An Improved Remora Optimization Algorithm with Lens Opposition-Based Learning for Global Optimization

The remora optimization algorithm (ROA) is one of the newly proposed swarm behavior-based algorithms. During the implementation of ROA, each search agent updates its position with the host, which leads to the algorithm difficult to converge and solve complex engineering problems. To alleviate these situations, Lens Opposition-Based Learning (LOBL) is added in this paper to optimize the performance of ROA when solving complex problems. The improved algorithm called IROA. LOBL is inspired by the imaging principle of the optical lens. Moreover, the LOBL strategy can improve the probability of the traditional ROA escaping from the local optimal solution. To bear out that the proposed IROA is functional, this paper selects 13 benchmark functions to prove the superiority of the improvement. By comparing IROA with five meta-heuristics, the experiment results illustrate that the IROA is more effective in working out benchmark optimization problems. Additionally, the performance of IROA in car crashworthiness design also reveal that IROA can get the optimal solution in solving practical optimization problems. Generally speaking, all the above experimental results validate the effectiveness of the proposed algorithm.