An evolutionary algorithm for inseparable problems based on immune clone.

Huijun Li,Ying Wang,Dan Zhang
DOI: https://doi.org/10.1145/3456529.3456551
2021-01-01
Abstract:Many problems in society and life can be solved by modeling as large-scale global optimization problems. The large-scale global optimization problems are very complex and difficult to solve. At present, for the CEC2013 large-scale test set, the existing algorithms cannot obtain satisfactory results. The large-scale global optimization problems remained to be solved, especially the inseparable problems among them. This paper proposes an algorithm that combines immune cloning algorithm with differential evolution algorithm. In this algorithm, clone immune operation and receptor editing mechanism are introduced to enhance the local searching ability, to improve the global optimization ability and convergence speed. The algorithm is used to solve the inseparable problems in the test set. The experimental results show that the algorithm can not only effectively avoid premature convergence, but also improve the convergence speed.
What problem does this paper attempt to address?