基于种群熵偏移平均加权的改进量子粒子群算法

Improved Quantum Particle Swarm Optimization Based on Population Entropy Offset Mean Weighting

量子粒子群算法具有更好的全局搜索能力,被视为对粒子群算法极为有效的改进,然而其运行过程中仍存在种群多样性衰减问题。为进一步提升量子粒子群算法的全局寻优能力,在基于加权平均最优位置的量子粒子群算法的基础上,提出了基于种群熵偏移平均加权的改进量子粒子群算法,将种群熵与加权范围中心偏移值进行动态关联,有效增强了算法搜索空间的遍历性,避免了算法早熟收敛。应用常规测试函数,与传统粒子群算法、量子粒子群算法和加权量子粒子群算法进行了对比分析,证明了文章提出的改进算法的有效性。

The Quantum Particle Swarm Optimization has better global search ability and is considered an extremely effective improvement to the Particle Swarm Optimization.However,there is still a problem of population diversity decay during its operation.In order to further enhance the global optimization ability of Quantum Particle Swarm Optimization,an improved Quantum Particle Swarm Optimization based on population entropy offset mean weighting is proposed,which is based on Quantum Particle Swarm Optimization in weighted mean optimal position.Dynamically associating the population entropy with the weighted range center offset value effectively enhances the traversal of the algorithm's search space and avoids premature convergence of the algorithm.By applying conventional test functions,a comparative analysis is conducted with traditional Particle Swarm Optimization,Quantum Particle Swarm Optimization,and weighted Quantum Particle Swarm Optimization,demonstrating the effectiveness of the improved algorithm proposed in the paper.