UWE-2 (University of Wurzburg Experimentalsatellit-2)
UWE is a follow-on picosatellite technology demonstration project within the CubeSat family standard, developed and built by students of the University of Würzburg, Germany. The overall objective is to demonstrate the capabilities of attitude determination and control in picosatellites. 1)
Figure 1: Computer rendered CAD model of the UWE-2 satellite (image credit: University of Würzburg)
The picosatellite design design is based on the CubeSat standards with regard to size (10 cm side length) and mass (≤ 1 kg). It includes all standard subsystems (e.g. power, on board data handling, telecommunication), but was optimized with respect to small size and low weight. The satellite, together with its electronic boards, was designed, developed and manufactured at the University Würzburg.
Special attention was paid to the low power consumption requirements, leading to the selection of the Hitachi H8 microprocessor. This computer is capable of running a complex operating system, providing access to a broad spectrum of applications and protocols. UWE-2 uses the µCLinux operating system (a GNU/Linux based operating system) for the OBDH (On Board Data Handling). To allow consistent implementation methods for every sensor, the SPI (Serial Peripheral Interface) bus system was selected as underlying physical communication layer. This master-slave bus provides very high transmission rates with low protocol overhead and easy implementation in the software controlling part (Figure 3).
Spacecraft power is generated from highly efficient triple-junction GaAs solar cells, providing an efficiency of 28%. The RF communication subsystem is based on a transceiver using the amateur frequency bands (VHF/UHF).
The determination of the actual attitude and orbit is a generally a basic requirement for virtually all spacecraft missions. However, the implementation of a fairly accurate ADS (Attitude Determination Subsystem) represents a great challenge in a picosatellite confinement due to the severe volume and mass restrictions. 2)
To enable an appropriate attitude determination on UWE-2, the picosatellite was equipped with redundant sensors for robust and accurate measurements. The sensor equipment is mainly based on commercial MEMS (Micro-Electromechanical System) components, which were selected with respect to low mass and power consumption. Complete 3-axis attitude sensing can neither be provided by magnetometers nor by sun sensors. Hence, a combination of both techniques was used to derive solutions of the equations for attitude determination. The magnetometers contribute measurements of the magnetic field vector induced by the Earth's magnetic field on the satellite body. The sun sensors detect the sun incidence angle to the satellite’s reference frame.
The UWE-2 satellite carries six pairs of perpendicularly mounted individual sun sensors, one on each panel. For complementary measurements three miniature gyros are included, contributing changes in attitude and a direct measurement of the axial spin rates. An accelerometer complements the gyros to form the inertial navigation system. TLE (Two-Line Element) data can - to some extent – provide ground measurements of the absolute position, such that an offline correction of the relative data on ground is possible.
Onboard position and time is provided by a GPS receiver. A Phoenix GPS receiver of DLR was integrated.
Onboard actuation is provided with small permanent magnets (they interact with the Earth's magnetic field thereby exerting a moment). This so-called passive technique offers a 2-axis stabilization capability; there remains 1 degree of freedom, which cannot be influenced. - Full 3-axis actuation with reaction wheels is being realized in the UWE-3 mission (in implementation phase as of 2009).
ADS software: The ADS of UWE-2 consists of the orbit generation subsystem, the subsystem for obtaining the reference sun vector and magnetic field vector and an EKF (Extended Kalman Filter), performing the data fusion and estimating the satellite’s attitude. Figure 2 depicts the complete functional block diagram of the attitude determination system. 3)
The orbit generator relies on the TLE (Two Line Element) set for the initial orbit elements required to generate the orbit positions. Since TLE data are generated in a specific way, the orbit elements are reconstructed by calculating the osculating values using the SGP4 (Simplified General Perturbations) propagation model. The orbit generator uses the general perturbations method to predict variations in the orbital parameters due to noise effects from the environment. The perturbations considered for UWE's orbit generator are third-body perturbations due to gravitational forces of the sun and the moon, perturbations due to non homogeneous gravity potential of the Earth and perturbations due to atmospheric drag.
The attitude estimator requires a reference magnetic field vector and sun vector information to compare the information provided by the magnetometer and the sun sensors. Due to constraints of storage space on board the satellite computer, storing large amount data for expected magnetic field values is not feasible. Instead, using the satellite positions provided by the orbit generator, the reference magnetic field value is estimated by using the IGRF (International Geomagnetic Reference Field) 2005 standard, mathematical model.
The attitude of the satellite is defined with respect to the orbit frame whose origin coincides with the geometrical center of the satellite. The center of mass and the geometrical center of the satellite are assumed to coincide. The satellite’s Roll-Pitch-Yaw frame rotates relative to the inertial frame. The attitude parameterization for UWE is done by using quaternions. Quaternion representation of rotation leads to convenient representation of attitude kinematics as linear equations. Additionally, the quaternion unit norm provides an easy way to preserve the orthogonality of the rotation matrix. Furthermore, the composite rotations can be performed in terms of quaternion multiplication without involving trigonometric functions.
Figure 2: Functional block diagram of the ADS (image credit: University of Würzburg)
The EKF (Extended Kalman Filter) package is at the heart of UWE's attitude estimation algorithm. The nonlinear state equations are linearized about the best estimate using the EKF. While the attitude kinematics are being estimated by the filter, the satellite dynamics are measured using the onboard gyro. The gyro drift errors are being compensated.
An appropriate performance of the attitude determination system can be achieved in this way (verified by simulation demonstrations). When these results can also be verified in orbit, then a good basis is created for coordinated operations of future satellite swarms, promising interesting capabilities in the field of Earth observations.
Figure 3: Data flow concept between the ADS and OBDH (image credit: University of Würzburg)
Figure 4: Photo of highly integrated integrated electronic boards inside UWE-2 (image credit: University of Würzburg)
Launch: UWE-2 (University of Würzburg) was launched as a secondary payload on Sept. 23, 2009 on a PSLV launcher from SDSC-SHAR on the east coast of India. Use of SPL (Single Picosatellite Launcher) of Astro- und Feinwerktechnik Adlershof GmbH (Berlin) for the deployment of the CubeSats. The primary payload on the flight is OceanSat-2 of ISRO (Indian Space Research Organization).
Further secondary payloads (CubeSats) on the flight are: BeeSat (Berlin Experimental Educational Satellite) mission of the TUB (Technical University of Berlin), Berlin, Germany, ITUpSat-1 (Istanbul Technical University PicoSatellite-1) of Istanbul Technical University, Turkey, and SwissCube of EPFL (Ecole Polytechnique Federale de Lausanne), Lausanne, Switzerland. 4)
All secondary payloads are using the SPL (Single Picosatellite Launcher) system for on-orbit deployment. SPL is a development of Astro und Feinwerktechnik Adlershof GmbH, Berlin, Germany.
Orbit: Sun-synchronous near circular orbit, altitude = 720 km, inclination = 98.28º, period = 99.31 min, the local equatorial crossing time is at 12:00 hours.
Figure 5: Integration of UWE-2 in the cleanroom (image credit: University of Würzburg)
1) Oliver Kurz, “Design and Implementation of an Attitude Determination System for the Cubesat UWE-2 (Hardware based),” Master's Thesis, University of Würzburg, Germany, Luleå University of Technology, Sweden, Dec. 4, 2007, URL: http://epubl.ltu.se/1653-0187/2008/032/LTU-PB-EX-08032-SE.pdf
2) Marco Schmidt, Karthik Ravandoor, Oliver Kurz, Stephan Busch, Klaus Schilling, “Attitude Determination for the Pico-Satellite UWE-2,” Proceedings of the 17th World Congress The International Federation of Automatic Control (IFAC), Seoul, Korea, July 6-11, 2008, URL: http://www.nt.ntnu.no/users/skoge/prost/proceedings/ifac2008/data/papers/4264.pdf
3) Marco Schmidt, Klaus Schilling, “An extensible on-board data handling software platform for pico satellites,” Acta Astronautica, Volume 63, Issues 11-12, December 2008, pp.1299-1304
The information compiled and edited in this article was provided by Herbert J. Kramer from his documentation of: "Observation of the Earth and Its Environment: Survey of Missions and Sensors" (Springer Verlag) as well as many other sources after the publication of the 4th edition in 2002. - Comments and corrections to this article are always welcome for further updates.