摘要
为了预测炸药爆轰参数并提高预测精度,提出一种将Hessian局部线性映射算法(HLLE)和支持向量机理论(SVM)相结合的预测方法。选取16种传统混合炸药作为研究对象,利用Hessian局部线性映射算法理论,对原始数据进行降维处理,再将降维后的数据作为输入,应用SVM对炸药爆速、爆热做回归预测,采用遗传算法(GA)选取最佳惩罚因子、核函数参数,将预测结果与文献值对比分析,预测相对误差在±3%之内。最后将该方法应用于新型熔铸混合炸药2,4-二氨基-3,5-二硝基-1-氧吡嗪(LLM-105)/1-甲基-3,5-二硝基-1,2,4-三唑(DNMT)和LLM-105/1-甲基-4,5-二硝基咪唑(4,5-MDNI)爆速预测,与试验值做比较,相对误差分别为2.91%、3.72%。结果表明,该方法对混合炸药爆轰参数预测误差较小,模型精度较高。
In order to improve the prediction accuracy of the detonation parameters,a new method based on Support Vector M achine（ SVM） theory and Hessian Local Linearly Embedding algorithm（ HLLE） was proposed to predict 16 traditional composite explosives. The original data after dimension reduction w ith HLLE,w as input to regressively predict the heat and velocity of composite explosion by SVM.T he best kernel function parameter and penalty factor are selected by Genetic Algorithm（ GA）. T he calculated results of the explosives almost agree w ith those of the literature,and the relative error is w ithin ± 3%. U sing the method,the values of detonation velocity for 2,4-diamino-3,5-dinitro pyrazine-1-oxide /1-methyl-3,5-dinitro-1,2,4-triazole and 2,4-diamino-3,5-dinitro pyrazine-1-oxide /1-methyl-4,5-dinitroimidazole w ere predicted and compared w ith experimental and the relative errors are 2. 91% and 3. 72%,respectively,show ing that the proposed method is comparatively accurate.
出处
《含能材料》
EI
CAS
CSCD
北大核心
2014年第2期221-225,共5页
Chinese Journal of Energetic Materials
基金
火炸药国防专项资助
