What Control Systems innovations are involved?
First generation AOCS components were quite specific to each type of orbit and application, made up a large amount of spacecraft mass, needed quite intensive ground support and relied on moving parts such as mechanical gyroscopes whose motion could perturb the spacecraft platform and were prone to premature breakdown.
The aim of current R&D within this sphere is to increase sensor autonomy and develop generic designs capable of serving applications, reducing the overall size and operational complexity of the AOCS subsystem while enhancing spacecraft controllability and boosting their reliability and operational lifetime - applying Active Pixel Sensor (APS) and micro-electronic mechanical system (MEMS) technology where possible.
New AOCS devices under development include autonomous star trackers capable of reliably identifying spacecraft attitude without any additional navigational information (known as “lost in space” capability). Also being developed are MEMS and Fibre Optic Gyroscopes without any moving parts, instead measuring the Coriolis effect on a vibrating ring or changes in light passing through a fibre optic loop to detect attitude angular rates.
Many technological developments are also undertaken in the area of modern Control Theory, in order to handle more efficiently the design and validation of more demanding space applications.
The ultimate goal is to provide space industry with a consistent set of highly efficient techniques and tools for the various engineering phases: modelling, controller design, and system verification, in order to meet costs and planning requirements, despite the fact that control engineering has to contend with space systems whose uncertainty, non-linearity and complexity are steadily increasing.
A wide scope of modelling tools are being developed to cover interdisciplinary effects as aerodynamics, structural and fluid couplings, together with trajectory optimization. Uncertainty modelling tools such as Linear Fractional Transformations (LFT) and Linear Parameter Varying (LPV) are developed to incoporate the effects of system tolerances, non-linearities, degradations and parametric variations.
For controller design, various multivariable techniques, following the successful introduction of H-infinity, are being further advanced, such as Linear Parameter Varying Control (LPV) and Linear Matrix Inequality (LMI) based techniques.
For system verification many efforts are ongoing to develop efficient Worst Case Analysis tools in order to provide efficient alternatives to the current Monte Carlo approach, which applies statistical sampling techniques to obtain probabilistic estimates of effectiveness.
Last update: 29 September 2009