Cooperative Control Strategy applied to on-orbit assembly of robotic modules
- Paper number
IAC-09.C1.4.2
- Author
Ms. Chiara Toglia, University of Rome "La Sapienza", Italy
- Coauthor
Mr. Daniel Kettler, Massachussets Institute of Technology (MIT), United States
- Coauthor
Prof. Steven Dubowsky, Massachussets Institute of Technology (MIT), United States
- Year
2009
- Abstract
Space systems can be assembled directly in orbit using self-assembling reconfigurable robotic satellites. This concept has many advantages. The modules are small, so production and launch costs are reduced. The launch mission can be planned using low cost standard modules. The modular approach also provides a natural redundancy and the ability to repair and refuel the system by replacing modules. However, this concept introduces many technical challenges which may lead to poor dynamic performance and instability if not addressed: changing inertia properties, reaction jet impingement problems during assembly, and limits on propellant storage because of the small size of the modules. Moreover, since propellant cannot be shared, constraints should be placed on the control authority so that fuel remaining resources are evenly distributed among the modules. Finally, a good control strategy should consider and exploit redundancy of sensors and actuators for better performance. This paper presents a methodology to control an assembly of robotic modules in orbit. A Cooperative Control strategy of the assembly is proposed. Such a strategy is optimal, i.e. assures minimal cost and minimal error trajectory, insures stability, avoids plume impingement, and best uses the combined resources (in terms of residual propellant) of each robot. Numerical simulation analysis of different configurations is presented, showing the effectiveness of the Cooperative Control. Redundancy of actuators is used to switch off poorly positioned thrusters that may damage adjacent modules and to redistribute consumption so that resource levels stay evenly apportioned. Measurement errors are reduced by combining data from each available sensor, thus exploiting sensors redundancy. A comparison between the coordinated control and a distributed control methodology, such that every module is independently controlled, is carried out, showing differences in terms of consumption and trajectory errors
- Abstract document
- Manuscript document
(absent)