证据独立性及其对目标识别的影响研究

doi:10.12305/j.issn.1001-506X.2022.01.24

摘要/Abstract

摘要：

证据理论在不确定信息的表达和处理上有独特的优势, 而证据独立是很多组合方法应用的前提, 但目前没有完全有效的判断方法。针对这一问题, 首先研究了证据组合的内涵和证据对识别框架的解释。然后从Smets证据独立性定义出发, 分析了Dempster组合规则中的独立性问题, 并针对小样本或无样本情形, 提出了一种证据独立性判断方法。同时，依据本方法对Zadeh悖论进行了分析和讨论。最后, 针对多传感器单目标识别问题, 研究了证据独立性对组合结果的影响。

关键词: 信息融合, 目标识别, 证据理论, 证据独立性, 相关证据

Abstract:

Evidence theory has unique advantages in the expression and processing of uncertain information, and evidence independence is the premise of the application of many combination methods, but there is no completely effective judgment method at present. To solve this problem, firstly, the connotation of evidence combination and the interpretation of evidence to the recognition framework are studied. Then, starting from the definition of Smets evidence independence, this paper analyzes the independence problem in Dempster combination rules, and puts forward a judgment method of evidence independence for the case of small samples or no samples. Finally, aiming at the problem of multi-sensor single target recognition, the influence of evidence independence on the combination results is studied.

Key words: information fusion, target recognition, evidence theory, evidence independence, relevant evidence

中图分类号:

TN95
C934

关欣, 于昊天, 衣晓. 证据独立性及其对目标识别的影响研究[J]. 系统工程与电子技术, 2022, 44(1): 192-198.

Xin GUAN, Haotian YU, Xiao YI. Research on evidence independence and its influence on target recognition[J]. Systems Engineering and Electronics, 2022, 44(1): 192-198.

图/表 6

图1

图2

表1

表2

表3

表4

参考文献 29

1	DEMPSTER A P . Upper and lower probabilities induced by a multivalued mapping[J]. The Annals of Mathematical Statistics, 1967, 38 (2): 325- 339. doi: 10.1214/aoms/1177698950
2	GLENN S . A mathematical theory of evidence[M]. Princeton: Princeton University Press, 1976.
3	DEZERT J, SMARANDACHE F. DSmT: a new paradigm shift for information fusion[C]//Proc. of the Fusion International Conference, 2004.
4	SMARANDECHE F , DEZERT J . Advances and applications of DSmT for information fusion (collected works) Vol.Ⅱ[M]. Rehoboth: American Research Press, 2006.
5	SMETS P. Combination of nondistinct evidences[C]//Proc. of the North American Fuzzy Information Processing, 1986.
6	SMETS P. The concept of distinct evidence[C]//Proc. of the 4th Conference on Information Processing and Management of Uncertainty in Knowledge-Based Systems, 1992: 789-794.
7	DEZERT J , SMARANDACHE F . An introduction to DSmT, advances and applications of DSmT for information fusion (collected works) Vol.Ⅲ[M]. Rehoboth: American Research Press, 2009.
8	FERSON S , NELSEN R B , HAJAGOS J , et al. Dependence in probabilistic modeling, Dempster-Shafer theory, and probability bounds analysis[M]. New Mexico: Sandia National Laboratories, 2004.
9	付耀文, 杨威, 庄钊文. 证据建模研究综述[J]. 系统工程与电子技术, 2013, 35 (6): 1160- 1167.
	FU Y W , YANG W , ZHUANG Z W . Review on evidence mo-deling[J]. Systems Engineering and Electronics, 2013, 35 (6): 1160- 1167.
10	FUNG R, CHONG C Y. Metaprobability and Dempster-Shafer in evidential reasoning[C]//Proc. of the 1st Conference on Uncertainty in Artificial Intelligence(UAI85), 1985.
11	VOORBRAAK F . On the justification of Dempster's rule of combination[J]. Artificial Intelligence, 1991, 48 (2): 171- 197. doi: 10.1016/0004-3702(91)90060-W
12	ALTINCAY H . On the independence requirement in Dempster-Shafer theory for combining classifiers providing statistical evidence[J]. Applied Intelligence, 2006, 25 (1): 73- 90. doi: 10.1007/s10489-006-8867-y
13	ZADEH L A . On the validity of Dempster's rule of combination[J]. Research Gate, 1979, 79 (24): 3- 28.
14	肖文, 王正友, 王耀德. 一种相关证据的合成规则[J]. 控制与决策, 2011, 26 (5): 773- 776.
	XIAO W , WANG Z Y , WANG Y D . Combination rule for depen-dent evidences[J]. Control and Decision, 2011, 26 (5): 773- 776.
15	DENOEUX T . Conjunctive and disjunctive combination of belief functions induced by nondistict bodies of evidence[J]. Artificial Intelligence, 2008, 172 (2): 234- 264.
16	KALLEL A , HÉGARAT M S L . Combination of partially non-distinct beliefs in the cautious-adaptive rule[J]. International Journal of Approximate Reasoning, 2009, 50 (7): 1000- 1021. doi: 10.1016/j.ijar.2009.03.006
17	CHEBBAH M , MARTIN A , YAGHLANE B B . Combining partially independent belief functions[J]. Decision Support Systems, 2015, 73, 37- 46. doi: 10.1016/j.dss.2015.02.017
18	CHOENNI S, BLOK H E, LEERTOUWER E. Handling uncertainty and ignorance in databases: a rule to combine depen-dent data[C]//Proc. of the 11th International Conference on Database Systems for Advanced Applications, 2006: 310-324.
19	苏晓燕, 韩文花, 许培达, 等. 相关证据研究综述[J]. 系统工程与电子技术, 2016, 38 (6): 1345- 1351.
	SU X Y , HAN W H , XU P D , et al. Review of combining dependent evidence[J]. Systems Engineering and Electronics, 2016, 38 (6): 1345- 1351.
20	蒋雯, 彭进业, 邓勇. 一种新的证据冲突表示方法[J]. 系统工程与电子技术, 2010, 32 (3): 562- 565.
	JIANG W , PENG J Y , DENG Y . New representation method of evidential conflict[J]. Systems Engineering and Electronics, 2010, 32 (3): 562- 565.
21	赵静, 关欣, 刘海桥. 冲突证据决策新方法及应用[J]. 北京航空航天大学学报, 2019, 4 (9): 1838- 1847.
	ZHAO J , GUAN X , LIU H Q . A new conflict evidence decision method and its application[J]. Journal of Beijing University of Aeronautics and Astronsutics, 2019, 45 (9): 1838- 1847.
22	赵静, 关欣, 衣晓, 等. 一种新的解决冲突问题的不确定性度量方法[J]. 控制与决策, 2020, 35 (6): 1307- 1315.
	ZHAO J , GUAN X , YI X , et al. A new uncertainty measurement method for conflict problem[J]. Control and Decision, 2020, 35 (6): 1307- 1315.
23	张燕君, 龙呈. 基于证据理论的目标识别方法[J]. 系统工程与电子技术, 2013, 35 (12): 2467- 2470. doi: 10.3969/j.issn.1001-506X.2013.12.04
	ZHAN Y J , LONG C . Target recognition method based on evidence theory[J]. Systems Engineering and Electronics, 2013, 35 (12): 2467- 2470. doi: 10.3969/j.issn.1001-506X.2013.12.04
24	胡丽芳, 关欣, 何友. 基于DSmT的多传感器目标识别[J]. 弹箭与制导学报, 2008, 28 (2): 186- 188. doi: 10.3969/j.issn.1673-9728.2008.02.058
	HU L F , GUAN X , HE Y . Multi-sensor target identification based on DSmT[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2008, 28 (2): 186- 188. doi: 10.3969/j.issn.1673-9728.2008.02.058
25	GUAN X , LIU H Q , YI X , et al. The improved combination rule of D numbers and its application in radiation source identification[J]. Mathematical Problems in Engineering, 2018, 2018, 1- 10.
26	胡丽芳, 关欣, 何友. 基于DSmT的递归目标识别融合方法[J]. 控制理论与应用, 2012, 29 (1): 79- 84.
	HU L F , GUAN X , HE Y . Recursive target identification fusion methods based on Dezert-Smarandache theory[J]. Control Theory & Applications, 2012, 29 (1): 79- 84.
27	何友, 王国宏, 关欣, 等. 信息融合理论及应用[M]. 北京: 电子工业出版社, 2010.
	HE Y , WANG G H , GUAN X , et al. Information fusion theory with applications[M]. Beijing: Publishing House of Electronics Industry, 2010.
28	DEZERT J . Foundations for a new theory of plausible and para-doxical reasoning[J]. Information & Security: An International Journal, 2002, 9 (1): 13- 57.
29	SMETS P , MAMDANI E , DUBOIS D , et al. Non-standard logics for automated reasoning[M]. London: Academic Press, 1988.

传感器	属性
传感器	C₁	C₂	C₃
A	x_A	x_A	-
B	-	x_B	x_B

传感器	属性
传感器	C₁	C₃
A	x_A	-
B	-	x_B

相关性	识别概率
独立	x_Ax_B
混合相关	(x_Ax_B, min{x_A, x_B})或$\left(x_{A} x_{B}, \frac{\left(x_{A}+x_{B}\right)}{2}\right) $
完全相关	$x_{2}=\frac{\left(x_{A}+x_{B}\right)}{2} $或[min{x_A, x_B}, max{x_A, x_B}]

传感器	属性
传感器	C₁	C₂	C₃
A	m_A=0.6θ+0.2θ+0.2U	m_A=0.6θ+0.2θ+0.2U	-
B	-	m_B=0.65θ+0.1θ+0.25U	m_B=0.65θ+0.1θ+0.25U

[1]	王彩云, 吴钇达, 王佳宁, 马璐, 赵焕玥. 基于改进的CNN和数据增强的SAR目标识别[J]. 系统工程与电子技术, 2022, 44(8): 2483-2487.
[2]	刘康, 何明浩, 韩俊, 王庚. 基于本征向量和Jousselme距离的高冲突证据融合方法[J]. 系统工程与电子技术, 2022, 44(7): 2175-2180.
[3]	孙晶明, 虞盛康, 孙俊. 基于元学习的雷达小样本目标识别方法及改进[J]. 系统工程与电子技术, 2022, 44(6): 1839-1845.
[4]	黄林, 龚立, 姜伟, 王康勃. 基于多源信息融合与HMM的剩余寿命预测[J]. 系统工程与电子技术, 2022, 44(5): 1747-1756.
[5]	周晓玲, 张朝霞, 鲁雅, 王倩, 王琨琨. 基于改进R-FCN的SAR图像识别[J]. 系统工程与电子技术, 2022, 44(4): 1202-1209.
[6]	孙晶明, 虞盛康, 孙俊. 基于深度学习的HRRP识别姿态敏感性分析[J]. 系统工程与电子技术, 2022, 44(3): 802-807.
[7]	王旋, 狄鹏, 尹东亮. 基于Lance距离和信度熵的冲突证据融合方法[J]. 系统工程与电子技术, 2022, 44(2): 592-602.
[8]	雷禹, 冷祥光, 周晓艳, 孙忠镇, 计科峰. 基于改进ResNet网络的复数SAR图像舰船目标识别方法[J]. 系统工程与电子技术, 2022, 44(12): 3652-3660.
[9]	薛春岭, 曹菲, 孙庆, 秦建强, 冯晓伟. 基于多特征信息融合的海面微弱目标检测[J]. 系统工程与电子技术, 2022, 44(11): 3338-3345.
[10]	刘旗, 张新禹, 刘永祥. 基于门控多尺度匹配网络的小样本SAR目标识别[J]. 系统工程与电子技术, 2022, 44(11): 3346-3356.
[11]	张瑜, 吴凯, 郭杰, 葛志闪, 张宝超. 基于数据质量评估的自适应序贯航迹关联算法[J]. 系统工程与电子技术, 2022, 44(11): 3477-3485.
[12]	王彩云, 姚晨, 吴钇达, 王佳宁, 李晓飞, 黄盼盼. 基于改进Dijkstra算法与时频域滤波的雷达目标识别[J]. 系统工程与电子技术, 2022, 44(10): 3090-3095.
[13]	李永祯, 刘业民, 庞晨, 黄大通. 基于分层极化特性的箔条云识别方法[J]. 系统工程与电子技术, 2021, 43(8): 2099-2107.
[14]	谢春思, 刘志赢, 桑雨. 基于特征匹配的舰载对陆导弹目标识别模型[J]. 系统工程与电子技术, 2021, 43(8): 2244-2253.
[15]	晏媛, 孙俊, 孙晶明, 于俊朋. 雷达小样本目标识别方法及应用分析[J]. 系统工程与电子技术, 2021, 43(3): 684-692.