Multi-objective optimization algorithm based on kinetic-molecular theory with memory

This article presents multi-objective optimization algorithm based on kinetic-molecular theory with memory (MOMKMTOA) for solving multi-objective optimization problem. In MOMKMTOA, the individuals are divided into the oblivion, the instantaneous library, the short-term library and the long-term library according to memory principles. It employs the model of updating memory and the Oblivion operator to improve the population diversity. The idea of non dominated quick sorting is introduced to distinguish the optimal results and speed up the convergence rate. The evolutionary mechanism of the original algorithm (KMTOA) is retained to maintain the advantages of population diversity. And the introduction of the guiding strategy of memory elite to avoid partial optimization. The algorithm is tested by CEC09 standard test function UFI similar to UFIO. The results show that MOMKMTOA has obvious advantages in accuracy, robustness and distribution.