A COMPUTATIONAL EFFICIENT SUBOPTIMAL ALGORITHM FOR DYNAMIC THRUSTER MANAGEMENT

摘要

This paper emphasizes on the design of a dynamic suboptimal thrust selectionalgorithm for reaction control system with redundant configured thrusters. Differingfrom using the thruster management function cone (TMF_cone) to constructspecial look-up tables that store the precomputed optimal bases of theauxiliary problem, this algorithm is based on the cascaded searching of the optimalor the suboptimal TMF_cone. During each iteration of the algorithm, anoptimal or near optimal TMF_cone for the unconstrained thruster managementproblem is first found. Then, minimum and maximum open times are takeninto account to manage commands exceeding the up or lower limit of the constraints.Control error is calculated for next iteration until it finally comes tozero and no more thrusters exceed the constraints. An Illustrative examplebased on the core module of the space station is provided to verify the effectivenessand correctness of the proposed method as well as to demonstrate itsadvantage in computational requirement and adaptiveness to thruster failure.