Residual Attention Network-Based Confidence Estimation Algorithm for Non-Holonomic Constraint in GNSS/INS Integrated Navigation System

计算机科学 全球定位系统 协方差 卫星系统 完整的 噪音(视频) 传感器融合 算法 全球导航卫星系统应用 约束(计算机辅助设计) 实时计算 惯性导航系统 人工智能 卡尔曼滤波器 惯性测量装置 残余物 导航系统 工程类 数学 电信 图像(数学) 统计 方向(向量空间) 几何学 机械工程
作者
Yimin Xiao,Haiyong Luo,Fang Zhao,Fan Wu,Xile Gao,Qu Wang,Lizhen Cui
出处
期刊:IEEE Transactions on Vehicular Technology [Institute of Electrical and Electronics Engineers]
卷期号:70 (11): 11404-11418 被引量:10
标识
DOI:10.1109/tvt.2021.3113500
摘要

Nowadays, the availability of accurate vehicle position becomes more and more indispensable. The GNSS/INS (Global Navigation Satellite Systems/Inertial Navigation System) is currently the most widely-used integrated navigation scheme for land vehicles, which is capable of provide high-accuracy and continuous positioning results in the open-sky environments. However, under the GNSS-denied conditions, the existing GNSS/INS integrated system often fails to provide reliable positioning results due to various and nonlinear errors contained in the MEMS (Micro-Electro-Mechanical System) IMU (Inertial Measurement Unit) measurements. To improve the positioning accuracy during GNSS outage, deep learning has been introduced into the GNSS/INS integrated system in recent years. In this paper, we propose a residual attention network-based confidence (i.e., measurement noise covariance) estimation algorithm for non-holonomic constraint in GNSS/INS integrated navigation system, which adopts a residual attention network to dynamically estimate the noise covariance of the pseudo-observation (i.e., non-holonomic constraint) for optimal Kalman filtering (KF) fusion. To emphasize the more representative features with larger weights for accurate noise covariance estimation, we introduce an attention mechanism to automatically assign proper weights to the learned features according to their contributions. We evaluate our proposed method on three practical road datasets and compare it with other seven methods including the traditional KF, Pure INS, KF with three deep learning networks, K-means, and the Input-Delayed Neural Networks based method. Extensive experimental results demonstrate that our proposed RA-NHC bounds the errors associated with velocities and achieves reasonable accuracy improvement in position and velocity estimation.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
1秒前
2秒前
可乐加冰完成签到,获得积分10
2秒前
David驳回了Ant应助
2秒前
3秒前
4秒前
量子星尘发布了新的文献求助10
5秒前
思源应助果粒多采纳,获得10
5秒前
6秒前
好滴捏发布了新的文献求助10
7秒前
bfs发布了新的文献求助10
8秒前
WN发布了新的文献求助10
9秒前
9秒前
慕青应助小白采纳,获得10
10秒前
AAACharlie发布了新的文献求助10
10秒前
热情的达发布了新的文献求助10
10秒前
orixero应助lucky李采纳,获得10
11秒前
11秒前
momo发布了新的文献求助10
12秒前
12秒前
13秒前
guo完成签到,获得积分10
14秒前
可期完成签到,获得积分10
15秒前
16秒前
wsj发布了新的文献求助10
16秒前
果粒多发布了新的文献求助10
17秒前
科目三应助ylq采纳,获得30
18秒前
liupc2019发布了新的文献求助20
19秒前
张雯思发布了新的文献求助10
22秒前
希望天下0贩的0应助momo采纳,获得10
22秒前
23秒前
24秒前
梦华完成签到 ,获得积分10
25秒前
26秒前
ylq发布了新的文献求助30
30秒前
okface关注了科研通微信公众号
30秒前
B2B发布了新的文献求助30
31秒前
脑洞疼应助ZZZ采纳,获得10
32秒前
小菡菡完成签到,获得积分10
34秒前
35秒前
高分求助中
A new approach to the extrapolation of accelerated life test data 1000
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 500
‘Unruly’ Children: Historical Fieldnotes and Learning Morality in a Taiwan Village (New Departures in Anthropology) 400
Indomethacinのヒトにおける経皮吸収 400
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
基于可调谐半导体激光吸收光谱技术泄漏气体检测系统的研究 350
Robot-supported joining of reinforcement textiles with one-sided sewing heads 320
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3989297
求助须知:如何正确求助?哪些是违规求助? 3531418
关于积分的说明 11253893
捐赠科研通 3270097
什么是DOI,文献DOI怎么找? 1804884
邀请新用户注册赠送积分活动 882087
科研通“疑难数据库(出版商)”最低求助积分说明 809158