基于改进灰狼算法优化SVR的航天侦察装备效能评估

doi:10.12305/j.issn.1001-506X.2021.10.25

摘要/Abstract

摘要：

定量评估航天侦察装备效能是武器装备体系建设的重要环节之一, 对装备发展和作战应用具有重要的现实意义。针对评估样本数据少、效能在多指标因素影响下变化规律非线性等条件下的效能评估问题, 提出一种基于改进灰狼(improved grey wolf optimizer, IGWO)算法优化的支持向量回归机(support vector regression, SVR)评估方法(IGWO-SVR)。引入反向学习策略及余弦非线性收敛因子改进灰狼优化算法收敛性能及全局寻优能力, 并将其应用于基于支持SVR效能评估参数的优化。基于航天侦察装备特点, 构建评估指标体系及航天侦察装备效能评估模型。最后, 通过对一定作战想定背景下航天侦察装备效能进行仿真评估, 验证了所提方法的合理性及优化模型的有效性。

关键词: 支持向量回归机, 效能评估, 航天侦察, 参数优化, 灰狼优化算法

Abstract:

Quantitative evaluation of operational effectiveness is an important part of reconnaissance satellite system (RSS) construction and has important practical significance for its development and combat application. In view of the problems such as small number of evaluation sample data and nonlinear change of performance under the influence of multiple index factors, a support vector regression (SVR) evaluation method based on improved grey wolf optimizer (IGWO) is proposed. Opposition-based learning strategy and cosine nonlinear convergence factor are introduced to improve the convergence performance and global optimization capability of GWO. IGWO algorithm is applied to the optimization of effectiveness evaluation parameters based on SVR. Based on the characteristics of RSS, the evaluation index system and effectiveness evaluation model are constructed. Finally, the validity of the model and the rationality of the method are verified by simulation evaluation of RSS under a certain operational scenario.

Key words: support vector regression machine, operational effectiveness evaluation, space reconnaissance, parameter optimization, grey wolf optimizer

中图分类号:

TJ86

韩驰, 熊伟. 基于改进灰狼算法优化SVR的航天侦察装备效能评估[J]. 系统工程与电子技术, 2021, 43(10): 2902-2910.

Chi HAN, Wei XIONG. Operational effectiveness evaluation of space reconnaissance equipment based on SVR optimized by improved grey wolf optimizer[J]. Systems Engineering and Electronics, 2021, 43(10): 2902-2910.

图/表 14

表1

表2

表3

表4

图1

图2

图3

图4

图5

表5

表6

图6

图7

表7

参考文献 33

1	徐一帆. 天基海洋移动目标监视的联合调度问题研究[D]. 长沙: 国防科学技术大学, 2011.
	XU Y F. Joint scheduling for space-based maritime moving targets surveillance[D]. Changsha: National University of Defense Technology, 2011.
2	崔鹏飞, 严洪森. 基于v-SVR的海洋环境下武器效能评估[J]. 计算机技术与发展, 2012, 22 (8): 20- 24.
	CUI P F , YAN H S . Operational effective-ness evaluation of weapon under influence of marine environment based on v-SVR[J]. Computer Technology and Development, 2012, 22 (8): 20- 24.
3	罗小明, 朱延雷, 何榕. 基于复杂适应系统的装备作战试验体系贡献度评估[J]. 装甲兵工程学院学报, 2015, 29 (2): 1- 6.
	LUO X M , ZHU Y L , HE R . Evaluation of contribution for equipment operational test system based on complex adaptive system[J]. Journal of Academy of Armored Force Engineering, 2015, 29 (2): 1- 6.
4	ZHENG L , ZHOU J W , HU H B , et al. Opera-tional effectiveness analysis of cluster submarine formation torpedo weapon system based on fuzzy AHP comprehensive evaluation[J]. ACM Trans.on Intelligent Systems and Technology, 2019, 2 (3): 17- 21.
5	ZHAO Q S, HU W T, XIA B Y, et al. The capability spaces complexity measure method of weapon system of systems[C]// Proc. of the IEEE International Systems Conference, 2020: 291-297.
6	SEBASTIAN M , JOSE M , MIRANDA J . IOWA-SVM: a density-based weighting strategy for SVM classification via OWA operators[J]. IEEE Trans.on Fuzzy Systems, 2020, 28 (9): 2143- 2150. doi: 10.1109/TFUZZ.2019.2930942
7	VAPNIK V N . The nature of statistical learning theory[M]. New York: Springer, 1995.
8	LEI J S , CHEN J F , CAO X . The regression prediction analysis of grouting concretion stone's strength based on SVR[J]. Advanced Materials Research, 2014, 29 (20): 171- 176.
9	LI S , FANG H , LIU X Y . Parameter optimization of support vector regression based on sine cosine algorithm[J]. Expert System, 2018, 91 (2): 63- 77.
10	黄发明, 殷坤龙, 张桂荣, 等. 多变量PSO-SVM模型预测滑坡地下水位[J]. 浙江大学学报(工学版), 2015, 49 (6): 1193- 1200.
	HUANG F M , YIN K L , ZHANG G R , et al. Prediction of groundwater level in landslide using multivariable PSO-SVM model[J]. Journal of Zhejiang University (Engineering Science), 2015, 49 (6): 1193- 1200.
11	曹庆奎, 赵斐. 基于遗传-支持向量回归的煤层底板突水量预测研究[J]. 煤炭学报, 2011, 36 (12): 2097- 2101.
	CAO Q K , ZHAO F . Forecast of water inrush quantity from coal floor based on genetic algorithm-support vector regression[J]. Journal of China Coal Society, 2011, 36 (12): 2097- 2101.
12	HATTA N M , ZAIN A M , SALLEHUDDIN R , et al. Recent studies on optimization method of grey wolf optimizer (GWO): a review[J]. Artificial Intelligence Review, 2018, 18 (2): 2651- 2683.
13	WANG X F , ZHAO H , HAN T , et al. A grey wolf optimizer using Gaussian estimation of distribution and its application in the multi-UAV multi-target urban tracking problem[J]. Applied Soft Computing, 2019, 78 (1): 240- 260.
14	TIZHOOSH H R. Opposition-based learning: a new scheme for machine intelligence[C]//Proc. of the IEEE International Confe-rence on Computational Intelligence for Modelling, Control & Automation, 2005: 695-701.
15	王田田, 王艳, 纪志成. 基于改进极限学习机的滚动轴承故障诊断[J]. 系统仿真学报, 2018, 30 (11): 4413- 4420.
	WANG T T , WANG Y , JI Z C . Fault diagnosis of rolling bearing based on improved extreme learning machine[J]. Journal of System Simulation, 2018, 30 (11): 4413- 4420.
16	魏雪, 吴清. 分段复合多尺度模糊熵和IGWO-SVM的脑电情感识别[J]. 计算机应用研究, 2019, 36 (11): 3310- 3314, 3356.
	WEI X , WU Q . EEG emotion recognition based on piecewise complex multi-scale fuzzy entropy and IGWO-SVM algorithm[J]. Application Research of Computers, 2019, 36 (11): 3310- 3314, 3356.
17	韩驰, 熊伟. 航天侦察装备体系指标关联信息挖掘研究[EB/OL]. [2020-11-02]. http://kns.cnki.net/kcms/detail/11.3092.V.20201030.1745.017.html.
	HAN C, XIONG W. Research on related information mining of space reconnaissance equipment system index[EB/OL]. [2020- 11-02]. http://kns.cnki.net/kcms/detail/11.3092.V.20201030.1745.017.html.
18	WANG Z , YANG C , OH S K , et al. Robust multi-linear fuzzy svr designed with the aid of fuzzy c-means clustering based on insensitive data information[J]. IEEE Access, 2020, 8, 184997- 185011. doi: 10.1109/ACCESS.2020.3030083
19	ZHU S , SUN J , LIU Y , et al. The air quality index trend forecasting based on improved error correction model and data preprocessing for 17 port cities in China[J]. Chemosphere, 2020, 252 (11): 126474.
20	BESSEDIK S A , DJEKIDEL R , AMEUR A . Performance of different kernel functions for LS-SVM-GWO to estimate flashover voltage of polluted insulators[J]. IET Science Measurement & Technology, 2018, 12 (6): 739- 745.
21	XU L W , WANG H , LIN W , et al. GWO-BP neural network based on performance prediction for mobile multiuser communi- cation networks[J]. IEEE Access, 2019, 7, 152690- 152700. doi: 10.1109/ACCESS.2019.2948475
22	ZHU S L , QIU X L , YIN Y R . Two-step-hybrid model based on data preprocessing and intelligent optimization algorithms (CS and GWO) for NO2 and SO2 forecasting[J]. Atmospheric Pollution Research, 2019, 10 (4): 1326- 1335. doi: 10.1016/j.apr.2019.03.004
23	SINGH N , SINGH S . A novel hybrid GWO-SCA approach for optimization problems[J]. Engineering Science and Technology, 2017, 20 (12): 1586- 1601.
24	LIU Y W, FAN S S, FENG Y, et al. Stockbridge damper identification of overhead power lines based on hog feature and GWO-SVM[C]//Proc. of the IEEE 3rd Conference on Energy Internet and Energy System Integration, 2019: 2466-2470.
25	林木, 李小波, 王彦锋, 等. 基于QFD和组合赋权TOPSIS的体系贡献率能效评估[J]. 系统工程与电子技术, 2019, 41 (8): 1802- 1809.
	LIN M , LI X B , WANG Y F , et al. Capability effectiveness evaluation of contribution ratio to system-of-systems based on QFD and combination weights TOPSIS[J]. Systems Engineering and Electronics, 2019, 41 (8): 1802- 1809.
26	YU Z , SHI X , ZHOU J , et al. Prediction of blast-induced rock movement during bench blasting: use of gray wolf optimizer and support vector regression[J]. Natural Resources Research, 2019, 29 (7): 843- 865.
27	LI W , ZHANG J . An innovated integrated model using singular spectrum analysis and support vector regression optimized by intelligent algorithm for rainfall forecasting[J]. Journal of Autonomous Intelligence, 2019, 2 (1): 46- 55. doi: 10.32629/jai.v2i1.37
28	LI B , LUO C Y , WANG Z Y . Application of GWO-SVM algorithm in arc detection of pantograph[J]. IEEE Access, 2020, 8, 173865- 173873.
29	ZHU A J , XU C P , LI Z , et al. Hybridizing grey wolf optimization with differential evolution for global optimization and test scheduling for 3D stacked SoC[J]. Journal of Systems Engineering and Electronics, 2015, 26 (2): 317- 328. doi: 10.1109/JSEE.2015.00037
30	HU Q H , ZHANG S G , YU M , et al. Short-term wind speed or power forecasting with heteroscedastic support vector regression[J]. IEEE Trans.on Sustainable Energy, 2016, 7 (1): 241- 249. doi: 10.1109/TSTE.2015.2480245
31	ȘENEL F A , GÖKÇE F , YVKSEL A S , et al. A novel hybrid PSO-GWO algorithm for optimization problems[J]. Engineering with Computers, 2018, 35 (3): 1359- 1373.
32	CHENG B , JIANG J , TAN Y J . A novel approach for WSoS capability requirement satisfactory degree evaluation using evidential reasoning[J]. System Engineering Theory and Practice, 2011, 31 (11): 2210- 2216.
33	LI X X , CHEN H G , LI D X , et al. Combat effectiveness prediction model using ELMAN feedback network[J]. Journal of System Simulation, 2015, 27 (1): 43- 49.

算法	描述
PSO	标准粒子群算法
GWO	GWO算法
IGWO_1	第2.2.1节策略IGWO算法
IGWO_2	第2.2.2节策略IGWO算法
IGWO	混合IGWO算法

参数	N	max	上限	下限
值	30	500	1e06	1e-03

测试函数	维度	范围	Min
${f_1} = \sum\nolimits_{i = 1}^n {x_i^2} $	30	[－100, 100]	0
${f_2} = \sum\nolimits_{i = 1}^n {\left\| {{x_i}} \right\|} + \prod\nolimits_{i = 1}^n {\left\| {{x_i}} \right\|} $	30	[－10, 10]	0
${f_3} = \sum\nolimits_{i = 1}^n {\left( {\sum\nolimits_{j = 1}^i {x_j^2} } \right)} $	30	[－100, 100]	0
${f_4} = \sum\nolimits_{i = 1}^n {\left[ {x_i^2 - 10\cos \left( {2\pi {x_i}} \right) + 10} \right]} $	30	[－5.12, 5.12]	0
${f_5} = - 20\exp \left[ { - \frac{1}{5}\sqrt {\left( {\sum\nolimits_{i = 1}^n {x_i^2} } \right)/n} } \right] - \exp \left[ {\left( {\sum\nolimits_{i = 1}^n {\cos \left( {2\pi {x_i}} \right)} } \right)/n} \right] + 20 + e$	30	[－32, 32]	0
${f_6} = \frac{1}{{4000}}\sum\limits_{i = 1}^n {x_i^2} - \prod\limits_{i = 1}^n {\cos \left( {\frac{{{x_i}}}{{\sqrt i }}} \right)} + 1$	30	[－600, 600]	0
${f_7} = 4x_1^2 - 2.1x_1^4 + x_1^6/3 + {x_1}{x_2} - 4x_2^2 + 4x_2^4$	2	[－5, 5]	-1.031 6
${f_8} = {\left( {{x_2} - \frac{{5.1}}{{4{\pi ^2}}}x_1^2 + \frac{5}{\pi }{x_1} - 6} \right)^2} + 10\left( {1 - \frac{1}{{8\pi }}} \right)\cos \;{x_1} + 10$	2	[－5, 5]	0.397 9
${f_9} = \left[ {1 + {{\left( {{x_1} + {x_2} + 1} \right)}^2} \cdot \left( {19 - 14{x_1} + 3x_1^2 - 14{x_2} + 6{x_1}{x_2} + 3x_2^2} \right)} \right] \cdot \left[ {30 + {{\left( {2{x_1} - 3{x_2}} \right)}^2} \cdot \left( {18 - 32{x_1} + 12x_1^2 + 48{x_2} - 36{x_1}{x_2} + 27x_2^2} \right)} \right]$	2	[－2, 2]	3

函数	PSO		GWO		IGWO_1		IGWO_2		IGWO
函数	Average	St.dev	Average	St.dev	Average	St.dev	Average	St.dev	Average	St.dev
f₁	3.67e-09	5.33e-09	3.79e-59	6.78e-59	1.02e-70	1.51e-70	8.47e-82	2.61e-81	7.72e-83	1.35e-82
f₂	1.59e-04	1.83e-04	8.74e-35	5.76e-35	2.98e-41	3.66e-41	3.96e-48	3.51e-48	2.07e-48	2.43e-48
f₃	14.304 2	4.827 6	4.21e-16	9.99e-16	1.04e-18	2.27e-18	4.31e-20	8.80e-20	9.45e-22	1.92e-21
f₄	49.748 9	10.103 0	0.442 0	1.397 7	5.68e-15	1.80e-14	0	0	0	0
f₅	1.45e-04	3.19e-04	1.76e-14	3.37e-15	1.23e-14	3.67e-15	7.64e-15	1.12e-15	7.99e-15	0
f₆	0.008 6	0.008 8	0.003 7	0.009 0	0.001 7	0.003 6	0.001 5	0.004 6	0	0
f₇	-1.031 6	4.24e-08	-1.031 6	7.32e-08	-1.031 6	1.30e-08	-1.031 6	0	-1.031 6	7.43e-09
f₈	0.397 9	7.72e-06	0.397 9	8.64e-06	0.397 9	2.08e-06	0.397 9	5.71e-07	0.397 9	0
f₉	3.000 0	5.41e-06	3.000 0	1.82e-06	3.000 0	1.52e-05	3.000 0	4.79e-06	3.000 0	1.26e-15

编号	半长轴/km	偏心率	轨道倾角/(°)	近地点辐角/(°)	升交点赤经/(°)	真近点角/(°)
LEO1-1/2/3	500	0	45.000 0	0	0	0/120/240
LEO2-4/5/6	500	0	45.109 2	0	89.889 8	30.16/150.16/270.16
LEO3-7/8/9	500	0	44.999 1	0	179.780 0	60.31/180.31/300.31
LEO4-10/11/12	500	0	44.889 7	0	269.891 0	90.16/210.16/330.16