火炮多柔体动力学结构优化研究  

Research on Flexible Multi-body Dynamics Structure Optimization of Artilleries

作　　者：萧辉[1] 杨国来[1] 孙全兆[1] XIAO Hui YANG Guo-lai SUN Quan-zhao(School of Mechanical Engineering, Nanjing University of Science and Technology, Nanjing 210094, Jiangsu, China)

机构地区：[1]南京理工大学机械工程学院,江苏南京210094

出　　处：《兵工学报》2017年第1期27-34,共8页Acta Armamentarii

基　　金：国家自然科学基金项目(11572158);国家"973"计划项目(1503613249)

摘　　要：针对现有动力学优化方法很难对多体系统中的柔性体进行结构优化的问题,提出结合多学科代理模型法和改进的非支配排序遗传算法的多柔体动力学优化方法。以火炮炮口振动参数为输出,柔体模态参数和部分火炮总体参数为输入,在已验证的多体刚柔耦合模型的基础上,采用径向基函数-反向传播神经网络建立了具有良好泛化能力和预测精度的代理模型。利用改进的非支配排序遗传算法对各炮口振动参数进行动力学优化,采用max-min准则从优化得到的Pareto最优解集中优选出一个较为兼顾各优化目标的解,与原模型结果对比,优化效果明显。该方法能用于火炮多柔体动力学结构优化,也为火炮总体结构设计和相关优化提供了一定的参考。In allusion to the problem of that the existing structural dynamics optimization methods cannot optimize the flexible parts in flexible multi-body dynamic systems, a method combining muhidisciplinary agent models and the improved nondominated sorting genetic algorithm （NSGA-II） is proposed. Based on an experimentally authenticated rigid-flexible coupled multibody model, a surrogate model with good gen- eralization ability and forecasting accuracy is established with RBF-BP neural network. In the proposed model, the muzzle vibration parameters are used as outputs, and the modal parameters of flexible part and some general structural parameters are taken as inputs. Nondominated sorting genetic algorithm is used to improve the muzzle vibration characteristics, and the max-min criterion is adopted to select a solution from the Pareto front. The optimized effect is compared with the optimized result of the original model. The result shows that the proposed method can be used for optimization of artillery flexible multi-body dy- namics structure.

关 键 词：兵器科学与技术 多柔体优化 炮口振动 径向基函数-反向传播神经网络 代理模型 遗传算法 

分 类 号：TJ302[兵器科学与技术—火炮、自动武器与弹药工程]