基于反向学习的DE-RSO混合优化算法

Opposition-Based Learning DE-RSO Hybrid Optimizer

摘要: 鼠群优化算法(Rat Swarm Optimizer, RSO)是一种可以解决全局优化问题的新型仿生优化算法，它的灵感主要来自于自然界种鼠群追逐猎物和与猎物搏斗的行为。然而，它具有收敛速度过慢和收敛精度不高的缺陷，为解决这一问题，本文提出了一种改进的鼠群优化算法——基于反向学习的DE-RSO混合优化算法(Opposition-Based Learning DE-RSO Hybrid Optimizer, OBLDE-RSO)。该算法使用了DE-RSO混合策略和反向学习策略，DE-RSO混合策略可以保持种群多样性并降低算法陷入局部最优的可能性，反向学习策略可以针对性地扩大个体的搜索范围，使个体以更高概率找到潜在的更加理想的求解区域。本文用29个IEEE CEC2017基准测试函数对OBLDE-RSO进行测试，并与其他经典算法的测试结果进行对比，实验结果表明，该算法在收敛精度和收敛速度方面都具有良好的性能。

Abstract: Rat Swarm Optimizer Algorithm (RSO) is a novel bio-inspired optimization algorithm that can solve global optimization problems. It is inspired by the behavior of wild rats chasing and fighting prey. However, it has the shortcomings of slow convergence speed and low convergence accuracy. To solve this problem, this paper proposes an improved rat swarm optimization algorithm—Opposition-Based Learning DE-RSO Hybrid Optimizer (OBLDE-RSO). The algorithm uses the DE-RSO hybrid strategy and the Opposition-Based Learning strategy. The DE-RSO hybrid strategy can maintain the diversity of the population and reduce the possibility of the algorithm falling into the local optimum. The Opposition-Based Learning strategy can expand the search range of the individual in a targeted manner and make the individual find a potentially more ideal solution area with a higher probability. In this paper, 29 IEEE CEC2017 benchmark functions are used to test OBLDE-RSO and compared with the test results of other classic algorithms. The experimental results show that the algorithm has good performance in terms of convergence accuracy and convergence speed.