High-Efficiency Thrust Vector Control Allocation

摘要

A generalized approach to the allocation of redundant thrust vector slew commands for multi-actuated launch vehicles is presented, where deflection constraints are expressed as omniaxial or elliptical deflection limits in gimbal axes. More importantly than in the aircraft control allocation problem, linear allocators (pseudoinverses) are preferred for large booster applications to facilitate accurate prediction of the control-structure interaction resulting from thrust vectoring effects. However, strictly linear transformations for the allocation of redundant controls cannot, in general, access all of the attainable moments for which there is a set of control effector positions that satisfies the constraints. In this paper, the control allocation efficiency of a certain class of linear allocators subject to multiple quadratic constraints is analyzed, and a novel single-pass control allocation scheme is proposed that augments the pseudoinverse near the boundary of the attainable set The controls are determined over a substantial volume of the attainable set using only a linear transformation; as such, the algorithm maintains compatibility with frequency-domain approaches to the analysis of the vehicle closed-loop elastic stability. Numerical results using a model of a winged reusable booster system illustrate the proposed technique's ability to access a larger fraction of the attainable set than a pseudoinverse alone.