|Thesis title:||Advanced Techniques for Satellites Modeling and Attitude Control|
|Research area:||Control Systems|
During the last decades modeling and simulation have become more and more important in research and development projects. The development of such tools is boosted by the necessity of reducing the design phase, both in terms of time and costs. This is even more important for space applications, where real tests are either impossible or prohibitively expensive. Generally speaking, simulators are software tools that can be used to analyze and assess system operations. Simulations are essential tools in the mission and spacecraft control design, because they provide a relatively inexpensive way for assessing the performance of our systems, try multiple configurations, establish trade-offs, verify the operation of control algorithms and strategies, and define and test innovative solutions. At the present day there are many spacecraft simulators available, but none of them seems to be able to cope with the necessary flexibility, modularity and extendibility requirements posed by the simulation of complex aerospace systems. This gap was covered by the Modelica Spaceflight Dynamics Library [Fig. 1], an object-oriented, multi domain environment, conceptually designed and developed by the author in the framework of this research project.
If the availability of an adaptable spacecraft simulator is vital to the design, development, testing and integration of a complex aerospace system, the operation of such a system requires the design of efficient and reliable control strategies. In this age of shrinking spacecraft size and smaller launch vehicle capacity, an ever growing need is evidenced for fitting more payload for more science return on a given spacecraft. Mass, volume, system complexity, reliability and cost are critical areas in the design of a spacecraft. In the light of this discussion, the investigation of control architectures making use of the class of compact and lightweight magnetic actuators takes on special relevance. Among the broad range of actuators available for spacecraft attitude control, magnetic actuators appear specially befitting in reason of their virtual inexpensiveness, simplicity and high reliability. In fact, because they do not require consumable to be carried onboard, when compared to traditional attitude control actuators, magnetic coils allow for higher efficiencies or, equivalently, longer mission operational lives or greater payload capabilities. Nevertheless, in spite of these apparent advantages, the use of magnetic actuators as a stand alone attitude control device is nowadays generally limited to small satellites on low Earth circular orbits. Indeed, the complexity inherent in the design problem, jointly with the significant savings affordable, make magnetic attitude control a very challenging and lively arena, where many open problems can still be found.
Magnetic attitude control has been a very active research topic in the last decades. However, in spite of this extensive research activity, no attention had been paid to the problem of developing design techniques for the frequently encountered case where magnetic actuators are aided in an active way. Among this class of design problems, the attitude control design for satellites equipped with magnetic actuators and a restricted support propulsion system based on impulsive thrusters is of great significance, and was thoroughly investigated with reference to the ESA's Swarm satellites [Fig. 2]. Two novel design approaches have been investigated and compared: a LTI design coupled with a control allocation problem and a joint optimal LPTV/IQC based robust design. Whereas both approaches have yielded comparable results in terms of consumables, the latter one allowed more than one order of magnitude reduction in the number of thrusters firings.
Another important issue addressed in this work, was the definition of an innovative analysis procedure for the robustness assessment of fully magnetically controlled satellites, operating in a partially uncertain environment. Despite its conservativeness, this procedure constitutes a major breakthrough, in that it is the first time that such an essential assessment can be made theoretically. The proposed analysis, based on IQC theory, provides (conservative) bounds on the robustness of three axis, fully magnetic, projection-based state feedback control laws, in presence of polytopic uncertainties on the magnetic field. Although the analysis was proposed for a satellite operating in a circular low Earth orbit, its application to several design cases, such as elliptic orbits, different magnetic coils configurations and/or preferential uncertainties on the magnetic field, is straightforward.
Finally a hybrid strategy for the robust magnetic unloading of a set of reaction wheels for three axis attitude control of ESA's MTG satellite was proposed. The proposed strategy consists of two branches. The first branch guarantees stability regardless of the uncertainties affecting both magnetic field and solar radiation pressure, provided that no saturation of the magnetic coil occurs. When this is the case, a residual is accumulated in the wheels angular momentum, which can only be eliminated by switching to the second branch. This is asymptotically stable, but is also sensible to magnetic field and solar radiation pressure deviations from their nominal values. Thus, as soon as the wheel’s residual has been removed, the hybrid strategy switches back to the first branch in order to maximize the design robustness.