Research On Control Allocation Method For Wheel-actuated Spacecraft Based On Effectivity Matrix Factorization

With the features of accurate output and consuming no propellant,Reaction Wheel(RW)is the commonly used actuator for spacecraft attitude control.Redundant RWs improve system reliability and controllability,while the redundancy obscures the mapping relationship between RW input and output,leading to some challenging issues in spacecraft control.This work based on input matrix factorization method has investigated RW Control Allocation(CA)problems,including construction of the universal CA method,design of allocation optimization algorithm and constrained CA for RW-actuated spacecraft.Since the existing RW CA algorithms are designed for specific situations and thus subject to poor adaptability,a universal dynamic CA method based on input matrix factorization is proposed.This work aided by the measure of control effectivity matrix factorization first constructs RW input and output spaces and articulates the mapping relationship between RW input and output spaces.Dividing RW input space into work space and null motion space further helps determine the one-to-one mapping between RW input and output: arbitrary output corresponds to unique input component in RW work space regardless of the amount of RW redundancy.The error optimization and allocation optimization in RW CA problem hence are separated into two independent parts and a universal CA solution form is proposed for different CA situations.The equivalence between proposed algorithm and specific method is validated through theoretical analysis and simulation.Appropriate exploitation of RW system redundancy is capable of achieving the optimization of the CA solutions,which can enhance the qualities of actuator motion and CA with CA accuracy guaranteed.A gradient-based method is proposed for RW allocation optimization process.Through designing the most rapid path in RW null motion space,allocation optimization is realized,without influence on RW output.This algorithm is specified and substantiated in RW input quadratic optimization and load balancing CA situations and RW energy dumping and unnecessary saturation avoidance strategies are presented.Simulation results validate the effectiveness of proposed methodConsidering the limitations of RW output,this work proposes an Attainable Moment Set(AMS)method for RW dynamic CA based on the three-dimensional point cloud.RW AMS and algorithm AMS model are constructed using three-dimensional point cloud principles and such model are converted to the path constraints in spacecraft attitude control.Spacecraft attitude control effect design model admissible within the AMS constraint thus is presented,with which prescribed spacecraft attitude performance is able to realize ideally.Meanwhile,an algorithm in RW null motion space is brought up to modify the allocation solution violating the admissible control set,which origins from the fact that control admissible set maps to AMS completely.This method modifies the CA solution to meet control admissible set constrains so as to pursue the full utilization of RW AMS.Simulation suggests that the proposed methods realize the prescribed spacecraft control performance and CA effect.