| The paper addresses the issue of reconfigurability allocation for satellite attitude control system.Combining the theory of reconfigurability and the characteristics of satellite attitude control system,three reconfigurability allocation methods are proposed. Finally simulation results verify theeffectiveness of the proposed methods.Firstly, the research background, meaning and purpose are surveyed, and then the history and thedevelopment of the reconfigurability and reconfigurability allocation are described. An overview ofthe present state of study and research about satellite attitude control system is elaborated, followedby the description of main contents of this article.Secondly, reconfigurability criteria based on controllable theory is proposed. Then, the degree ofreconfigurability is defined as the ratio of the number of reconfigurable faults to the number of allfaults. Moreover, two optimal reconfigurability allocation models are analyzed.Thirdly, in the respect of reconfigurability allocation with no constraint, the departed allocationmethods are compared with each other according to application ranges. Then, in order to overcomethe lack of reconfigurability statistical data and ambiguity of human knowledge, the reconfigurabilityallocation method based on fuzzy comprehensive evaluation is presented for the systems underunrestricted condition. The simulation results shows the effectiveness of the proposed method.Finally, in the respect of reconfigurability allocation with constraints, according to the number ofoptimization goals, the reconfigurability allocation with constraints is divided into two issues: singleobject optimization and multi-objective optimization. Then the reconfigurability allocation based onheuristic algorithm is proposed for the former. This method can solve the problem of resourcesoptimization allocation with constraint conditions, simulation result comes to the maximum ofreconfigurable degree. Moreover, to solve the problem of optimization allocation for conflictingobjectives, different evaluation functions are put forward on the latter issue. The optimal result isconcluded by weighing feasible results. The simulation results demonstrate superiority of theproposed methods. |
