摘要
针对大型液体运载火箭控制参数设计依赖人工经验,设计过程耗时长,难以得出很好设计结果等问题,提出了一种控制参数智能设计方法。结合稳定裕度指标和截止频率约束,设计了一种分段式目标函数,避免单一稳定裕度值过大、其他稳定裕度很小甚至不稳定的情况。通过分析参数对系统频域特性的影响,将校正网络与控制增益分层优化,提出了一种基于优势更新的改进进化策略方法优化校正网络参数,基于Powell算法优化控制增益。仿真结果表明,提出的目标函数通用性强,提出的改进进化策略收敛速度快、收敛性好,优于其他启发式算法,优化结果优于人工设计结果,具有较高的工程应用价值。
Aiming at the problems that the design of control parameters of large liquid launch vehicle depends on manual experience,the design process takes a long time,and it is difficult to get good design results,an intelligent design method of control parameters is proposed.A piecewise objective function is proposed with combining stability margin index and cut-off frequency constraint to avoid the situation that single margin value is too large while other margins are small.The corrector network and control gains are optimized hierarchically based on analysis on frequency domain.The corrector network parameters are optimized with a modified evolution strategy with advantage updating technique.The control gains are optimized based on Powell.Numerical simulations have demonstrated that the proposed piecewise objective function is universal,the proposed modified evolution strategy has fast convergence rate and good convergence which is better than other heuristic algorithms.The optimization result is better than that of manual result and has high engineering application value.
作者
张荣升
袁晗
王紫扬
秦旭东
ZHANG Rongsheng;YUAN Han;WANG Ziyang;QIN Xudong(Beijing Institute of Astronautical Systems Engineering,Beijing 100076,China)
出处
《宇航学报》
EI
CAS
CSCD
北大核心
2023年第12期1883-1893,共11页
Journal of Astronautics
关键词
运载火箭
姿态控制
参数优化
启发式算法
智能设计
Launch vehicle
Attitude control
Parameter optimization
Heuristic algorithm
Intelligent design
