自适应跟踪算法设计+文献综述
目标发生运动状态的改变时，单一模型的跟踪效果会大打折扣。本文主要目的是利用卡尔曼递推算法设计一个能同时实现跟踪精度高，并且能自动转换滤波模型的自适应滤波器，以方便地追踪机动目标。22093
本文首先介绍目标模型、坐标系的选取以及目标跟踪的原理和方法，重点讨论卡尔曼滤波算法。接着，在此基础上，研究了机动检测和自适应滤波算法。然后编写Matlab代码，用卡尔曼滤波算法对匀速CV模型和匀加速CA模型滤波仿真。最后，针对匀速和匀加速运动之间的转换问题，提出模型切换算法。仿真结果显示带有模型切换的卡尔曼滤波精度高，在目标跟踪中有着非常出色的表现。
关键词 目标跟踪；卡尔曼滤波；自适应滤波；模型转换
毕业论文设计说明书（论文）外文摘要
Title Adaptive Target Tracking Algorithm
Abstract
With the development of communications and positioning technologies in modern society, target tracking has been applied in both military and people's livelihood. The usually used filters for maneuvering target tracking have filter, Kalman filter and particle filter etc. They are applied to communication, radar, navigation, automation and so on.
When the target’s motion state changed, the tracking performance of a single model will be decreased. In this thesis a adaptive filter is designed which can realize both high accuracy, real-time tracking and automatic model transformation.
Firstly， the principle of target tracking was summarized, including how to choose target model and coordinate system. Details of Kalman filter are discussed. Secondly, the maneuver detection and adaptive filtering algorithm are researched. Then, the Kalman filtering programs with Matlab are developed to simulate constant-velocity model and constant-acceleration model. Finally, to the transformation issues between above two models, model switching algorithm is put forward. The simulation results showed that the proposed algorithm with model switching has high precision and outstanding performance compared with conventional filters.
Keywords Target tracking; Kalman filtering; Adaptive filtering; Model transformation
目 次 I
1 绪论 1
1．1 研究背景 1
1．2 国内外发展现状 1
1．3 本文主要研究及结构 2
2 目标跟踪滤波原理与方法 3
2．1 跟踪坐标系选择 3
2．2 目标模型 4
2．3 基本的跟踪滤波与预测方法 9
3 卡尔曼滤波 13
3．1 卡尔曼滤波的发展 13
3．2 卡尔曼滤波基本算法 15
3．3 卡尔曼滤波的性质 17
4 机动目标跟踪中的自适应滤波 18
4．1 自适应滤波算法 18
4．2 CV模型自适应滤波 20
4．3 CA模型自适应滤波 22
5 算法设计与仿真 24
5．1 CV模型代码及仿真 24
5．2 CA模型代码及仿真 28
5．3 自适应模型切换设计 30
结 论 35