MANAGEMENT OF THE OPTIMIZATION ALGORITHM ON THE BASIS OF MODELING ARTIFICIAL IMMUNE SYSTEM
DOI:
https://doi.org/10.34185/1991-7848.itmm.2021.01.045Keywords:
algorithm, mutation, artificial neural systems, compression, operator, optimization, radius of similarityAbstract
The efficiency improvement of the known optimization algorithm based on modeling of the artificial immune system due to the adaptive population compression operator is proposed. The radius of similarity of individuals, which is responsible for whether they can be represented in the next generation, is proposed to be proportional to the radius of mutation of cells - search agents. In this case, the radius of the mutation, and accordingly the radius of similarity proportional to it, should gradually decrease during the operation of the algorithm, in accordance with the optimal solution achievement and proportionally to the iteration number. The proposed approach was tested on a number of problems in real and binary space. The results of solving the test problems showed the high efficiency of the proposed algorithmic approach.
References
Gao Sh. Revisiting Immunological Algorithms for Numerical Optimization / Shangce Gao, Yuki Todo, Beibei Chen, Jianhong Chai // IC: International Journal of Immune Computation. 2013. Vol.1. Pages 9-17.
Zheldak T. A. The algorithm of artificial immune system simulation with Saaty selection operator and onedimensional local search / T. A. Zheldak, V. V. Slesarev, I. G. Gulina // Naukovyj visnyk Nacional'nogo girnychogo universytetu. 2016. №5 s. 149-156.
Zheldak T.A. Optymizacija dvoetapnogo vyrobnyctva metaloprokatu shljahom formuvannja zlyvkiv / T.A. Zheldak, L.S. Korjashkina // Komp'juterne modeljuvannja ta optymizacija skladnyh system. CMOCS-2020: tezy dopovidej 6-i' mizhn. nauk.-tehn. konf. Dnipro: DVNZ UDHTU. 2020. s. 80-81. https://doi:10.32434/CMOCS-2020.
