An Ultra-Wideband Autonomous Integrity Monitoring Algorithm Based on Post-Test Residuals
1. 摘要
引入RAIM对UWB作故障检测,对发生故障的历元降低权重
2. 引言
The primary research focus in UWB positioning technology centers on non-line-of-sight issues, error modeling, positioning algorithms, and anchor node placement, all of which are vital considerations when utilizing UWB for indoor positioning.
High accuracy and credibility are both critical in positioning. Nevertheless, the evaluation method for determining the credibility of UWB positioning results is currently suboptimal.
- Thus, this paper employs the doubledifference UWB positioning method (DUWB)
- Receiver Autonomous Integrity Monitoring (RAIM) theory is utilized in UWB by employing the post-test residuals
3.定位模型
用的三维位置,什么坐标系没说明
线性化方程,然后用最小二乘求解
Post-test可能是先定位后检测故障,这不是正常流程吗
故障检测模型(RAIM)
下面是正常流程,残差平方和作检测量,服从卡方分布,标准化,根据设定
P
f
a
P_{fa}
Pfa计算检测阈值
公式8应该是有问题的,不用除以(n-t),是
S
S
S服从
n
−
t
n-t
n−t自由度的卡方分布,后面求HPL写的乱七八糟,不再说明
这里放我自己的求PL理解。
RAIM中的PL的计算
σ
D
\sigma \sqrt{D}
σD是系统的偏差和,通过
D
=
∣
r
∣
2
σ
2
D=\frac{\lvert r \rvert^2}{\sigma ^2}
D=σ2∣r∣2反算,是系统的偏差的模,然后映射到每个卫星斜率上面,每个卫星斜率分为水平斜率和垂直斜率,具体就是投影矩阵对应的行和列,这里不再说明怎么取值。
后面值得说明是
S
S
S矩阵是什么
故障如何排除,如何降低权重一点没说
实验
分为静态实验和动态实验
两组仪器,反正就是好,单UWB误差调的有点离谱,全文完。