固定翼无人机姿态估计的互补滤波器

需积分: 14 63 浏览量更新于2024-09-07 收藏 560KB PDF 举报

"AComplementaryFilterforAttitudeEstimationofaFixed-WingUAV，这篇论文探讨了如何在固定翼无人机（UAV）的姿态估计中利用低成本惯性测量单元（IMU）的测量数据来设计非线性互补滤波器。" 在航空领域，特别是无人机技术中，姿态估计是至关重要的，它涉及到对飞行器的姿态（如俯仰、翻滚和偏航角）的准确计算。互补滤波器是一种常用于传感器融合的技术，它能有效地结合来自不同传感器的数据，以提高系统对环境变化的响应和准确性。在这个特定的论文中，作者提出了一个针对固定翼无人机姿态估计的非线性互补滤波器方案。该滤波器的设计思路是利用加速度计输出进行低频姿态估计，而通过整合陀螺仪输出进行高频姿态估计。加速度计通常能够提供重力分量的信息，但在飞机转弯时，还会包含因为空气动力学效应产生的机架加速度。这部分加速度主要由向心力引起，当飞机做曲线飞行时尤为明显。为了分离出重力分量，论文提出使用一个简单的向心力模型，这个模型基于额外的空速测量数据。同时，通过引入一个一阶动态模型来考虑攻角的影响，以更精确地估计出重力方向。攻角是描述空气流与飞机翼面相对角度的关键参数，其动态变化会影响飞机的升力和飞行特性。实验部分，论文使用实际世界的数据集对提出的滤波器进行了测试，并将其性能与其他输出进行了比较。这样的评估有助于验证滤波器的有效性和可靠性，特别是在动态飞行条件下的表现。 "A complementary filter for attitude estimation of a fixed-wing UAV" 这篇论文为固定翼无人机的姿态估计提供了一种创新方法，通过巧妙地处理加速度计和陀螺仪的数据，实现了更准确的姿态解算，这对于无人机自主导航和控制系统的优化具有重要意义。

A Complementary Filter for Attitude Estimation of a Fixed-Wing UAV

Mark Euston, Paul Coote, Robert Mahony, Jonghyuk Kim and Tarek Hamel

Abstract— This paper considers the question of using a non-

linear complementary ﬁlter for attitude estimation of ﬁxed-wing

unmanned aerial vehicle (UAV) given only measurements from a

low-cost inertial measurement unit. A nonlinear complementary

ﬁlter is proposed that combines accelerometer output for low

frequency attitude estimation with integrated gyrometer output

for high frequency estimation. The raw accelerometer output

includes a component for the airframe acceleration, that occurs

primarily as the aircraft turns, as well as the gravitational

acceleration that is required for the ﬁlter. The airframe ac-

celeration is estimated using a simple centripetal force model

(based on additional airspeed measurements), augmented by

a ﬁrst order dynamic model for angle-of-attack, and used to

obtain estimates of the gravitational direction independent of

the airplane manoeuvres. Experimental results are provided

on a real-world data set and the performance of the ﬁlter is

evaluated against the output from a full GPS/INS that was

available for the data set.

I. INTRODUCTION

Attitude determination is an essential task for an Un-

manned Aerial Vehicle (UAV). With the growing range of

applications in UAV’s, and the push to make vehicles cheaper

and more reliable, it is of interest to develop robust and

simple algorithms for attitude estimation [1], [9]. There is

a large literature on attitude ﬁltering techniques, see for

example the recent review article by Crassidis et al. [6].

Most of the advanced ﬁlter techniques (particle ﬁltering, etc.)

are computationally demanding and unsuitable for the small

scale embedded processors in UAV avionic systems. The two

methods that are commonly used are extended Kalman ﬁlter-

ing (EKF) or some form of constant gain state observer, often

termed a complimentary ﬁlter due to its frequency ﬁltering

properties for linear systems [12]. Extended Kalman Filtering

has been studied for a range of aerospace applications [7],

[9], [6], [16]. Such ﬁlters, however, have proved difﬁcult to

apply robustly [14], [4], [16]. In practice, many applications

use simple linear single-input single-output complementary

ﬁlters [16], [5]. In recent work, a number of authors have

developed nonlinear analogous of single-input single-output

(SISO) ﬁlters for attitude estimation [15], [19], [18], [13],

[11], [2]. To implement these schemes on a UAV using

inertial measurement unit (IMU) data the accelerometer

output is used to estimate the gravitational direction. The

recent work by the authors [8], [12] allows the full estimation

of vehicle attitude (up to a constant heading error) as well as

gyro biases based just on the accelerometer and gyrometer

M. Euston, P. Coote, J. Kim and R. Mahony are with Department of

Engineering, Australian National University, ACT, 0200, Australia. e-mail:

firstname.surname@anu.edu.au.

T. Hamel is with I3S-CNRS, Nice-Sophia Antipolis, France. e-mail:

thamel@i3s.unice.fr.

data. The ﬁlter fails, however, when the vehicle dynamics

are sufﬁciently large that accelerometer output no longer

provides a good estimate of the gravitational direction. This

is particularly the case for a ﬁxed wing UAV manoeuvering

in a limited space and making repeated rapid turns.

In this paper, we develop a nonlinear complementary ﬁlter,

augmented by a simple ﬁrst order model of vehicle dynamics

that provides excellent attitude estimates for a ﬁxed wing

UAV. The key contribution is to develop a model of the

non-inertial acceleration of the airframe that can be used

to compensate the accelerometer output to obtain a zero

bias estimate of the gravitational direction. The model is

based on a simple centripetal force model derived from the

airspeed and the rate of turn of the vehicle. However, the

angle-of-attack of the airplane is signiﬁcantly higher during

a sharp turn, and this must be modelled to correctly align

the compensation terms for the accelerometer output. We

address this problem by incorporating a simple ﬁrst order

model of the angle-of-attack dynamics of the airframe driven

by the pitch rate measurement obtained from the gyrometer

output. The combined system is simple to implement and

achieves excellent performance, given the minimal data that

is available. The algorithm is veriﬁed on experimental data

from a ﬁxed wing aerial robotic vehicle. The performance of

the algorithm is conﬁrmed by comparison with an attitude

estimate obtained from a full INS/GPS stochastic ﬁlter that

has been run on the experimental data.

II. EXPLICIT COMPLEMENTARY FILTER

A complementary ﬁlter for attitude estimation performs

low-pass ﬁltering on a low-frequency attitude estimate, ob-

tained from accelerometer data, and high-pass ﬁltering on a

biased high-frequency attitude estimate, obtained by direct

integration of gyrometer output, and fuses these estimates

together to obtain an all-pass estimate of attitude. When

the pitch and roll of a airplane are modelled as decoupled

processes, a SISO ﬁlter can designed for each signal that

uses the the angle between the accelerometer output and

the body-ﬁxed-frame as attitude reference and the separate

gyrometer axis output as velocity reference in a classical

linear complementary ﬁlter [3]. When a low pass estimate

of the full attitude can be reconstructed from the IMU

measurements, for example if magnetometer data is also

available and the full coupled rotation matrix for attitude

can be computed as an algebraic function of the gravitational

and magnetic ﬁelds measured in the body-ﬁxed-frame, then

nonlinear extensions of the complementary ﬁlters have been

available for ﬁfteen years [15], [18]. Magnetometers are

rarely useful on small scale UAVs due to the perturbation of

下载后可阅读完整内容，剩余5页未读，立即下载

zhangnage123

粉丝: 0

固定翼无人机姿态估计的互补滤波器

IMU pose complementary filter.zip

A Complementary Filter for Attitude Estimation of a Fixed-Wing U

Novel design of tri-band bandpass filter based on fractal-shaped geometry of a complementary single split ring resonator

Design Optimization of a Complementary and Modular Linear Flux-Switching Permanent Magnet Motor for Long Stator Applications

Novel Capacitance Coupling Complementary Dual-Direction SCR for High-Voltage ESD

A miniaturized system for measurement of the refractive index of sub-microliter liquid

Hilbert-shaped complementary single split ring resonator and low-pass filter with ultra-wide stopband, excellent selectivity and low insertion-loss

Selective Conversion from p-Type to n-Type of Printed Bottom-Gate Carbon Nanotube Thin-Film Transistors and Application in Complementary Metal-Oxide-Semiconductor Inverters

Hilbert-shaped complementary ring resonator and application to enhanced-performance low pass filter with high selectivity

A Method Research for Generating Complete Complementary Sequences

最新资源