基于PSO-LSN算法的冲击响应谱时域波形合成方法研究  

Time domain waveform synthesis method of shock response spectrum based on PSO-LSN algorithm

作　　者：蒋辰玮 王军评[1] 严侠[1] JIANG Chenwei;WANG Junping;YAN Xia(Institute of Systems Engineering,China Academy of Engineering Physics,Mianyang 621900,China)

机构地区：[1]中国工程物理研究院总体工程研究所,四川绵阳621900

出　　处：《振动与冲击》2024年第23期102-107,118,共7页Journal of Vibration and Shock

基　　金：院长基金自立项目(YZJJZL2023050)。

摘　　要：针对冲击响应谱(shock response spectrum,SRS)试验过程中,试件受非线性、局部共振等因素影响导致控制易出现局部超差,需多次修正迭代时域基波波形参数的问题,通过分析冲击响应谱试验结果主要影响因素及变化机理,提出基于自适应学习的空间邻域驱动策略粒子群算法(particle swarm optimization based on learning spatial neighborhood driven,PSO-LSN)。根据粒子邻域相似性增强局部空间搜索能力,共享最优位置与速度信息,并结合自适应学习机制调整更新步长,实现对基于合成基波法的冲击响应谱时域波形合成优化。结果表明,基于PSO-LSN算法的时域波形合成在迭代前期对决策域空间有着较好的全局搜索能力,随着迭代次数的增加,其局部精细搜索能力明显提升,可获得高精准度的仿真计算结果,有效验证了算法的准确性和实用性,可为进一步提升冲击响应谱时域波形合成计算精度提供支撑。Here,aiming at the problem of local out-of-tolerance caused by nonlinearity,local resonance and other factors affecting specimen control in shock response spectrum(SRS)tests and requiring multiple revisions and iterations of time-domain fundamental wave waveform parameters,a spatial neighborhood driven strategy particle swarm optimization particle swarm optimization based on learning spatial neighborhood driven(PSO-LSN)algorithm based on adaptive learning was proposed by analyzing main affecting factors and change mechanism of SRS test results.According to the particle neighborhood similarity,the local spatial search capability was enhanced to share the optimal position and velocity information,adjust update step size through adaptive learning mechanism,and realize optimization of SRS time-domain waveform synthesis based on the synthetic fundamental wave method.The results showed that the time-domain waveform synthesis based on PSO-LSN algorithm in early stage of iteration has better global search ability for decision domain space;with increase in iteration number,its local fine search ability is obviously improved to be able to obtain high-precision simulation calculation results,and effectively verify the correctness and practicality of the algorithm;it can provide a support for further improving accuracy of SRS time-domain waveform synthesis calculation.

关 键 词：冲击响应谱(SRS) 时域波形合成 粒子群算法(PSO) 空间邻域驱动策略 自适应学习机制 

分 类 号：TH212[机械工程—机械制造及自动化] TH213.3