基于雷达观测的目标跟踪与评估系统
发布时间：2021-01-23 17:11
多目标跟踪是指在传感器获取的带有噪声的测量基础上,建立被监测的多个目标的轨迹的过程。多目标跟踪对于军队以及民用监视具有重要的实际意义。多目标跟踪的战略目标是从获得的数据中追踪目标,这涉及到众多特别复杂的因素,例如跟踪目标的数量可变,传感器噪声,错误的警报以及数据关联的不确定性等。针对这些问题,本文选用概率假设密度滤波器作为核心方法,研究实现了以该滤波器为基础的多种状态估计算法。势概率假设密度滤波器也被认为是解决方案之一,但它比PHD滤波器在计算上更为复杂。同样地,我们也在本文中实现了针对该滤波器的状态估计。在概率假设密度滤波的实现方面,本文采用了粒子滤波实现的SMC-PHD滤波器,因为它能够与非线性和非高斯动力学联系在一起。之后,本文主要对四种状态估计方法进行了研究和对比。在多目标轨迹评估方面,本文使用OSPA以获得最佳轨迹集合距离,以评估两个轨迹集合的相似程度。最后,本文将上述算法研究集成于基于雷达传感器的目标跟踪演示系统中,以提供可视化的多目标跟踪结果和算法性能的多角度评估。
【文章来源】：哈尔滨工业大学黑龙江省 211工程院校 985工程院校
【文章页数】：79 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
Chapter 1 Introduction
1.1 Background
1.2 Literature review
1.2.1 Target tracking
1.2.2 Radar tracking
1.2.3 Tracking performance evaluation
1.3 Main content
Chapter 2 Multiple target trajectory evaluation
2.1 Hausdorff metric for Multi-Object trajectories
2.2 Optimal Mass Transfer （OMAT） metric
2.3 Evaluation metric （OSPA）
2.4 Chapter summary
Chapter 3 Multi-target tracking based on Random Finite Sets filtering
3.1 The Probability Hypothesis Density filter
3.2 Implementing of the PHD Filter
3.3 The Cardinalized PHD （CPHD） filter
3.3.1 Performing the CPHD estimation
3.3.2 Computational Complexity of the CPHD Filter
3.4 Simulation results and analysis
3.5 Chapter Summary
Chapter 4 Multi-target state estimation
4.1 Weight component of particles
4.2 Multi-target state estimation methods
4.2.1 ZHAO method
4.2.2 Ristic method
4.2.3 MEAP estimator
4.2.4 EAP estimator
4.2.5 K-means algorithm
4.3 Chapter Summary
Chapter 5 System design and implementation
5.1 General structure of the system
5.2 The GUI application
5.3 The simulation settings
5.4 Radar simulation
5.4.1 Radar parameters
5.4.2 Radar control
5.5 Chapter summary
Conclusion
结论
References
Acknowledgements
【参考文献】：
期刊论文
[1]A measurement-driven adaptive probability hypothesis density filter for multitarget tracking[J]. Si Weijian,Wang Liwei,Qu Zhiyu. Chinese Journal of Aeronautics. 2015(06)
本文编号：2995593
【文章来源】：哈尔滨工业大学黑龙江省 211工程院校 985工程院校
【文章页数】：79 页
【学位级别】：硕士
【文章目录】：
摘要
Abstract
Chapter 1 Introduction
1.1 Background
1.2 Literature review
1.2.1 Target tracking
1.2.2 Radar tracking
1.2.3 Tracking performance evaluation
1.3 Main content
Chapter 2 Multiple target trajectory evaluation
2.1 Hausdorff metric for Multi-Object trajectories
2.2 Optimal Mass Transfer （OMAT） metric
2.3 Evaluation metric （OSPA）
2.4 Chapter summary
Chapter 3 Multi-target tracking based on Random Finite Sets filtering
3.1 The Probability Hypothesis Density filter
3.2 Implementing of the PHD Filter
3.3 The Cardinalized PHD （CPHD） filter
3.3.1 Performing the CPHD estimation
3.3.2 Computational Complexity of the CPHD Filter
3.4 Simulation results and analysis
3.5 Chapter Summary
Chapter 4 Multi-target state estimation
4.1 Weight component of particles
4.2 Multi-target state estimation methods
4.2.1 ZHAO method
4.2.2 Ristic method
4.2.3 MEAP estimator
4.2.4 EAP estimator
4.2.5 K-means algorithm
4.3 Chapter Summary
Chapter 5 System design and implementation
5.1 General structure of the system
5.2 The GUI application
5.3 The simulation settings
5.4 Radar simulation
5.4.1 Radar parameters
5.4.2 Radar control
5.5 Chapter summary
Conclusion
结论
References
Acknowledgements
【参考文献】：
期刊论文
[1]A measurement-driven adaptive probability hypothesis density filter for multitarget tracking[J]. Si Weijian,Wang Liwei,Qu Zhiyu. Chinese Journal of Aeronautics. 2015(06)
本文编号：2995593
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2995593.html
