LSMA-TLBO: A hybrid SMA-TLBO algorithm with levy flight based mutation for numerical optimization and engineering design problems

The objective of this paper is to present a novel hybrid algorithm LSMA-TLBO for solving the numerical and engineering design optimization problems. The proposed algorithm utilizes the features of Slime Mould Algorithm (SMA) and the Teaching-Learning Based Optimization (TLBO) to balance the exploitation and exploration ability. In addition, a levy flight based mutation is introduced in this study to maximize the exploration ability of the algorithm. SMA is one of the meta-heuristic algorithms (MHAs) which mimics the oscillation mechanism of slime mould in nature. It is an efficient technique that has been successfully applied to the complex optimization problems (OPs) due to its global exploration ability. However, despite its popularity, the SMA also suffers from some shortcomings, such as local optima stagnation, slow convergence rate and improper balance between exploitation and exploration. In view of this weaknesses, the hybrid algorithm LSMA-TLBO is proposed considering the good global search ability of SMA, fast convergence of TLBO and the features of the levy flight based mutation to maximize the exploration ability. The performance of the proposed LSMA-TLBO is evaluated on 23 different benchmark functions including uni-modal and multi-modal problems. In addition, IEEE CEC-C06 2019 test suite and six engineering design problems are also considered to examine the practicable abilities of the proposed method. The results are compared with several recently proposed state-of-the-art algorithms and the simulation results show that the proposed method outperforms other algorithms. Further, the LSMA-TLBO is executed through statistical testing of hypothesis namely Wilcoxon's rank-sum test. It is concluded that the proposed LSMA-TLBO is more competitive and effective algorithm to solve real-world complex OPs.

Automated summary

This paper presents a novel hybrid algorithm LSMA-TLBO for solving numerical and engineering design optimization problems. The algorithm combines the features of Slime Mould Algorithm and Teaching-Learning Based Optimization to balance exploitation and exploration ability, with the introduction of a levy flight based mutation to maximize exploration ability.