Abstract: For the shortcomings of the marine predator algorithm, such as slow convergence speed and difficult to escape from the local optimum, an improved marine predator algorithm is proposed. Firstly, chaotic mapping is combined with opposition learning strategy to generate high-quality initial prey population while ensuring ergodicity and randomness. Secondly, the adaptive t-distribution mutation operator is introduced to update the population to increase the diversity of the population and avoid falling into the local optimum. For the updated population, it is divided into elite group and learning group according to fitness. The learning group learns from the average dimension of prey in the elite group, and the prey in the elite group learns from each other to further improve the population quality and search accuracy. 15 test functions are selected and compared to verify that the improved algorithm can effectively improve the convergence speed and optimization accuracy of the original algorithm. Finally, the improved algorithm is applied to wireless sensor network coverage optimization, the experimental results show that the improved algorithm improves the network coverage, and the optimized node distribution is more uniform.

Key words: chaotic mapping, opposition-based learning, adaptive t-distribution, group learning, marine predator algorithm

摘要： 针对海洋捕食者算法存在收敛速度慢、不易逃出局部最优的缺点，提出了一种改进海洋捕食者算法。将混沌映射与对立学习策略相结合，在保证遍历性和随机性的同时，生成高质量的初始猎物种群。引入自适应t分布变异算子更新种群，增加种群多样性，避免陷入局部最优。对更新后的种群，按照适应度分为精英组和学习组，学习组向精英组猎物的平均维度进行学习，精英组内的猎物相互维度学习，进一步提高种群质量和搜索精度。选取15个测试函数，通过对比测试，验证了改进后的算法可以有效提高原算法的收敛速度和寻优精度。将改进后的算法应用于无线传感器网络覆盖优化，实验结果显示，改进后的算法提高了网络覆盖率，优化后的节点分布更加均匀。

关键词: 混沌映射, 对立学习, 自适应t分布, 分组学习, 海洋捕食者算法