基于优化算法的弹药贮存寿命预测方法  

Ammunition Storage Life Prediction Method Based on Optimization Algorithm

作　　者：冯昌林 邵渤涵 程雨森 FENG Chang-lin;SHAO Bo-han;CHENG Yu-sen(Unit 92942,People's Liberation Army,Beijing 102400,China;Systems Engineering Research Institute of CSSC,Beijing 100089,China;Naval University of Engineering,Wuhan 430033,China)

机构地区：[1]中国人民解放军92942部队,北京102400 [2]中国船舶集团有限公司系统工程研究院,北京100089 [3]海军工程大学,武汉430033

出　　处：《装备环境工程》2023年第1期8-15,共8页Equipment Environmental Engineering

摘　　要：目的 实现对缺失及不足的制导弹药贮存失效数据预测及补充的能力。方法 首先通过4种不同的预测算法(GA-BP、PSO-BP、GA-SVM、PSO-SVM),对自然贮存条件下弹药贮存失效数据进行预测,其次根据最小二乘拟合法,实现弹药贮存寿命评估模型的构建,再通过寿命评估模型,计算出不同方法下对应的贮存寿命。结果 通过不同模型的构建,4种预测方法与无优化条件下均能实现弹药贮存失效数据的预测,并且在规定可靠度,GA-BP和PSO-BP预测精度比另外2种方法更低。结论 GA-SVM与PSO-SVM更适合弹药贮存失效数据的预测,且效果更好。The work aims to achieve the ability to predict and supplement the missing and insufficient BM drug storage failure data. First, four different prediction algorithms(GA-BP, PSO-BP, GA-SVM, and PSO-SVM) were used to predict the storage failure data of ammunition under natural storage conditions. Second, the ammunition storage life assessment model was constructed according to the least squares. At last, the life assessment model was used to calculate the corresponding storage life under different methods. Prediction of ammunition storage failure data can be achieved by all four prediction methods and optimization-free conditions. And in the specified reliability, the accuracy of GA-BP and PSO-BP predictions is lower compared to the other two methods. GA-SVM and PSO-SVM are better suited to predicting ammunition storage failure data and are more effective.

关 键 词：弹药 贮存寿命 遗传算法 粒子群算法 BP神经网络 支持向量机 

分 类 号：TJ413[兵器科学与技术—火炮、自动武器与弹药工程]