多级优化算法在火炮总体结构设计中的应用  

Application of Multilevel Optimization Algorithm in Artillery Integrated Design

作　　者：单春来 刘朋科[1] 古斌 贺琦 徐宏英 SHAN Chunlai;LIU Pengke;GU Bin;HE Qi;XU Hongying(Northwest Institute of Mechanical and Electrical Engineering,Xianyang 712099,Shaanxi,China)

机构地区：[1]西北机电工程研究所,陕西咸阳712099

出　　处：《兵工学报》2022年第1期11-19,共9页Acta Armamentarii

基　　金：武器装备预先研究项目(2020年)。

摘　　要：随着现代战争模式对武器装备性能要求不断提高,需要新的设计方法以实现火炮总体结构参数最优设计,为此提出一种能够综合考虑各方面性能的火炮总体结构优化设计方法。建立火炮多体动力学的参数化模型,基于最优拉丁超立方试验设计方法得到设计参数的灵敏度和大量的输入-输出数据,由各参数的灵敏度合理计算不同射击工况及不同评价指标下的权重值,由输入-输出数据结合神经网络算法构建近似函数。搭建多级优化算法模型,计算收敛后即可得到火炮的最优设计方案。结果表明:与初始方案相比,优化后的设计方案稳定性有明显提高,评价指标平均改善22.27%;该方法从发射动力学角度可为火炮总体结构最优设计提供参考。The weapons and equipment are required to have high-mobilility,high-precision,long-range and lightweight in modern warfare.A noverall optimization design method is presented,in which the all aspects of performance of artillery are comprehensively considered.A parametric model of artillery is established,the sensitivity coefficients are calculated using the optimal Latin hypercube experimental design,and a large number of input-output data can be obtained at the same time.The weight values under different firing conditions and different evaluation indexes are calculated from the sensitivity coefficients,and the approximate functions are constructed based on input-output data.A multilevel optimization algorithm is set up based on these results above,and an optimal design scheme would be obtained after the algorithm convergence.The result shows that the evaluation indexes of the optimal design scheme are averagely increased by 22.27%,which is obviously improved compared with the initial design scheme.

关 键 词：火炮 动力学 稳定性 总体设计 多级优化算法 

分 类 号：TJ302[兵器科学与技术—火炮、自动武器与弹药工程]