基于深度表征学习和遗传算法的军用座舱色彩设计方法

doi:10.12382/bgxb.2023.0456

摘要/Abstract

摘要：

军用座舱色彩设计作为载人军事设备工业设计中较为主观的一部分,设计的合理性至关重要。为提高军用座舱色彩设计的科学性,提出一种基于深度表征学习和遗传算法的军用座舱配色方法。利用深度表征学习模型预测军用座舱配色方案,并根据色彩感知理论建立军用座舱配色模型,将其作为生成方案的限制条件。同时,引入交互式遗传算法到智能配色系统中,通过人工引导的方式优化神经网络的参数,对预测的配色方案进行有效迭代。实验结果表明:该方法生成的配色方案符合军用座舱配色模型,结合遗传算法的模型预测准确率比单一的深度表征模型提高了16%~18%。相较于人工色彩设计方案,军用座舱智能配色方法生成的方案满意度略优、设计周期缩短了80%~88%,色彩稳定性提高了6%~12%。

关键词: 军用座舱配色, 深度表征学习, 交互式遗传算法, 色彩感知

Abstract:

The color design of military cockpits is a subjective part of industrial design for manned military equipment, and the rationality of the design is crucial. A military cockpit color matching method based on deep representation learning and genetic algorithm is proposed to improve the scientific nature of military cockpit color design. This method uses a deep representation learning model to predict the military cockpit color schemes and establishes a military cockpit color matching model based on color perception theory, which serves as a constraint condition for generating schemes. At the same time, an interactive genetic algorithm is introduced into the intelligent color matching system to optimize the parameters of neural network through manual guidance, and effectively iterate the predicted color matching schemes. The results show that the color matching schemes generated by the proposed method comply with the military cockpit color matching model, and the prediction accuracy of the proposed model combined with genetic algorithm is improved by 16%~18% compared to a single deep representation model. Compared to manual color design schemes, the satisfaction of the color schemes generated by the military cockpit intelligent color matching method is slightly higher, the design cycle is shortened by 80%~88%, and the color stability is improved by 6%~12%.

Key words: military cockpit color matching, deep representation learning, interactive genetic algorithm, color perception

中图分类号:

TP183

苏胜, 顾森, 宋志强, 刘萍. 基于深度表征学习和遗传算法的军用座舱色彩设计方法[J]. 兵工学报, 2024, 45(4): 1060-1069.

SU Sheng, GU Sen, SONG Zhiqiang, LIU Ping. Military Cockpit Color Design Method Based on Deep Representation Learning and Genetic Algorithm[J]. Acta Armamentarii, 2024, 45(4): 1060-1069.

图/表 11

表1 军用舱室配色模型

Table 1 Military cockpit color matching model

色彩种类	色彩取值			色彩意象
色彩种类	H	S	B	色彩意象
皮肤色	0~280	0~91	5~88	深沉,安全,现代,理性
填充物色	16~350	2~86	25~100	和谐,舒适,大气,清爽
特殊色	18~357	3~87	2~87	人性,柔和,解压,轻快

图1 面向军用座舱配色的深度表征学习模型的架构图

Fig.1 Architecture diagram of deep representation learning model for military cockpit color matching

表2 算法关键参数表

Table 2 Key hyper-parameters of algorithm

算法关键参数	数值
迭代步数	50
学习率	0.01
潜层维度	256
随机失活率	0.5

图2 基于深度表征学习的军用座舱配色方案生成算法框架

Fig.2 Algorithm framework of military cockpit color matching scheme generation based on deep representation learning

图3 基于军用座舱配色方案优化的交互式遗传算法系统框架

Fig.3 Interactive genetic algorithm system framework based on military cockpit color scheme matching optimization

表3 配色方案预测准确率对比

Table 3 Comparison of predictive accuracies of color schemes%

算法	座舱型号
算法	S*5型	C/*5型	7*A型
深度表征模型	47	37	53
基于深度表征模型的遗传算法	64	55	69

表4 配色周期对比

Table 4 Comparison of color matching timesh

方法	座舱型号
方法	S*5型	C/*5型	7*A型
设计师配色	32	32	32
基于深度表征学习和遗传算法的军用座舱色彩设计方法	5.1	3.8	6.5

表5 配色方案稳定性对比

Table 5 Comparison of color scheme stabilities%

方案	座舱型号
方案	S*5型	C/*5型	7*A型
MEPCD模型生成方案	38	34	21
设计师配色设计方案	91	94	88
基于深度表征学习和遗传算法的军用座舱色彩设计方法生成方案	100	100	100

表6 配色方案满意度对比

Table 6 Comparison of color scheme satisfactions%

方案	座舱型号
方案	S*5型	C/*5型	7*A型
设计师设计方案	67	54	61
基于深度表征学习和遗传算法的军用座舱色彩设计方法生成方案	70	69	73

图4 载入针对性学习样本

Fig.4 Loading targeted learning samples

图5 某型多管火箭炮座舱配色预测

Fig.5 Prediction of color matching of a certain type of multi-barrel rocket launcher cockpit

参考文献 35

[1]	TAO W Y, WU Y, LI W F, et al. Influence of classroom colour environment on college students' emotions during campus lockdown in the COVID-19 post-pandemic era—a case study in Harbin, China[J]. Buildings, 2022, 12(11): 1873.
[2]	李中付, 余浩, 时粉周, 等. 船舶舱室色彩搭配对船员工效影响研究[J]. 船舶工程, 2014, 36(1):82-86.
	LI Z F, YU H, SHI F Z, et al. Researches on experimentation with influence of cabin color on crew's performance[J]. Ship Engineering, 2014, 36(1):87-91. (in Chinese)
[3]	黄璞. 色彩美学在船舶舱室设计中的应用[J]. 包装工程, 2022, 43(12):273-277 HUANG P.
	Application of color aesthetics in ship cabin design[J]. Packaging Engineering, 2022, 43(12):273-277. (in Chinese)
[4]	魏蓉, 宋志强, 苏胜. 基于色彩模型的车载指挥方舱设计[J]. 包装工程, 2020, 41(6):86-93.
	WEI R, SONG Z Q, SU S. Design of vehicle-mounted commend cabin based on color model[J]. Packaging Engineering, 2020, 41(6): 86-93. (in Chinese)
[5]	李中付, 丁猛, 余浩, 等. 基于色彩心理与工效的船舱室色彩搭配原则[J]. 船舶工程, 2015, 37(增刊1):203-205, 260.
	LI Z F, DING M, YU H, et al. Fundamental requirements of cabin color design with psychology of color vision and ergonomics[J]. Ship Engineering, 2015, 37(S1): 203-205, 260. (in Chinese)
[6]	朱昱宁, 徐博群, 刘肖健. 基于交互式遗传算法的参考图像辅助配色设计[J]. 包装工程, 2020, 41(2):181-188.
	ZHU Y N, XU B Q, LIU X J. Reference image aided color matching design based on interactive genetic algorithm[J]. Packaging Engineering, 2020, 41(2): 181-188. (in Chinese)
[7]	DING M, DONG W. Multiemotional product color design using gray theory and nondominated sorting genetic algorithm-Ⅲ[J]. Color Research and Application, 2020, 45(1): 142-155.
[8]	刘肖健, 曹愉静, 赵露唏. 传统纹样的色彩网络模型及配色设计辅助技术[J]. 计算机集成制造系统, 2016, 22(4): 897-905.
	LIU X J, CAO Y J, ZHAO L X. Color networks of traditional cultural patterns and color design aiding technology[J]. Computer Integrated Manufacturing Systems, 2016, 22(4): 897-905. (in Chinese)
[9]	刘肖健, 贾莹莹, 徐博群. 配色设计研究现状综述[J]. 包装工程, 2021, 42(14):92-100.
	LIU X J, JIA Y Y, XU B Q. Review on the study of color design[J]. Packaging Engineering, 2021, 42(14): 92-100. (in Chinese)
[10]	BITTERMANN M S. Insight into color aesthetics from probabilistic perception modeling[J]. Color Research & Application, 2018, 43(4): 527-543.
[11]	HU G S, LIU Z, WANG Y, et al. Calculative modeling for quantified semantics‐color mapping[J]. Color Research and Application, 2020, 45(3): 465-476.
[12]	DING M, BAI Z H. Product color emotional design adaptive to product shape feature variation[J]. Color Research & Application, 2019, 44(5): 811-823.
[13]	苏胜, 宋志强, 闫敏, 等. 军用座舱赋色原则研究[J]. 西安工业大学学报, 2022, 42(3):317-323.
	SU S, SONG Z Q, YAN M, et al. Research on and application of factor extraction model for color design of military cabin[J]. Journal of Xi'an Technological University, 2022, 42(3):317-323. (in Chinese)
[14]	苏胜, 宋志强, 顾森, 等. 军用舱室色彩感知下配色方法研究[J]. 西安工业大学学报, 2022, 42(1):96-104.
	SU S, SONG Z Q, GU S, et al. Color matching method based on color perception for military cabins[J]. Journal of Xi'an Technological University, 2022, 42(1): 96-104. (in Chinese)
[15]	LECUN Y, BENGIO Y, HINTON G. Deep learning[J]. Nature, 2015, 521(7553): 436-444.
[16]	GUO Y M, LIU Y, OERLEMANS A, et al. Deep learning for visual understanding: a review[J]. Neurocomputing, 2016, 187: 27-48.
[17]	HAO X, ZHANG G G, MA S. Deep learning[J]. International Journal of Semantic Computing, 2016, 10(3): 417-439.
[18]	张凤, 张超, 杨华民, 等. 作战辅助系统中基于深度学习的投影图像颜色补偿方法[J]. 兵工学报, 2021, 42(11):2418-2423. doi: 10.3969/j.issn.1000-1093.2021.11.015
	ZHANG F, ZHANG C, YANG H M, et al. Deep learning-based color compensation method for projected images in combat assistant system[J]. Acta Armamentarii, 2021, 42(11): 2418-2423. (in Chinese) doi: 10.3969/j.issn.1000-1093.2021.11.015
[19]	BAU D, LIU S, WANG T Z, et al. Rewriting a deep generative model[C]// Proceedings of the 16th European Conference on Computer Vision. Glasgow, UK: Springer, 2020: 351-369.
[20]	WARNES A. Saudi international airshow[J]. Airforces Monthly, 2019(374):30-35.
[21]	LIU F. CASC in Zhuhai airshow 2012[J]. Aerospace China, 2012, 13(4):2-5.
[22]	CHO S B. Towards creative evolutionary systems with interactive genetic algorithm[J]. Applied Intelligence, 2002, 16(2):129-138.
[23]	LEE J H, KIM H S, CHO S B. Accelerating evolution by direct manipulation for interactive fashion design[C]// Proceedings of the 4th International Conference on Computational Intelligence and Multimedia Applications. Yokusika, Japan: IEEE, 2001.
[24]	DAWKINS R. The evolution of evolvability[M]//Langton C. Artificial life. New York, NY, US: Routledge, 2019: 201-220.
[25]	JAKSA R, TAKAGI H. Tuning of image parameters by interactive evolutionary computation[C]// Proceedings of 2003 IEEE International Conference on Systems, Man and Cybernetics. Washington, DC, US: IEEE, 2003.
[26]	BAIK S, BALA J, HADJARIAN A, et al. Visualization for genetic evolution of target movement in battle fields[C]// Proceedings of the Computational Science - {ICCS} 2005. Atlanta, GA, US: Springer-Verlag, 2005.
[27]	BANDTE O, MALINCHIK S. A broad and narrow approach to interactive evolutionary design - an aircraft design example[C]// Proceedings of the Genetic and Evolutionary Computation - GECCO 2004. Seattle, WA, US: Springer-Verlag, 2004.
[28]	TAKAGI H. Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation[J]. Proceedings of the IEEE, 2001, 89(9):1275-1296.
[29]	KIM H S, CHO S B. Application of interactive genetic algorithm to fashion design[J]. Engineering Applications of Artificial Intelligence, 2000, 13(6):635-644.
[30]	CONGDON C B, MAZZA R H. GenTree: an interactive genetic algorithms system for designing 3D polygonal tree models[C]// Proceedings of the Genetic and Evolutionary Computation - GECCO 2003. Chicago, IL, US: Springer-Verlag, 2003.
[31]	FAROOQ H, ZAKARIA N, SIDDIQUE M T. An interactive visualization of genetic algorithm on 2-D graph[J]. International Journal of Software Science & Computational Intelligence, 2012, 4(1):34-54.
[32]	REN J, GONG D W, SUN X Y, et al. Interactive genetic algorithms with variational population size[C]// Proceedings of the 5th International Conference on Intelligent Computing. Ulsan, South Korea: Springer-Verlag, 2009.
[33]	DENG W, ZHANG X, ZHOU Y, et al. An enhanced fast non-dominated solution sorting genetic algorithm for multi-objective problems[J]. Information Sciences, 2022, 585: 441-453.
[34]	JALALI Z, NOORZAI E, HEIDARI S. Design and optimization of form and facade of an office building using the genetic algorithm[J]. Science and Technology for the Built Environment, 2020, 26(2): 128-140.
[35]	GIASSI M, GÖTEMAN M. Layout design of wave energy parks by a genetic algorithm[J]. Ocean Engineering, 2018, 154: 252-261.