Naked mole-rat algorithm (NMRA) is a new swarm intelligence technique based on the mating patterns of NMRs present in nature. The algorithm though is very simple and linear in nature but suffers from poor exploration during the initial stages and poor exploitation towards the end. Thus to overcome these problems and estimate the effect of basic parameters of NMRA, six new inertia weight strategies and five new mutation operators have been employed. After careful investigation, a new Lévy mutated NMRA (LNMRA) is proposed. The new algorithm employs combined properties of inertia weights and mutation operators altogether. For performance evaluation, the proposed algorithms are subjected to variable initial population and dimension sizes and testing is done on CEC 2005, CEC 2014 benchmark problems and real world optimization problem of dual band-notched ultra-wideband (UWB) antenna design. Experimental and statistical results show that the proposed LNMRA is better with respect to other algorithms under comparison.

裸鼠鼠算法（NMRA）是一种新的群智能技术，它基于自然界中存在的NMR的交配模式。尽管该算法本质上非常简单且线性，但在初始阶段就缺乏探索性，而在最后阶段却受到不良利用。因此，为了克服这些问题并估计NMRA基本参数的效果，已采​​用了六种新的惯性权重策略和五种新的突变算子。经过仔细研究，提出了一种新的Lévy突变NMRA（LNMRA）。新算法完全利用了惯性权重和变异算子的组合属性。为了进行性能评估，对提出的算法进行了可变的初始填充和尺寸调整，并在CEC 2005上进行了测试，CEC 2014基准测试问题和双带陷波超宽带（UWB）天线设计的现实世界优化问题。实验和统计结果表明，相比于其他算法，所提出的LNMRA更好。