VECTOR shows its outstanding performance against a FOG unit
INTRODUCTION
Precise attitude determination is an essential process for aircraft navigation and control, and also one of the biggest challenges facing the development of any cutting-edge Air Data Attitude and Heading Reference System (ADAHRS) and Inertial Navigation System (INS).
To achieve the highest levels of precision in attitude estimation, UAV Navigation has invested in years of research and development, and is continuously looking for ways to improve the system.
The VECTOR autopilot is based around the POLAR, UAV Navigation’s high end ADAHRS-INS unit. The POLAR unit comprises an Inertial Measurement Unit (IMU), an Air Data System (ADS) and a precision, multi-constellation GNSS. The POLAR provides the heading, attitude and velocity information required by the twin redundant Flight Control CPUs in the VECTOR in order to provide control in all flight modes and to execute the required flight plan. POLAR is also available to developers as a standalone product.
POLAR is the only high-performance ADAHRS unit to offer a quaternion-based, drift-compensated sensor fusion algorithm operating with full 32-bit floating point precision at update rates as high as 500 Hz. A quaternion-based solution ensures reliable and efficient operation without the traditional problems associated with gimbal-lock. The high update rates ensure the availability of system states with minimal latency, a crucial requirement for high-performance control systems.
To demonstrate the POLAR unit’s cutting edge estimation capability, testing was carried out against a high precision reference system. The tests took place in March (fixed-wing platform, >400 kg) and June (rotary-wing platform, >600 kg) of 2018.
SETUP AND PROCEDURE
The test flights setup consisted of a POLAR unit fixed rigidly along with a tactical-grade Fiber Optic Gyroscope (FOG) based AHRS and INS solution manufactured by Kearfott. The Kearfott FOG unit was used as the reference system.
The fixed-wing test platform was a medium size, multi-payload unmanned aerial vehicle (UAV) designed for tactical long-endurance missions.
Type |
UAV |
Length |
6 m |
Wingspan |
10 m |
Gross weight |
450 kg |
Cruise Speed |
130 Km/h |
Table 1 - Fixed-wing testing platform specifications
The rotary-wing test platform was a piston-engined, manned, light helicopter.
Type |
Manned |
Length |
9 m |
Rotor Diameter |
10 m |
Gross weight |
1000 kg |
Cruise Speed |
164 Km/h |
Table 2 - Rotary-wing testing platform specifications
Several test flights were carried out, performing a variety of manoeuvres with differing dynamics (Table 3). In order to test the effect of a GNSS degraded environment on the precision of the ADAHRS unit, the GNSS signal reception was remotely disabled during the second half of every test flight.
Figure 1 - 2D trajectory view
Key:
A: Hover + negative longitudinal velocity 1
B: Hover + negative longitudinal velocity 2
C/D: Loiter right + loiter left radius 400/200 m.
Figure 2 - Velocity (top) and altitude (bottom) progression during flight. (Pictures from the UAV Navigation software Visionair Analytics)
Key:
E: Longitudinal acceleration 0-120 km/h.
F: Longitudinal deceleration 120-0 km/h.
G: Longitudinal deceleration 160-0 km/h.
H: Abrupt descent 650-100 m.
I: Abrupt ascent 100-350 m.
RESULTS
The results from the tests reveal root-mean squared (RMS) errors less than 0.5° in roll and pitch angles, independently of GNSS signal availability.
Manoeuvre |
Pitch RMS Error [deg] |
Roll RMS Error [deg] |
Pitch error standard deviation [deg] |
Roll error standard deviation [deg] |
(A) Hover + negative longitudinal velocity 1 |
0.13 |
0.068 |
0.10 |
0.06 |
(B) Hover + negative longitudinal velocity 2 |
0.18 |
0.27 |
0.18 |
0.26 |
(C/D) Loiter right + loiter left radius 400/200 m |
0.23 |
0.3 |
0.22 |
0.27 |
(E) Longitudinal acceleration 0-120 km/h |
0.11 |
0.14 |
0.11 |
0.12 |
(F) Longitudinal deceleration 120-0 km/h |
0.068 |
0.13 |
0.10 |
0.27 |
(G) Longitudinal deceleration 160-0 km/h |
0.29 |
0.27 |
0.25 |
0.27 |
(H) Abrupt descent 650-100 m |
0.1 |
0.27 |
0.10 |
0.26 |
(I) Abrupt ascent 100-350 m |
0.12 |
0.16 |
0.11 |
0.13 |
Complete flight |
0.17 |
0.23 |
0.17 |
0.19 |
Table 3 - Polar Estimation Analysis Results (March 2018)
Figure 3 - Pitch angle comparison - Manoeuvre A
Figure 4 - Roll angle comparison - Manoeuvre A
Figure 5 - Pitch angle comparison - Manoeuvre B
Figure 6 - Roll angle comparison - Manoeuvre B
Figure 7 - Pitch angle comparison - Manoeuvre C
Figure 8 - Roll angle comparison - Manoeuvre C
Figure 9 - Pitch angle comparison - Manoeuvre E
Figure 10 - Roll angle comparison - Manoeuvre E
Figure 11 - Pitch angle comparison - Manoeuvre F
Figure 12 - Roll angle comparison - Manoeuvre F
Figure 13 - Pitch angle comparison - Manoeuvre G
Figure 14 - Roll angle comparison - Manoeuvre G
Figure 15 - Pitch angle comparison - Manoeuvre H
Figure 16 - Roll angle comparison - Manoeuvre H
Figure 17 - Pitch angle comparison - Manoeuvre I
Figure 18 - Roll angle comparison - Manoeuvre I
Figure 21 - Pitch angle error - Complete flight
Figure 22 - Roll angle error - Complete flight
Figure 3 - Pitch angle comparison - Complete flight
Figure 4 - Roll angle comparison - Complete flight
CONCLUSIONS
The tests successfully showed that the POLAR unit within the VECTOR gives outstanding performance across a wide range of dynamic flight conditions, even when the GNSS system is unavailable. This is particularly impressive given that the system is using industrial grade MEMS-based sensors.
This exceptional precision is made possible by minimizing errors in all areas of the attitude estimation process.
UAV Navigation’s flight control solutions are used by a variety of Tier 1 aerospace manufacturers in a wide range of Unmanned Aerial Vehicles (UAV). These include high-performance tactical unmanned planes, aerial targets, mini-UAVs and helicopters.
The cornerstone of the Company’s success is a comprehensive, in-house capability to develop AHRS units, flight control algorithms and to fuse the data provided by multiple sensors (GNSS, airspeed, magnetometers, gyroscopes, accelerometers etc). The autopilots are characterized by their reliability and robustness, with over 60,000 flight hours in all kinds of weather and on all kinds of platform.