摘要
舰载机出动架次率作为衡量航母战斗力的关键指标,对航母舰载机系统的安全高效运行十分重要。建立根据实时数据预测当前出动架次率的模型,将会为航母指挥官的实时调度提供重要参考。首先,从指标原始数据出发,基于大数据关联度分析、社区发现及主成分分析法,确定指标之间的树状关系,从而建立稀疏深度神经网络。同时,为了保证更好的训练效果,选取标准化、L2正则化、Adam优化器作为神经网络的优化算法进行训练。仿真结果表明,在航母舰载机持续性出动任务下,所提方法能够实现对舰载机出动架次率的快速、准确、实时预测。
As a key indicator to measure the combat effectiveness of an aircraft carrier,the carrier aircraft’s sortie rate is very important for the safe and efficient operation of the carrier-based aircraft system.Establishing a model that predicts the current sortie rate based on real-time data will provide an important reference for the aircraft carrier commander’s real-time scheduling.Firstly,starting from the original data of indicators,based on big data correlation analysis,community discovery,and principal component analysis,the tree-like relationship between indicators is determined,so as to establish a sparse deep neural network.At the same time,in order to ensure better training effect,standardization,L2 regularization,and Adam optimizer are selected as the optimization algorithm of the neural network.The simulation results show that the proposed method can achieve fast,accurate and real-time prediction of the sortie rate of carrier aircraft under the mission of continuous dispatch.
作者
邓嘉宁
李海旭
安强林
沙恩来
王泽
吴宇
DENG Jianing;LI Haixu;AN Qianglin;SHA Enlai;WANG Ze;WU Yu(College of Aerospace Engineering,Chongqing University,Chongqing 400044,China;China Shipbuilding System Engineering Research Institute,Beijing 100094,China)
出处
《系统工程与电子技术》
EI
CSCD
北大核心
2023年第11期3515-3523,共9页
Systems Engineering and Electronics
