一种改进的原子搜索算法

doi:10.16182/j.issn1004731x.joss.20-0824

摘要/Abstract

摘要：

原子搜索算法（atom search algorithm，ASO）是模仿自然界中原子运动而提出的一种新型优化算法，针对ASO在求解复杂函数时存在易早熟及收敛速度慢的问题，提出了一种改进ASO算法（improved atomic search algorithm，IASO）。IASO加入了原子个体历史最优解产生的约束力来修正ASO的加速度，增强全局搜索能力。自适应更新2个乘数系数来协调算法的全局搜索和局部开发能力。适时采用高斯变异策略来重新更新原子位置，提高跳出早熟的能力。对14个基准函数进行仿真实验，对比其他算法，IASO在收敛速度、收敛精度方面表现出优越的性能。

关键词: 原子优化算法, 函数优化, 自适应, 高斯变异, 收敛精度, 测试函数

Abstract:

The atom search algorithm (ASO) is a new optimization algorithm proposed by imitating the movement of atoms in the natural world. An improved atomic search algorithm (IASO) is proposed to address the problems of prematureness and slow convergence of ASO in solving complex functions. IASO adds the binding force generated by the historical optimal solution of individual atoms to correct the acceleration of ASO and enhance the global search capability. The two multiplier coefficients are adaptively updated to coordinate the algorithm's global search and local development capabilities. The Gaussian mutation strategy is used to re-update the atomic position and improve the ability to jump out of precocity. Carrying out simulation experiments on 14 benchmark functions and comparing other algorithms, IASO shows superior performance in terms of convergence speed and convergence accuracy.

Key words: atomic optimization algorithm, function optimization, self-adaptation, Gaussian mutation, convergence accuracy, test function

中图分类号:

TP301.6

李建锋, 卢迪, 李贺香. 一种改进的原子搜索算法[J]. 系统仿真学报, 2022, 34(3): 490-502.

Jianfeng Li, Di Lu, Hexiang Li. An Improved Atomic Search Algorithm[J]. Journal of System Simulation, 2022, 34(3): 490-502.

图/表 9

图1

图2

表1

表2

基准测试函数

Benchmark function	Dim	Feature	Range	f_min
$f 1 x = ∑ i = 1 n x i 2$	30	[-100,100]	[-100,100]	0
$f 2 x = ∑ i = 1 n x i + ∏ i = 1 n x i$	30	[-10,10]	[-10,10]	0
$f 3 x = ∑ i = 1 n ∑ j = 1 i x j 2$	30	[-100,100]	[-100,100]	0
$f 4 x = m a x i x i, 1 ≤ i ≤ D$	30	[-100,100]	[-100,100]	0
$f 5 x = ∑ i = 1 n x i + 0.5 2$	30	[-100,100]	[-100,100]	0
$f 6 x = ∑ i = 1 n x i 4 + r a n d 0,1$	30	[-1.28,1.28]	[-1.28,1.28]	0
$f 7 x = ∑ i = 1 n - x i s i n x i$	30	[-500,500]	[-500,500]	-418.982 9×5
$f 8 x = ∑ i = 1 n x i 2 - 10 c o s 2 π x i + 10$	30	[-5.12,5.12]	[-5.12,5.12]	0
$f 9 x = - 20 e x p - 0.2 1 n ∑ i - 1 n x i 2 - e x p 1 n ∑ i = 1 n c o s 2 π x i + 20 + e$	30	[-32,32]	[-32,32]	0
$f 10 x = 1 4 ? 000 ∑ i = 1 n x i 2 - ∏ i = 1 n c o s x i i + 1$	30	[-600,600]	[-600,600]	0
$f 11 x = 0.1 s i n 2 3 π x 1 + ∑ i - 1 n x i - 1 2 1 + s i n 2 3 π x 1 + 1 + x n - 1 2 1 + s i n 2 2 π x n + ∑ i = 1 n u x i, 5,100,4$	30	[-50,50]	[-50,50]	0
$f 12 x = 1500 + ∑ j = 1 25 1 j + ∑ i = 1 2 x i - a i j 6 - 1$	2	[-65,65]	[-65,65]	1
$f 13 x = ∑ i = 1 11 a i - x 1 b i 2 + b i x 2 b i 2 + b i x 3 + x 4 2$	4	[-5,5]	[-5,5]	0.000 30
$f 14 = - ∑ i = 1 7 X - a i X - a i T + C i - 1$	4	[0.10]	[0.10]	-10.402 8

表2

图3

图4

表3

测试函数实验结果

Function	Result	Algorithm
Function	Result	GWO	MFO	WOA	ASO	IASO
$f 1$	Best	0	0.000 1E-11	0	0.003 6E-21	0
	Worst	0.786 4E-58	0.144 1E-11	0.907 2E-80	0.100 0E-21	0.166 5E-110
	Mean	0.045 1E-58	0.028 5E-11	0.030 2E-80	0.034 6E-21	0.005 6E-110
	Std	0.146 0E-58	0.041 2E-11	0.165 6E-80	0.027 8E-21	0.030 4E-110
$f 2$	Best	0.000 3E-31	0.003 8E-7	0	0.008 0E-9	0
	Worst	0.365 6E-31	0.206 8E-7	0.171 2E-52	0.168 4E-9	0.777 5E-80
	Mean	0.054 3E-31	0.055 3E-7	0.016 9E-52	0.043 6E-9	0.025 9E-80
	Std	0.083 3E-31	0.053 7E-7	0.042 7E-52	0.032 0E-9	0.141 9E-80
$f 3$	Best	0	0	1.484 2	1.176 7E-17	0
	Worst	0.101 7E-27	0.005 1	6.981 9	0.007 8E-17	0.285 1E-142
	Mean	0.008 2E-27	0.008 3	4.309 4	0.000 4E-17	0.018 1E-142
	Std	0.023 4E-27	0.019 1	1.171 1	0.001 5E-17	0.068 9E-142
$f 4$	Best	0	0.001 6	0.002 5	0.070 3E-11	0
	Worst	0.196 2E-16	1.551 9	16.573 5	0.992 3E-11	0.654 1E-72
	Mean	0.021 7E-16	0.243 9	1.383 9	0.346 1E-11	0.021 8E-72
	Std	0.041 4E-16	0.296 7	3.447 6	0.249 7E-11	0.119 4E-72
$f 5$	Best	0	0	0.000 1	0	0
	Worst	0.252 2E-5	0.152 5E-12	0.010 3	0.116 7E-22	0.654 5E-22
	Mean	0.016 8E-5	0.236 0E-12	0.001 1	0.033 8E-22	0.205 1E-22
	Std	0.063 8E-5	0.033 3E-12	0.001 8E-2	0.029 9E-22	0.172 3E-22
$f 6$	Best	0.000 1E-3	0.003 3	0.013 4E-2	0.002 7	0.263 3E-4
	Worst	0.001 5E-3	0.024 4	0.003 0E-2	0.043 0	3.147 1E-4
	Mean	0.000 6E-3	0.009 3	0.002 9E-2	0.013 5	1.204 0E-4
	Std	0.000 4E-3	0.005 9	0.389 3	0.009 2	0.745 0E-4
$f 7$	Best	-3.182 7E+3	-3.833 0E+3	-4.189 6E+3	-3.538 4E+3	-3.071 4E+3
	Worst	-2.045 6E+3	-2.521 1E+3	-2.199 6E+3	-2.373 7E+3	-2.886 5E+3
	Mean	-2.702 4E+3	-3.213 9E+3	-3.467 7E+3	-2.783 6E+3	-3.587 0E+3
	Std	0.293 2E+3	0.339 6E+3	0.637 6E+3	0.283 7E+3	0.278 8E+3
$f 8$	Best	0	4.974 8	0	1.989 9	0
	Worst	6.486 2E-13	78.743 5	4.175 1E-13	16.914 3	0
	Mean	1.103 1E-13	25.316 9	3.180 2E-13	8.656 1	0
	Std	2.116 5E-13	15.587 8	1.043 9E-13	3.223 0	0
$f 9$	Best	0.444 1E-14	0	0.008 9E-13	0.070 0E-10	0.888 2E-15
	Worst	0.799 4E-14	19.950 2E-6	0.151 0E-13	0.433 8E-10	0.888 2E-15
	Mean	0.680 9E-14	0.719 9E-6	0.050 3E-13	0.197 5E-10	0.888 2E-15
	Std	0.170 3E-14	3.644 4E-6	0.033 7E-13	0.087 5E-10	0
$f 10$	Best	0	0.017 2	0	0	0
	Worst	0.094 5E-14	0.666 4	0.330 7E-14	0.027 0E-13	0
	Mean	0.024 6E-14	0.155 5	0.048 3E-14	0.004 4E-13	0
	Std	0.021 4E-14	0.133 2	0.087 5E-14	0.007 8E-13	0
$f 11$	Best	0.000 0	0.000 0	0.000 8	0.004 4	0
	Worst	0.201 5	0.011 0	0.211 3	0.302 1	0.804 3E-22
	Mean	0.022 9	0.003 3	0.042 2	0.055 0	0.211 1E-22
	Std	0.049 9	0.005 1	0.049 5	0.061 0	0.190 8E-22
$f 12$	Best	0.998 0	0.998 0	0.998 0	0.998 0	0.998 0
	Worst	12.670 5	7.874 0	10.763 2	0.001 4	1.992 0
	Mean	3.676 0	2.249 2	3.025 8	3.968 9	1.064 5
	Std	3.875 0	1.918 4	3.354 9	1.398 4	0.252 2
$f 13$	Best	0.000 3	0.000 6	0.000 3	0.716 2	0.000 3
	Worst	0.056 6	0.020 4	0.001 5	0.001 4	0.000 7
	Mean	0.005 7	0.002 1	0.000 7	0.001 0	0.000 5
	Std	0.012 2	0.003 9	0.000 3	0.000 2	0.000 1
$f 14$	Best	-10.402 7	-10.402 9	-10.401 0	-10.402 9	-10.402 9
	Worst	-5.127 3	-1.837 6	-2.765 3	-2.751 9	-5.128 8
	Mean	-10.225 8	-6.457 5	-8.493 2	-6.147 9	-10.227 1
	Std	0.962 9	3.095 1	2.653 7	1.396 9	0.962 9

表3

图5

表4

Wilcoxon秩和检验结果

function	Significance level P
function	GWO	MFO	WOA	ASO
$f 1$	0	0	0	0
$f 2$	0	0	0	0
$f 3$	0	0	0	0
$f 4$	0	0	0	0
$f 5$	0	0	0	0.158 0
$f 6$	0	0	0	0
$f 7$	0	0.021 6	0.333 7	0
$f 8$	0	0	0	0
$f 9$	0.000 2	0.001 2	0.000 3	0.001 2
$f 10$	0	0	0.000 1	0
$f 11$	0	0	0	0
$f 12$	0	0.075	0	0.001 9
$f 13$	0.004 9	0	0.001 2	0

表4

参考文献 19

1	Mann Mukesh, Tomar Pradeep,Om Prakash Sangwan. Bio-Inspired Metaheuristics: Evolving and Prioritizing Software Test Data[J]. Applied Intelligence March 2018 (S0924-669X), 2018, 48(3): 687-702,
2	林诗洁, 董晨, 陈明志, 等新型群智能优化算法综述 [J]. 计算机工程与应用, 2018, 54(12): 6-14.
	Lin Shijie, Dong Chen, Cheng Minzhi, et al. Summary of New Group Intelligent Optimization Algorithms[J]. Computer Engineering and Applications, 2018, 54(12): 6-14.
3	Holland, John H. Building Blocks, Cohort Genetic Algorithms, and Hyperplane-Defined Functions[J]. Evolutionary Computation (S1063-6560) 2000, 8(4):373-391.
4	Paolo Rocca, Oliveri Giacomo, Massa Andrea. Differential Evolution as Applied to Electromagnetics[J]. IEEE Antennas & Propagation Magazine (S1558-4143), 2011, 53(1): 3058-3059.
5	Mirjalili S, Mirjalili S M, Lewis A. Grey Wolf Optimizer[J]. Advances in Engineering Software (S0965-9978), 2014, 69(3): 46-61.
6	Mirjalili S, Lewis A. The Whale Optimization Algorithm[J]. Advances in Engineering Software (S0965-9978), 2016, 95(5): 51-67.
7	刘生建, 杨艳, 周永权. 一种群体智能算法——狮群算法[J]. 模式识别与人工智能, 2018, 31(5): 431-441.
	Liu Shengjian, Yang Yan, Zhou Yongquan. A Swarm Intelligence Algorithm-Lion Swarm Optimization[J]. Pattern Recognition and Artificial Intelligence, 2018, 31(5): 431-441.
8	Rashedi E, Nezamabadi-Pour H, Saryazdi S. GSA:A Gravitational Search Algorithm[J]. Information Sciences (S0020-0255), 2009, 179(13): 2232-2248.
9	Hatamlou, Abdolreza. Black hole: A New Heuristic Optimization Approach for Data Clustering[J]. Information ences (S0020-0255), 2013, 222: 175-184.
10	Deng L, Yang P, Liu W. An Improved Genetic Algorithm [C]// 2019 IEEE 5th International Conference on Computer and Communications (ICCC). Chengdu, China: IEEE, 2019: 47-51,.
11	丁青锋, 尹晓宇. 差分进化算法综述[J]. 智能系统学报,2017, 12(4): 431-442.
	Ding Qingfeng, Yin Xiaoyu. Research Survey Of Differential Evolution Algorithms[J]. CAAI Transactions on Intelligent Systems 2017, 12(4): 431-442.
12	Song Z, Tang C, Chen X, et al. A Self-Adaptive Mechanism Embedded Gravitational Search Algorithm[C]// 2019 12th International Symposium on Computational Intelligence and Design (ISCID). Hangzhou, China: Zhejiang University Press 2019: 108-112.
13	Zhao X, Fang Y, Ma Z, et al. An Ameliorated Moth-Flame Optimization Algorithm[C]// 2018 37th Chinese Control Conference (CCC). Wuhan: TCCT, 2018: 2372-2377.
14	Sun Y, Wang X, Chen Y, et al. A modified whale Optimization Algorithm For Large-Scale Global Optimization Problems[J]. Expert Systems with Applications (S0957-4174), 2018, 114: 563-577.
15	Zhao W, Wang L, Zhang Z. A Novel Atom Search Optimization for Dispersion Coefficient Estimation In Groundwater[J]. Future Generation Computer Systems(S0167-739X), 2018, 91: 601-610.
16	Yang W, Chen L, et al. Multi/Many-Objective Particle Swarm Optimization Algorithm Based on Competition Mechanism[J]. Computational Intelligence and Neuroence (S1687-5265), 2020, 2020: 1-26.
17	Guo X Y, Brault P. Early Stages Of Silicon Nitride Film Growth Studied By Molecular Dynamics Simulations[J]. Surface Science (S0039-6028), 2001, 488(1/2): 133-140.
18	高庆吉, 王瑞雪, 谈政. 基于高斯变异和自适应参考点的MOPSO优化算法[J]. 计算机应用与软件, 2019, 36(9): 255-259
	Gao Qingji, Wang Ruixue, Tan Zheng.Multi-Objective Particle Swarm Optimization Based on Gaussian Mutation and Adaptive Reference Point[J]. Computer Applications and Software, 2019, 36(9): 255-259
19	Derrac J, García S, Molina D, et al. A Practical Tutorial On The Use of Nonparametric Statistical Tests as a Methodology for Comparing Evolutionary and Swarm Intelligence Algorithms[J].Swarm and Evolutionary Computation (S2210-6502), 2011, 1(1): 3-18

[1]	朱海龙, 贾如侠, 张亮, 贺维. 基于证据推理和置信规则库的涡扇发动机故障预测[J]. 系统仿真学报, 2022, 34(9): 2074-2086.
[2]	王启明, 蒋江月, 吕志超, 张汉祖. 基于改进MPC的协同自适应巡航控制策略研究[J]. 系统仿真学报, 2022, 34(9): 2087-2097.
[3]	张会林, 金玉洁, 杨海马. ANFIS优化磁链滑模观测器的PMSM无传感器控制[J]. 系统仿真学报, 2022, 34(8): 1682-1690.
[4]	张玉超, 蒋沅, 代冀阳. 三阶多机器人协同编队动态避障控制[J]. 系统仿真学报, 2022, 34(8): 1762-1774.
[5]	张思贤, 杨艺, 张猛, 米鹏博. 高效的多特征自适应相关滤波跟踪器[J]. 系统仿真学报, 2022, 34(8): 1864-1873.
[6]	聂慧, 李康顺, 苏洋. 一种量化因子自适应学习量化训练算法[J]. 系统仿真学报, 2022, 34(7): 1639-1650.
[7]	李智杰, 石昊琦, 李昌华, 张颉. 基于改进遗传算法的影像中心布局优化方法[J]. 系统仿真学报, 2022, 34(6): 1173-1184.
[8]	段绍米, 罗会龙, 刘海鹏. 人群搜索和樽海鞘群的混合算法优化PID参数[J]. 系统仿真学报, 2022, 34(6): 1230-1246.
[9]	郭宇飞, 赵康, 海永清. 面向有限元分析的三角网格布尔运算方法[J]. 系统仿真学报, 2022, 34(5): 1003-1014.
[10]	马立新, 程颍. 计及可中断负荷的园区综合能源系统优化调度[J]. 系统仿真学报, 2022, 34(4): 817-825.
[11]	李兆强, 张时雨. 基于快速RRT算法的三维路径规划算法研究[J]. 系统仿真学报, 2022, 34(3): 503-511.
[12]	赵静, 王鹏, 丁筱茜, 蒋国平, 徐丰羽, 孙雁飞. 基于观测器的四旋翼容错控制及仿真研究[J]. 系统仿真学报, 2022, 34(1): 1-10.
[13]	闫秀英, 党苗苗. 基于改进多目标粒子群算法的家庭用电时段优化[J]. 系统仿真学报, 2022, 34(1): 70-78.
[14]	王宁, 代冀阳, 应进, 李叶鼎, 鲁亮亮. 基于自适应扩展势场的多无人机航迹规划仿真[J]. 系统仿真学报, 2021, 33(9): 2147-2156.
[15]	黄星源, 李岩屹. 基于双Q学习算法的干扰资源分配策略[J]. 系统仿真学报, 2021, 33(8): 1801-1808.