哈里斯鹰优化算法是一种模拟哈里斯鹰在自然界中合作捕猎行为的新型群体智能算法，具有参数少、易实现的特点，广泛应用于复杂优化问题求解。然而，在处理高度复杂的优化问题时，该算法可能会存在陷入局部最优状态和稳定性不足的问题，其收敛速度往往较慢。针对该算法存在的不足之处，本文提出两种改进的哈里斯鹰优化算法。

首先，提出了一种基于混合策略改进的哈里斯鹰优化算法，在种群初始化阶段利用佳点集策略，提高算法的遍历性；在全局搜索阶段引入双曲正余弦权重因子提高算法的全局搜索能力；在局部搜索阶段引入柯西变异算子，帮助算法跳出局部最优；在每次迭代结束后采用了重启策略，提高了算法的收敛精度和后期的搜索能力。本文选取12个测试函数对改进算法进行仿真实验并对结果使用Wilcoxon秩和检验来检验算法的性能，结果表明改进后的算法能够有效克服原哈里斯鹰优化算法中收敛过早和迭代后期易陷入局部最优的缺陷，并提高了算法的收敛精度、稳定性和收敛速度。

其次，在混合策略改进的哈里斯鹰优化算法的基础上，提出了一种基于透镜成像和混合变异的哈里斯鹰优化算法。在种群初始化阶段取消佳点集策略，引入的重启策略和变异策略，会增加种群的多样性；取消双曲正余弦惯性权重策略，引入用透镜成像原理的反向学习策略，作为一个独立的更新机制存在，最后在重启策略之后引入混合变异策略，采用多个突变操作子产生不同的个体，防止算法搜索陷入局部最优。通过实验数据和收敛曲线分析基于透镜成像和混合变异的哈里斯鹰优化算法，验证此算法优于混合策略改进的哈里斯鹰优化算法。并且通过对压力容器设计问题求解结果对比，验证了基于透镜成像和混合变异的哈里斯鹰优化算法在工程优化问题上有很好的适用性。

最后，基于主成分分析方法对数据进行特征提取，利用基于透镜成像和混合变异的哈里斯鹰优化算法对随机森林进行参数调整，构建IHHO-RF模型进行空气质量预测研究，利用评价指标评估模型的预测效果，验证了所建立的模型具有更佳的预测效果。

Harris hawks optimization algorithm is a new group intelligence algorithm that simulates the cooperative hunting behavior of Harris hawks in nature. It has the characteristics of few parameters and easy to realize, and is widely used in solving complex optimization problems. However, when dealing with highly complex optimization problems, the algorithm may fall into local optimal states and lack of stability, which often converge more slowly. Aiming at the shortcomings of this algorithm, this paper proposes two improved Harris hawks optimization algorithms.

Firstly, a modified harris hawks optimization algorithm based on hybrid strategy improvement was proposed. In the population initialization stage, the strategy of the best point set was utilized to enhance the algorithm's exploration ability; in the global search stage, the hyperbolic sine and cosine weight factors were introduced to improve the algorithm's global search capability; in the local search stage, the Cauchy mutation operator was introduced to help the algorithm escape from the local optimum; and in each iteration after the end, the restart strategy was adopted to improve the algorithm's convergence accuracy and search ability in the later stage. Twelve test functions were selected for simulation experiments on the improved algorithm, and the Wilcoxon rank sum test was used to verify the performance of the algorithm. The results show that the improved algorithm can effectively overcome the defects of the original harris hawks optimization algorithm, such as premature convergence and easy falling into local optimum in the later iterations, and improve the algorithm's convergence accuracy, stability and convergence speed.

Secondly, based on the hybrid strategy improved harris hawks optimization algorithm, an improved harris hawks optimization algorithm based on lens imaging and hybrid mutation (IHHO) is proposed. In the population initialization stage, the strategy of the best point set is cancelled. The introduced restart strategy and mutation strategy will increase the diversity of the population; the hyperbolic sine and cosine inertia weight strategy is cancelled, and the reverse learning strategy based on the lens imaging principle is introduced as an independent update mechanism. Finally, after the restart strategy, the hybrid mutation strategy is introduced, and multiple mutation operators are used to generate different individuals to prevent the algorithm from getting stuck in local optimum. Through the analysis of experimental data and convergence curves of the harris hawks optimization algorithm based on lens imaging and hybrid mutation, it is verified that this algorithm is superior to the improved harris hawks optimization algorithm with a hybrid strategy. Moreover, by comparing the solution results of the pressure vessel design problem, it is verified that the harris hawks optimization algorithm based on lens imaging and hybrid mutation has good applicability in engineering optimization problems.

Finally, based on the principal component analysis method, the features of air quality data are extracted, and the IHHO algorithm is used to adjust and optimize the parameters of RF to construct the IHHO-RF model for air quality prediction research. The prediction effect of the model is evaluated using evaluation indicators, and it is verified that the established model has better prediction performance.

[1]Kirkpatrick S, Gelatt C D, Jr, Vecchi MP. Optimization by simulated annealing[J]. Science, 1987, 220:671-680.

[2]Muhlenbein H, Schlierkamp-Voosen D. Predictive models for the breeder genetic algorithm: I Continuous parameter optimization[J]. Evolutionary Computation, 1993, 1: 25-49.

[3]Kennedy J. Particle swarm optimization[C]. Proceeding of IEEE International Conference, Neural Networks, Perth, Australia, 2011: 1942-1948.

[4]Dorigo M, Birattari M, Stutzle T. Ant colony optimization[J]. IEEE Computational Intelligence Magazine, 2006, 1(4): 28-39.

[5]Karaboga D, Basturk B. A powerful and efficient algorithm for numerical function optimization: artificial bee colony (ABC) algorithm[J]. J. Global Optimization, 2007, 39 (3): 459-471.

[6]Rashedi E, Nezamabadi-pour H, Saryazdi S. GSA: A Gravitational Search Algorithm[J]. Information Sciences, 2009, 179(13): 2232-2248.

[7]Rao R, Savsani V, Vakharia D. Teaching-Learning-Based Optimization: An optimization method for continuous non-linear large scale problems[J]. Information Sciences, 2011, 183(1): 1-15.

[8]Lam S Y A, Li K O V. Chemical-Reaction-Inspired Metaheuristic for Optimization[J]. IEEE Trans. Evolutionary Computation, 2010, 14 (3): 381-399.

[9]Du P, Wang J, Hao Y, et al. A novel hybrid model based on multi-objective Harris hawks optimization algorithm for daily PM2.5 and PM10 forecasting[J]. Applies Soft Computing, 2020, 96: 106620.

[10]Uddin P M, Majumder A, Barma D J, et al. An Improved Multi-Objective Harris Hawks Optimization Algorithm for Solving EDM Problems[J]. Arabian Journal for Science and Engineering, 2024: 1-46.

[11]Kaveh A, Rahmani P, Eslamlou D A. An efficient hybrid approach based on Harris Hawks optimization and imperialist competitive algorithm for structural optimization[J]. Engineering with Computers, 2021, 38: 1-29.

[12]Huang M, Zhao H, Chen Y. Research on SAR image quality evaluation method based on improved harris hawk optimization algorithm and XGBoost[J]. Scientific reports, 2024, 14(1): 28364.

[13]张海林, 陈泯融. 基于混合策略的改进哈里斯鹰优化算法[J]. 计算机系统应用, 2023, 32(01): 166-178.

[14]聂春芳. 融合黄金正弦和随机游走的哈里斯鹰优化算法[J]. 智能计算机与应用, 2021, 11(07): 113-119+123.

[15]Li C, Si Q, Zhao J, et al. A robot path planning method using improved Harris Hawks optimization algorithm[J]. Measurement and Control, 2024, 57(4): 469-482.

[16]钱亮, 黄伟, 杨建卫. 基于HHO-ELM的光伏阵列故障诊断方法研究[J]. 电源技术, 2024, 48(02): 345-350.

[17]黄力, 宋爽, 刘闯, 等. 基于IHHO-HKELM输电线路覆冰预测模型[J]. 电力科学与技术学报, 2024, 39(4): 33-41.

[18]Zhou Y, Hao Z. Multi-Strategy Improved Whale Optimization Algorithm and Its Engineering Applications[J]. Biomimetics, 2025, 10(1): 47-47.

[19]程金芮, 金瑾, 张朝龙, 等. 自适应策略优化的粒子群优化算法在神经网络架构搜索中的应用[J]. 计算机应用, 2024, 44(S1): 60-64.

[20]Wang H, Yu X. A ranking improved teaching-learning-based optimization algorithm for parameters identification of photovoltaic models[J]. Applied Soft Computing, 2024, 167: 112371-112371.

[21]冯增喜, 何鑫, 崔巍, 等.混合随机反向学习和高斯变异的混沌松鼠搜索算法[J]. 计算机集成制造系统, 2023, 29(02): 604-615.

[22]Sun Y, Mutalib S, Tian L. Improved Whale Optimization Algorithm with LSTM for Stock Index Prediction[J]. International Journal of Advanced Computer Science and Applications, 2025, 16(1): 283-295.

[23]Wang J, Liu H, Qi X, et al. Tool wear prediction based on SVR optimized by hybrid differential evolution and grey wolf optimization algorithms[J]. CIRP Journal of Manufacturing Science and Technology, 2024, 55: 129-140.

[24]Mei Y, Deng Y. Indoor Visible Light Fingerprint Location Method Based on Marine Predator Algorithm-Optimized Least Squares Support Vector Machine[J]. Applied Sciences, 2024, 14(16): 7416-7416.

[25]Binbusayyis A, Sha M. Stability Prediction in Smart Grid Using PSO Optimized XGBoost Algorithm with Dynamic Inertia Weight Updation[J]. Computer Modeling in Engineering & Sciences, 2024, 142(1): 909-931.

[26]Song H, Wang J, Bei J, et al. Modified snake optimizer based multi-level thresholding for color image segmentation of agricultural diseases[J]. Expert Systems With Applications, 2024, 255: 124624-124624.

[27]Wang S, Guo X, Sun T, et al. Short-Term Photovoltaic Power Forecasting Based on the VMD-IDBO-DHKELM Model[J]. Energies, 2025, 18(2): 403-403.

[28]贾鹤鸣, 刘庆鑫, 刘宇翔, 等. 融合动态反向学习的阿奎拉鹰与哈里斯鹰混合优化算法[J]. 智能系统学报, 2023, 18(01): 104-116.

[29]高建瓴, 喻明毫. 混合差分进化与二次插值改进哈里斯鹰算法[J]. 小型微型计算机系统, 2023, 44(07): 1352-1359.

[30]郭雨鑫, 刘升, 高文欣, 等. 精英反向学习与黄金正弦优化的HHO算法[J]. 计算机工程与应用, 2022, 58(10): 153-161.

[31]吴艳敏, 刘家旗, 王璐, 等. 基于改进哈里斯鹰优化算法的配电网动态重构[J]. 科学技术与工程, 2024, 24(08): 3251-3259.

[32]丁鑫, 郭云川, 张长胜, 等. 递进式融合多策略的改进哈里斯鹰优化算法[J]. 小型微型计算机系统, 2024, 45(09): 2126-2136.

[33]尹德鑫, 张琳娜, 张达敏, 等. 基于混沌透镜成像学习的哈里斯鹰算法及其应用[J]. 传感技术学报, 2021, 34 (11): 1463-1474.

[34]陈功, 曾国辉, 黄勃, 等. 融合互利共生和透镜成像学习的HHO算法[J]. 计算机工程与应用, 2022, 58 (10): 76-86..

[35]钱亮, 黄伟, 杨建卫. 基于HHO-ELM的光伏阵列故障诊断方法研究[J]. 电源技术, 2024, 48(02): 345-350.

[36]白雪, 白永国. 基于改进哈里斯鹰算法的SVM与特征选择同步优化[J]. 计算机工程与设计, 2023, 44(05): 1537-1546.

[37]黄霖, 符强, 童楠. 基于自适应调整哈里斯鹰优化算法求解机器人路径规划问题[J]. 计算机应用, 2023, 43(12): 3840-3847.

[38]张林, 沈佳颖, 胡传陆, 等. 基于信噪比的学习型哈里斯鹰算法[J]. 北京航空航天大学学报, 1-17[2025-03-04].

[39]Kardani, Navid, Bardhan, et al. A novel improved Harris Hawks optimization algorithm coupled with ELM for predicting permeability of tight carbonates[J]. Engineering with Computers, 2022, 38: 4323-4346.

[40]Cheng Z, Yong Z, Xuhua P, et al. Improved Harris Hawks Optimization algorithm based on quantum correction and Nelder-Mead simplex method[J]. Mathematical biosciences and engineering: MBE, 2022, 19(8): 7606-7648.

[41]Saikrishnan V, Karthikeyan M. Hybrid harris hawk-arithmetic optimization with deep learning-driven object detection and classification for surveillance video analysis[J]. Multimedia Tools and Applications, 2024: 1-22.

[42]Wen L ,Jianjun J ,Ming X , et al. Lens-imaging learning Harris hawks optimizer for global optimization and its application to feature selection[J]. Expert Systems With Applications, 2022, 202: 117255.

[43]朱菊香, 谷卫, 罗丹悦, 等. 基于CEEMD-ISSA-LSTM的空气质量预测[J]. 国外电子测量技术, 2022, 41(11): 120-129.

[44]方晓萍, 陈秀銮, 褚琦, 等. 基于PSOGSA-LSTM模型的长江经济带空气质量指数预测研究[J]. 数理统计与管理, 2023, 42(01): 14-25.

[45]张爱琳, 白丽娜, 韩液, 等. 基于多变量分数阶灰色模型的郑州市空气质量预测[J]. 安全与环境学报, 2022, 22(04): 2258-2269.

[46]Tang C, Wang Z, Wei Y, et al. A novel hybrid prediction model of air quality index based on variational modal decomposition and CEEMDAN-SE-GRU[J]. Process Safety and Environmental Protection, 2024, 191: 2572-2588.

[47]曹泽轩, 王晓峰, 谢志新, 等. 融合多策略改进的哈里斯鹰优化算法[J]. 郑州大学学报(理学版), 2023, 55(06): 22-28.

[48]丁敬国, 郭锦华. 基于主成分分析协同随机森林算法的热连轧带钢宽度预测[J]. 东北大学学报(自然科学版), 2021, 42(09): 1268-1274+1289.

[49]华罗庚, 王元. 数论在近似分析中的应用[M]. 北京: 科学出版社, 1978.

[50]李书群, 陈钰, 杨雨婷, 等. 基于佳点集人工鱼群的点云配准算法[J]. 合肥工业大学学报(自然科学版), 2023, 46(09): 1203-1209.

[51]冯柯茹. 基于多追随和交叉进化的樽海鞘群算法[J]. 科学技术创新, 2021, (24): 12-13.

[52]Bai J, Li Y, Zheng M, et al. A sinh cosh optimizer[J]. Knowledge-Based Systems, 2023, 282: 111081.

[53]何庆, 林杰, 徐航. 混合柯西变异和均匀分布的蝗虫优化算法[J]. 控制与决策, 2021, 36(07): 1558-1568.

[54]赵峰, 徐丹华. 基于混沌扰动与柯西变异的HHO算法[J]. 安徽大学学报(自然科学版), 2023, 47(04): 25-34.

[55]Naik K M, Panda R, Wunnava A, et al. A leader Harris hawks optimization for 2-D Masi entropy-based multilevel image thresholding[J]. Multimedia Tools and Applications, 2021, 80(28-29): 1-41.

[56]Nadimi-Shahraki H M, Taghian S, Mirjalili S. An Improved Grey Wolf Optimizer for Solving Engineering Problems[J]. Expert Systems With Applications, 2021, 166: 113917.

[57]徐航, 张达敏, 王依柔, 等. 混合策略改进鲸鱼优化算法[J]. 计算机工程与设计, 2020, 41(12): 3397-3404.

[58]贾鹤鸣, 陈丽珍, 力尚龙, 等. 透镜成像反向学习的精英池侏儒猫鼬优化算法[J]. 计算机工程与应用, 2023, 59(24): 131-139.

[59]周鹏, 董朝轶, 陈晓艳, 等. 基于Tent混沌和透镜成像学习策略的平衡优化器算法[J]. 控制与决策, 2023, 38(06): 1569-1576.

[60]Wang Y, Cai Z, Zhang Q. Differential Evolution With Composite Trial Vector Generation Strategies and Control Parameters[J]. IEEE Transactions on Evolutionary Computation, 2011, 15(1): 55-66.

[61]吴擎, 徐惟罡, 张春江. 基于师生交流机制的改进类电磁机制算法[J]. 计算机集成制造系统, 2020, 26(04): 1033-1042.

[62]许海超, 李子君, 姜爱霞, 等. 山东省空气质量指数的时空分布特征[J]. 济南大学学报(自然科学版), 2017, 31(02): 168-175.

[63]刘康丽. 基于范围划分的区域大气污染联防联控机制优化研究[D]. 杭州: 浙江工业大学, 2020.

[64]康新悦. 基于模糊时间序列模型的空气质量指数研究[D]. 南京: 南京师范大学, 2021.

[65]樊雪丰. 赤峰市空气质量变化特征及气象影响分析[J]. 内蒙古科技与经济, 2024, 15: 100-103.

[66]赵欣. 基于因子分析的LSTM-TCN-Attention空气质量预测模型[D]. 开封: 河南大学, 2023.

[67]叶宜涛. 基于人工神经网络的哈尔滨市空气质量分析及预测[D]. 哈尔滨: 东北农业大学, 2022.

[68]Zhang H, Gong Q. A Novel Ultra-High Voltage Direct Current Line Fault Diagnosis Method Based on Principal Component Analysis and Kernel Density Estimation[J]. Sensors, 2025, 25(3): 642-642.