Amirreza Kosari - Assistant professor, University of Tehran, Faculty of new science and technology
Samane Kaviri - M.Sc. Student, University of Tehran,
Behzad Moshiri - Professor, University of Tehran
Mahdi Fakoor - Assistant Professor, University of Tehran

In this work we consider a new technique for the control allocation problem for an over-actuated GEO Communications satellite. Control allocator system computesthe required commands that should be applied to the optimum distributed thrusters on the outer surface of the satellite so thata certain set of torques would be generated using this actuators.The number of actuators is greater than the number of torque necessary to be compensated, and the controlling ranges of theactuators are constrained to the certain pre-specified linear active domains. Control allocation problem is formulated asan optimization problem so that all of the available degrees of freedom can be utilized. When sufficient control power exists,secondary performance objectives could be achieved. The proposed methodology enables reconfiguration control in the presence of actuator failures, while the controller keeps beingunchanged. The differential evolution optimization method is used to allocate control commends. The performance andaccuracy of the proposed method have been verified using pseudo-inverse and linear programming methods. The case study results applied to a typical GEO satellite control subsystem design, demonstrates the effectiveness of the differential evolution method in comparison with the other algorithms.

Control Allocation, Satellite Attitude Control, Control Constraints, Actuator Failure