多传感器编队目标跟踪技术

本书是关于多传感器编队目标跟踪方法的一部专著，是作者们对国内外近30年来该领域研究进展和自身研究成果的总结。全书由6章组成，主要内容有：基础知识概述，编队目标航迹起始方法，复杂背景下集中式多传感器编队目标跟踪方法，集中式多传感器机动编队目标跟踪方法，系统误差下编队目标航迹关联方法，建议与展望。

博士，海军航空工程学院信息融合研究所综合研究室副主任兼院士秘书、讲师。研究领域为多传感器多目标跟踪、航迹关联、大数据技术等。作为课题组长或技术总师承担国家自然基金、总装预研基金等多项，发表学术论文多项。获山东省优秀科技成果创新奖和海军优秀硕士学位论文奖。

第1章 绪 论······································································································ 1

1.1 研究背景········································································································· 1

1.2 国内外研究现状····························································································· 2

1.2.1 航迹起始····························································································· 2

1.2.2 航迹维持····························································································· 3

1.2.3 机动跟踪····························································································· 3

1.3 多传感器编队目标跟踪技术中有待解决的一些关键问题························· 4

1.3.1 杂波环境下编队目标航迹起始技术················································ 4

1.3.2 复杂环境下集中式多传感器编队目标跟踪技术···························· 5

1.3.3 集中式多传感器机动编队目标跟踪技术········································ 5

1.3.4 系统误差下编队目标航迹关联技术················································ 6

1.4 本书的主要内容及安排················································································· 7

第2章 编队目标航迹起始算法·········································································· 8

2.1 引言················································································································· 8

2.2 基于相对位置矢量的编队目标灰色航迹起始算法····································· 8

2.2.1 基于循环阈值模型的编队预分割·················································· 10

2.2.2 基于编队中心点的预互联······························································ 11

2.2.3 RPV-FTGTI 算法············································································· 12

2.2.4 编队内目标航迹的确认·································································· 18

2.2.5 编队目标状态矩阵的建立······························································ 19

2.2.6 仿真比较与分析·············································································· 20

2.2.7 讨论··································································································· 34

2.3 集中式多传感器编队目标灰色航迹起始算法················································ 35

2.3.1 多传感器编队目标航迹起始框架·················································· 35

2.3.2 多传感器预互联编队内杂波的剔除·············································· 36

2.3.3 多传感器编队内量测合并模型······················································ 37

2.3.4 航迹得分模型的建立······································································ 38

2.4 基于运动状态的集中式多传感器编队目标航迹起始算法························40

多传感器编队目标跟踪

·VIII·

2.4.1 同状态航迹子编队获取模型·························································· 40

2.4.2 多传感器同状态编队关联模型······················································ 45

2.4.3 编队内航迹精确关联合并模型······················································ 45

2.5 仿真比较与分析··························································································· 46

2.5.1 仿真环境··························································································· 47

2.5.2 仿真结果及分析·············································································· 47

2.6 本章小结······································································································· 54

第3章 复杂背景下集中式多传感器编队目标跟踪算法································· 56

3.1 引言··············································································································· 56

3.2 系统描述······································································································· 56

3.3 云雨杂波和带状干扰剔除模型··································································· 57

3.3.1 云雨杂波剔除模型·········································································· 58

3.3.2 带状干扰剔除模型·········································································· 60

3.3.3 验证分析··························································································· 61

3.4 基于模板匹配的集中式多传感器编队目标跟踪算法······························· 63

3.4.1 基于编队整体的预互联·································································· 63

3.4.2 模板匹配模型的建立······································································ 65

3.4.3 编队内航迹的状态更新·································································· 69

3.4.4 讨论··································································································· 69

3.5 基于形状方位描述符的集中式多传感器编队目标粒子滤波算法··········· 69

3.5.1 编队目标形状矢量的建立······························································ 70

3.5.2 相似度模型的建立·········································································· 72

3.5.3 冗余图像的剔除·············································································· 74

3.5.4 基于粒子滤波的状态更新······························································ 74

3.6 仿真比较与分析··························································································· 75

3.6.1 仿真环境··························································································· 75

3.6.2 仿真结果··························································································· 76

3.6.3 仿真分析··························································································· 78

3.7 本章小结······································································································· 79

第4章 集中式多传感器机动编队目标跟踪算法············································· 81

4.1 引言··············································································································· 81

4.2 典型机动编队目标跟踪模型的建立··························································· 82

目 录

·IX·

4.2.1 编队整体机动跟踪模型的建立······················································ 82

4.2.2 编队分裂跟踪模型的建立······························································ 85

4.2.3 编队合并跟踪模型的建立······························································ 87

4.2.4 编队分散跟踪模型的建立······························································ 89

4.3 变结构JPDA机动编队目标跟踪算法······················································· 91

4.3.1 事件的定义······················································································· 92

4.3.2 编队确认矩阵的建立······································································ 93

4.3.3 编队互联矩阵的建立······································································ 93

4.3.4 编队确认矩阵的拆分······································································ 95

4.3.5 概率的计算······················································································· 97

4.3.6 编队内航迹的状态更新································································ 100

4.4 扩展广义S-维分配机动编队目标跟踪算法············································ 101

4.4.1 基本模型的建立············································································ 102

4.4.2 编队量测的划分············································································ 103

4.4.3 3-维分配问题的构造····································································· 106

4.4.4 广义S-维分配问题的构造···························································· 107

4.4.5 编队内航迹的状态更新································································ 107

4.5 仿真比较与分析························································································· 108

4.5.1 仿真环境························································································· 108

4.5.2 仿真结果························································································· 110

4.5.3 仿真分析························································································· 113

4.6 本章小结····································································································· 114

第5章 系统误差下编队目标航迹关联算法·················································· 116

5.1 引言············································································································· 116

5.2 系统误差下基于双重模糊拓扑的编队目标航迹关联算法····················· 116

5.2.1 基于循环阈值模型的编队航迹识别············································ 117

5.2.2 第一重模糊拓扑关联模型···························································· 118

5.2.3 第二重模糊拓扑关联模型···························································· 123

5.3 系统误差下基于误差补偿的编队目标航迹关联算法····························· 125

5.3.1 编队航迹状态识别模型································································ 125

5.3.2 编队航迹系统误差估计模型························································ 127

5.3.3 误差补偿和编队内航迹的精确关联············································ 130

5.3.4 讨论································································································· 130

多传感器编队目标跟踪

·X·

5.4 仿真比较与分析························································································· 131

5.4.1 仿真环境························································································· 131

5.4.2 仿真结果及分析············································································ 132

5.5 本章小结····································································································· 134

第6章 结论及展望·························································································· 135

附录A 式(2-17)中阈值参数ε 的推导··························································· 140

附录B 式(5-19)的推导····················································································· 144

参考文献·············································································································· 148

CONTENTS

Chapter 1 Introduction···························································································· 1

1.1 Background of Research··············································································· 1

1.2 Internal and Oversea Research Actualities ··················································· 2

1.2.1 Track Initiation ·················································································· 2

1.2.2 Track Maintenance ············································································ 3

1.2.3 Maneuvering Tracking ······································································ 3

1.3 The Key Problem to Be Resolved in Multi-sensor Formation Targets

Tracking Technique ········································································································ 4

1.3.1 Formation Targets Track Initiation Technique with Clutter·············· 4

1.3.2 Centralized Multi-sensor Formation Targets Tracking Technique

with the Complicated Background ········································································ 5

1.3.3 Centralized Multi-sensor Maneuvering Formation Targets Tracking

Technique ··············································································································· 5

1.3.4 Track Correlation Technique of the Formation Targets with

Systematic Errors ··································································································· 6

1.4 Main Content and Arragement of Dissertation············································· 7

Chapter 2 Formation Targets Track Initiation Algorithm ······································· 8

2.1 Introduction··································································································· 8

2.2 Formation Targets Gray Track Initiation Algorithm Based on Relative

Position Vector················································································································ 8

2.2.1 Preparative Division of the Formation Targets Based on the

Circulatory Threshold Model··············································································· 10

2.2.2 Preparative Association Based on the Formation Center················ 11

2.2.3 RPV-FTGTI Algorithm ··································································· 12

2.2.4 Validation of the Tracks in the Formation······································· 18

2.2.5 Establishment of the Formation Target State Matrix ······················ 19

2.2.6 Simulation Comparision and Analysis············································ 20

2.2.7 Discussion ······················································································· 34

2.3 Centralized Multi-sensor Formation Targets Gray Track Initiation

Algorithm ····················································································································· 35

2.3.1 Multi-sensor Formation Targets Track Initiation Frame ················· 35

2.3.2 Multi-sensor Clutter Deletion in Preparative Associated

多传感器编队目标跟踪

·XII·

Formations ··········································································································· 36

2.3.3 Multi-sensor Measurement Mergence Model in the Formation ····· 37

2.3.4 Establishment of the Track Score Model ········································ 38

2.4 Centralized Multi-sensor Formation Targets Track Initiation Algorithm

Based on Moving State································································································· 40

2.4.1 Same-state Track SubFormation Obtainment Model······················ 40

2.4.2 Multi-sensor Same-state Formation Association Model················· 45

2.4.3 Accurate Association and Mergence Model of the Formation

Tracks··················································································································· 45

2.5 Simulation Comparision and Analysis························································ 46

2.5.1 Simulation Envirenment··································································· 47

2.5.2 Simulation Results and Analysis ······················································ 47

2.6 Summary····································································································· 54

Chapter 3 Centralized Multi-sensor Formation Targets Tracking Algorithm with the

Complicated Background ····························································································· 56

3.1 Introduction································································································· 56

3.2 System Description ····················································································· 56

3.3 Deletion Models of the Cloud-rain Clutter and the Narrow-Band

Interference··················································································································· 57

3.3.1 Cloud-rain Clutter Deletion Model ·················································· 58

3.3.2 Narrow-Band Interference Deletion Model ····································· 60

3.3.3 Validation and Analysis ···································································· 61

3.4 Centralized Multi-sensor Formation Targets Tracking Algorithm Based on

Template Matching······································································································· 63

3.4.1 Preparative Association Based on the Whole Formation ················· 63

3.4.2 Establishment of the Template Matching Model ····························· 65

3.4.3 State Update of the Tracks in the Formation···································· 69

3.4.4 Discussion························································································· 69

3.5 Centralized Multi-sensor Formation Targets Particle Filter Based on Shape

and Azimuth Descriptor································································································ 69

3.5.1 Establishment of the Formation Targets Shape Vector····················· 70

3.5.2 Establishment of the Resemble Model············································· 72

3.5.3 Deletion of the Redundant Picture ··················································· 74

3.5.4 State Update Based on Particle Filter··············································· 74

CONTENTS

·XIII·

3.6 Simulation Comparision and Analysis························································ 75

3.6.1 Simulation Envirenment··································································· 75

3.6.2 Simulation Results············································································ 76

3.6.3 Simulation Analysis·········································································· 78

3.7 Summary····································································································· 79

Chapter 4 Centralized Multi-sensor Maneuvering Formation Targets Tracking

Algorithm ····················································································································· 81

4.1 Introduction································································································· 81

4.2 Establishment of Typical Maneuvering Formation Targets Tracking

Models ·························································································································· 82

4.2.1 Establishment of the Formation Whole Maneuver Tracking

Model ··················································································································· 82

4.2.2 Establishment of the Formation Splitting Tracking Model·············· 85

4.2.3 Establishment of the Formation merging Tracking Model ·············· 87

4.2.4 Establishment of the Formation dispersing Tracking Model ··········· 89

4.3 Maneuvering Formation Targets Tracking Algorithm Based on Different

Structure JPDA Technique···························································································· 91

4.3.1 Event Definition ··············································································· 92

4.3.2 Establishment of the Formation Validation Matrix ·························· 93

4.3.3 Establishment of the Formation Association Matrix························ 93

4.3.4 Splitting of the Formation Validation Matrix ··································· 95

4.3.5 Calculation of the Probability··························································· 97

4.3.6 State Update of the Tracks in the Formation·································· 100

4.4 Maneuvering Formation Targets Tracking Algorithm Based on Patulous

Generalized S-D Assignment Technique···································································· 101

4.4.1 Establishment of the Basic Model·················································· 102

4.4.2 Partition of the Measurements of the Formation Targets ··············· 103

4.4.3 Conformation of 3-D Assignment Problem ··································· 106

4.4.4 Conformation of Generalized S-D Assignment Problem ··········· 107

4.4.5 State Update of the Tracks in the Formation·································· 107

4.5 Simulation Comparision and Analysis······················································ 108

4.5.1 Simulation Envirenment································································· 108

4.5.2 Simulation Results·········································································· 110

4.5.3 Simulation Analysis········································································ 113

多传感器编队目标跟踪

·XIV·

4.6 Summary··································································································· 114

Chapter 5 Formation Targets Track Correlation Algorithm with Systematic

Errors ···························································································································116

5.1 Introduction······························································································· 116

5.2 Formation Targets Track Correlation Algorithm with Systematic Errors

Based on Double Fussy Topology·············································································· 116

5.2.1 Formation Tracks Identification Based on Circulatory Threshold

Model ················································································································· 117

5.2.2 The First Scale Fussy Topology Model·········································· 118

5.2.3 The Second Scale Fussy Topology Model ····································· 123

5.3 Formation Targets Track Correlation Algorithm with Systematic Errors

Based on Error Compensation···················································································· 125

5.3.1 Formation Track State Identification Model ·································· 125

5.3.2 Formation Track Systematic Error Estimation Model ··················· 127

5.3.3 Error Compensation and Formation Track Accurate

Correlation ········································································································· 130

5.3.4 Discussion······················································································· 130

5.4 Simulation Comparision and Analysis······················································ 131

5.4.1 Simulation Envirenment································································· 131

5.4.2 Simulation Results and Analysis ···················································· 132

5.5 Summary··································································································· 134

Chapter 6 Conclusions and Prospects ·································································· 135

Appendix A Illation of the Threshold Parameter ε in Formula (2-17) ············ 140

Appendix B Illation of Formula (5-19)····························································· 144

References············································································································ 148