a method that can optimize not only parameters but also the structure of the model. Therefore, it is possible to search over a wide range without being restricted by human design. However, this considerably increases the search space, and it is easy to fall into a local solution. In this study, the artificial bee colony algorithm is combined with the element description method to improve its search ability. The artificial bee colony algorithm is known to be effective for high-dimensional and multimodal problems. The performance of the proposed method is validated using a heat sealing system in packaging machinery. The proposed method is evaluated in comparison with the genetic algorithm, which is a conventional method. Experiments confirm that the local solution avoidance performance of the artificial bee colony algorithm is significantly better than that of the genetic algorithm.
KW - artificial bee colony algorithm
KW - element description method
KW - genetic algorithm
KW - particle swarm optimization
KW - system identification
KW - temperature control
UR - http://www.scopus.com/inward/record.url?scp=85141925207&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85141925207&partnerID=8YFLogxK
U2 - 10.1541/ieejjia.21005358
DO - 10.1541/ieejjia.21005358
M3 - Article
AN - SCOPUS:85141925207
SN - 2187-1094
VL - 11
SP - 643
EP - 649
JO - IEEJ Journal of Industry Applications
JF - IEEJ Journal of Industry Applications
IS - 5
ER -