摘要
提出两种改进型模型预测静态规划(MPSP)制导律算法,以提高在弱初始猜测条件下的收敛性。首先,将MPSP算法视作应用于欠定系统求根问题的牛顿迭代法;然后,基于线搜索和信赖域策略提出两种改进型MPSP算法;最后,以多约束条件下导弹末制导问题作为算例进行仿真分析,结果表明两种改进型MPSP算法在保证计算效率的前提下,具备更强的收敛性。
The two improved model predictive static programming(MPSP)algorithms are presented to address the drawback of poor convergence property.Firstly,the MPSP is treated as a Newton-type method for underdetermined systems.Then,the line-search and trust region strategies are introduced to improve the performance of the MPSP algorithm.Finally,numerical simulations are conducted for missile terminal guidance problems.Results show that the improved algorithms have excellent convergence performance as well as high computational efficiency.
作者
胡任祎
贺彦峰
史丽楠
马洋洋
泮斌峰
Hu Renyi;He Yanfeng;Shi Linan;Ma Yangyang;Pan Binfeng(Beijing Aerospace Automatic Control Institute,Beijing 100854,China;School of Astronautics,Northwestern Polytechnical University,Xi'an 710072,China)
出处
《航天控制》
CSCD
北大核心
2022年第3期22-27,共6页
Aerospace Control
关键词
模型预测静态规划
线搜索
信赖域
导弹
末制导
Model predictive static programming
Line search
Trust region
Missile
Terminal guidance
