基于代理模型的武器装备体系优化算法研究

Research on the Optimization Algorithms for Weapon Equipment Systems Based on the Surrogate Model

体系优化是装备体系研究的核心问题,但是基于仿真的武器装备体系优化方法存在寻优效率低、费用高昂的缺陷。本文结合武器装备体系优化问题的特点,提出基于代理模型的武器装备体系优化算法,其中选用Kriging模型作为代理模型,通过均匀设计方法生成初始样本点,运用EI函数进行代理模型更新,采用最速下降法进行迭代优化。示例验证表明,较之仿真方法及多项式响应曲面方法,该算法具有较高的寻优精度和收敛速度,对提高武器装备体系优化的效率具有较高的理论和实用价值。

System optimization is one of the core issues in the research of the weapon equipment sys- tems. However, to overcome the low efficiency and high cost of the optimization methods of equipment systems based on simulation, an optimization algorithm of weapon equipment systems based on the Krig- ing surrogate model and uniform design is presented, in which the surrogate model is updated by the EI （expected improvement） function, and the steepest descent method is used in the optimization algo- rithm. The optimization algorithm put forward in this paper has high precision and high speed of conver- gence, and also has a high theoretical and practical value in solving non-liner problems.