结合ANN及改进MOPSO的天线布局多目标优化

doi:10.12305/j.issn.1001-506X.2026.05.17

摘要/Abstract

摘要：

针对车载多天线布局复杂度高、目标冲突性强等问题，基于神经网络及改进多目标粒子群优化（multi-objective particle swarm optimization，MOPSO）算法提出一种综合智能优化方法。以天线耦合度、方向图畸变度和驻波比为优化目标，以天线坐标与输入阻抗为决策变量，通过径向基神经网络快速预测天线性能，规避传统全波仿真的计算瓶颈。引入自适应网格算法及动态变异机制增强MOPSO算法的全局搜索能力，高效逼近Pareto最优前沿。进而，基于模糊集合理论对Pareto 最优解集进行选优。实验表明，相较于非支配排序遗传算法II等，所提方法对各目标函数的综合优化性能最佳，分别改善了23.2%、20.6%和5.8%，证明了其有效性和优越性。

关键词: 天线布局, 人工神经网络, 多目标优化, 粒子群算法, 模糊集合理论

Abstract:

To address the challenges of high complexity and strong conflicting objectives in vehicular multi-antenna placement, an integrated intelligent optimization method is proposed based on neural networks and improved multi-objective particle swarm optimization （MOPSO） algorithm. Take the antenna coupling degree, radiation pattern distortion degree, and standing wave ratio as optimization objectives, and the antenna coordinates and input impedances as decision variables. utilize the radial basis neural network to predict antenna performance rapidly, thereby circumventing the computational bottlenecks of traditional full-wave simulation. The MOPSO algorithm is improved through adaptive grid algorithm and dynamic mutation mechanism to strengthen the global search capability, and approximate the Pareto optimal frontier efficiently. Furthermore, the Pareto optimal solution set is prioritized with fuzzy set theory. Experimental demonstrate that, compared to algorithms such as non-dominated sorting genetic algorithm-II, the proposed method achieves the best comprehensive optimization performance for all objective functions, with improvements of 23.2%, 20.6%, and 5.8%, respectively, which validates its effectiveness and superiority.

Key words: antenna placement, artificial neural network, multi-objective optimization, particle swarm optimization, fuzzy set theory

中图分类号:

TM 743

刘璐瑶, 金晓, 蔡金良, 张武才. 结合ANN及改进MOPSO的天线布局多目标优化[J]. 系统工程与电子技术, 2026, 48(5): 1607-1614.

Luyao LIU, Xiao JIN, Jinliang CAI, Wucai ZHANG. Multi-objective optimization of antenna placement using integrated ANN and improved MOPSO[J]. Systems Engineering and Electronics, 2026, 48(5): 1607-1614.

图/表 12

图1

图2

图3

表1

图4

图5

表2

图6

图7

表3

表4

表5

参考文献 30

1	胡超. 基于ANP的车载通信系统电磁兼容性能综合评估[D]. 西安: 西安电子科技大学, 2009.
	HU C. Study on comprehensively evaluating systematic electromagnetic compatibility for vehicular communication system based on ANP[D]. Xi’an: Xidian University, 2009.
2	HANAFI A, KAMEL R A, ELHAK I, et al. Indoor antenna placement automation with meta-heuristic optimization[C]//Proc. of the International Conference on Machine Intelligence and Smart Innovation, 2024: 107−112.
3	马小姝, 李宇龙, 严浪. 传统多目标优化方法和多目标遗传算法的比较综述[J]. 电气传动自动化, 2010, 32 (3): 48- 50. doi: 10.3969/j.issn.1005-7277.2010.03.012
	MA X S, LI Y L, YAN L. Comparsion review of traditional multi-objective optimization methods and multi-objective genetic algorithm[J]. Electric Drive Automation, 2010, 32 (3): 48- 50. doi: 10.3969/j.issn.1005-7277.2010.03.012
4	XIE Z W, LI Y L, YANG S Y. A hybrid multi-objective optimization method and its application to electromagnetic device designs[J]. Applied Sciences, 2022, 12, 12110. doi: 10.3390/app122312110
5	ZHU D Z, WERNER P L, WERNER D H. Design and optimization of 3-D frequency-selective surfaces based on a multiobjective lazy ant colony optimization algorithm[J]. IEEE Trans. on Antennas and Propagation, 2017, 65 (12): 7137- 7149. doi: 10.1109/TAP.2017.2766660
6	MANSOR M A, SATHASIVAM S, KASIHMUDDIN M S M. Artificial immune system algorithm in VLSI circuit configuration[C]//Proc. of the 24th National Symposium on Mathematical Sceiences, 2017.
7	ZHENG Y Z, ZHOU L B, WANG J, et al. A neural network and NSGA-II based multi-objective optimization design method for permanent magnet synchronous machine[C]//Proc. of the 19th International Symposium on Electromagnetic Fields in Mechatronics, Electrical and Electronic Engineering, 2019.
8	GUO W, DONG Y Q. Enhancing energy-efficient building design: a multi-agent-assisted MOEA/D approach for multi-objective optimization[J]. Energy Informatics, 2024, 7, 102. doi: 10.1186/s42162-024-00406-3
9	ABIDO M A. Multiobjective evolutionary algorithms for electric power dispatch problem[J]. IEEE Trans. on Evolutionary Computation, 2006, 10 (3): 315- 329. doi: 10.1109/TEVC.2005.857073
10	KOPER E M, WOOD W D, SCHNEIDER S W. Aircraft antenna coupling minimization using genetic algorithms and approximations[J]. IEEE Trans. on Aerospace and Electronic Systems, 2004, 40 (2): 742- 751. doi: 10.1109/TAES.2004.1310019
11	丁雪. 基于遗传算法的车载天线布局优化[D]. 西安: 西安电子科技大学, 2014.
	DING X. Position optimal design of vehicular antennae via genetic algorithm[D]. Xi’an: Xidian University, 2014.
12	EMMETT C B, FLINT J A, SEAGER R D. Multiple objective optimization for antenna diversity on airborne platforms[J]. IET Science, Measurement & Technology, 2023, 17(1): 1−10.
13	HARIDAS V, VAITHEESWARAN S M, PRATHIKSHA M V. NSGA II Pareto optimal solutions for antenna locations on a commuter aircraft[C]//Proc. of the IEEE Space, Aerospace and Defence Conference, 2024: 1035−1039.
14	UZER A. Improvement of the diffraction coefficient of GTD by using multiplicative calculus[J]. IEEE Trans. on Antennas and Propagation, 2014, 62 (7): 3741- 3748. doi: 10.1109/TAP.2014.2319852
15	王伟. 车载多天线耦合效应及天线布局研究[D]. 西安: 西安电子科技大学, 2019.
	WANG W. Study on coupling effect and layout of vehicular multi-antennas[D]. Xi’an: Xidian University, 2019.
16	DEVABHAKTUNI V, BUNTING C F, GREEN D, et al. A new ANN-based modeling approach for rapid EMI/EMC analysis of PCB and shielding enclosures[J]. IEEE Trans. on Electromagnetic Compatibility, 2013, 55 (2): 385- 394. doi: 10.1109/TEMC.2012.2214223
17	BURGHARDT F, GARBE H. Comparison of predictions between artificial neural networks and Gaussian processes in EMC−investi gations[C]//Proc. of the International Symposium on Electromagnetic Compatibility, 2019: 921−926.
18	WU L, HE D P, AI B, et al. Artificial neural network based path loss prediction for wireless communication network[J]. IEEE Access, 2020, 8, 199523- 199538. doi: 10.1109/ACCESS.2020.3035209
19	SHI D, LV N, LIU Y, et al. The electromagnetic radiated emission prediction of PCB with the improved artificial neural network model[C]//Proc. of the IEEE International Symposium on Electromagnetic Compatibility & Signal/Power Integrity, 2017: 50−54.
20	XIE H Y, LIU Y, LI Y, et al. A prediction model based on artificial neural network for E1 HEMP coupling with distribution power lines[J]. IEEE Trans. on Power Delivery, 2022, 37 (6): 5337- 5344. doi: 10.1109/TPWRD.2022.3176118
21	MAO G P, ZHANG C, SHI K K, et al. Prediction of the performance and exhaust emissions of ethanol-diesel engine using different neural network[J]. Energy Sources, Part A: Recovery, Utilization, and Environmental Effects, 2022, 44 (2): 4949- 4963.
22	YU M X, ZHENG X P, ZHAO C H. Research on the prediction method of clock tester calibration data based on radial basis function neural network[J]. Electronics, 2023, 12 (22): 4677. doi: 10.3390/electronics12224677
23	XI L, LI L, LI L L, et al. Parameter optimization of titanium alloy considering energy efficiency and tool wear based on RBFNN-MOPSO algorithm in milling[J]. Journal of Manufacturing Processes, 2024, 122, 97- 111. doi: 10.1016/j.jmapro.2024.05.070
24	CHEN B H, WEN G J, HE X, et al. Application of adaptive grid-based multi-objective particle swarm optimization algorithm for directional drilling trajectory design[J]. Geoenergy Science and Engineering, 2023, 222, 211431. doi: 10.1016/j.geoen.2023.211431
25	KUO R J, GOSUMOLO M, ZULVIA F E. Multi-objective particle swarm optimization algorithm using adaptive archive grid for numerical association rule mining[J]. Neural Computing and Applications, 2019, 31, 3559- 3572. doi: 10.1007/s00521-017-3278-z
26	邹康格, 刘衍民. 基于自适应网格混合机制的多目标粒子群算法[J]. 重庆工商大学学报（自然科学版）, 2022, 39 (2): 14- 23. doi: 10.16055/j.issn.1672-058X.2022.0002.003
	ZOU K G, LIU Y M. Multi-objective particle swarm optimization based on adaptive mesh mixing mechanism[J]. Journal of Chongqing Technology and Business University （Natural Science Edition）, 2022, 39 (2): 14- 23. doi: 10.16055/j.issn.1672-058X.2022.0002.003
27	SHI S, DING W L, ZHAO Z, et al. Application and comparison of RBF and BP neural networks for lithology identification of Permian volcanic rocks in the Shunbei area of the Tarim Basin in China[J]. Journal of Earth System Science, 2024, 133, 181. doi: 10.1007/s12040-024-02352-w
28	UKOIMA K N, OKORO O A, OBI P I, et al. A modiﬁed multi-objective particle swarm optimization （M-MOPSO） for optimal sizing of a solar-wind-battery hybrid renewable energy system[J]. Solar Compass, 2024, 12, 100082. doi: 10.1016/j.solcom.2024.100082
29	GUO J Q, JIANG Z L, YING J L, et al. Optimal allocation model of port emergency resources based on the improved multi-objective particle swarm algorithm and TOPSIS method[J]. Marine Pollution Bulletin, 2024, 209, 117214. doi: 10.1016/j.marpolbul.2024.117214
30	杨超, 李以农, 郑玲, 等. 基于多目标粒子群算法的电磁主动悬架作动器优化[J]. 机械工程学报, 2019, 55 (19): 154- 166.
	YANG C, LI Y N, ZHENG L, et al. Optimum of electromagnetic active suspension actuator using multi-objective particle swarm optimization algorithm[J]. Journal of Mechanical Engineering, 2019, 55 (19): 154- 166.

坐标	最小值	最大值	坐标	最小值	最大值
$ {x}_{1} $	–1000	1000	$ {y}_{1} $	500	2 000
$ {x}_{\text{2}} $	0	1000	$ {y}_{2} $	–2500	–500
$ {x}_{\text{3}} $	–1000	–500	$ {y}_{3} $	–1000	0
$ {x}_{\text{4}} $	–1500	–500	$ {y}_{4} $	–2500	–1500

参数	取值
种群数量	50
最大迭代次数	100
初始惯性权重	0.8
最终惯性权重	0.2
外部存档数	250
个体经验系数 c₁	1.5
社会经验系数 c₂	1.5
网格等分数量	30
限定非劣解集的数量上限	20

时间	$ {x}_{1} $	$ {y}_{1} $	$ {z}_{1} $	$ {x}_{2} $	$ {y}_{2} $	$ {z}_{2} $	$ {x}_{3} $	$ {y}_{3} $	$ {z}_{3} $	$ {x}_{4} $	$ {y}_{4} $	$ {z}_{4} $
优化前	–300	1030	50	410	–1920	50	–640	–520	50	–940	–1640	50
优化后	–870	1390	44	390	–1530	51	–590	–300	49	–920	–1890	49

指标	NSGA-II	MOEA/D	MOPSO	改进MOPSO
天线耦合度	0.08	0.08	0.08	0.09
方向图畸变度	3.7	4.3	4.5	3.1
驻波比	1.2	1.3	1.3	1.1
HV	0.65	0.68	0.58	0.75
IGD	0.24	0.20	0.32	0.16
SP	0.15	0.18	0.22	0.11

算法	参数
标准MOPSO	惯性权重：0.5；局部速度/全局速度：1.5
NSGA-II	交叉概率：0.9，变异概率：1/12
MOEA/D	邻域大小：20；分解方法：Tchebycheff