基于综合灰关联序模型的残差门控循环神经网络位标器零部件选配  

Selective Assembly for Coordinator Parts by RNGRU Based on Comprehensive Grey Relational Order Model

作　　者：钟百鸿 王琳[1] 钟诗胜 ZHONG Baihong;WANG Lin;ZHONG Shisheng(School of Ocean Engineering,Harbin Institute of Technology at Weihai,Weihai,Shandong,264209)

机构地区：[1]哈尔滨工业大学(威海)海洋工程学院,威海264209

出　　处：《中国机械工程》2021年第3期314-320,356,共8页China Mechanical Engineering

基　　金：国家自然科学基金(51705100);中央高校基本科研业务费专项资金(HITNSRIF2019078)。

摘　　要：位标器漂移性能是决定精确制导武器跟踪能力与制导精度的关键因素之一,通过位标器零部件选配可改善位标器漂移性能。针对位标器零部件一次选配成功率低的问题,提出综合灰关联序(CGRO)模型并对影响位标器零部件选配的装配参数进行关联分析,得到了影响位标器零部件选配的关键装配参数;建立残差门控循环神经网络(RNGRU)模型,实现了位标器零部件的选配。以影响位标器漂移性能的陀螺转子与调漂螺钉装配为例,应用CGRO模型对陀螺转子的装配参数进行关联分析,得到了影响两者装配的关键装配参数;基于RNGRU模型实现了对调漂螺钉质量的回归预测。实验结果表明,所提方法能够实现位标器零部件的选配,其预测精度优于传统门控循环神经网络。The drift performance of coordinator,one of the key factors that determined the tracking ability and guidance accuracy of the precision-guided weapon directly,might be improved by the selective assembly for coordinator parts.In view of the low success rate of the first selective assembly for coordinator parts,a CGRO model was proposed to perform the relevancy on assembly parameters affecting the selective assembly for coordinator parts,and key assembly parameters affecting the selective assembly for coordinator parts were obtained.The RNGRU model was established to realize the selective assembly for coordinator parts.Taking the selective assembly between the gyro rotor and drift-adjusting screw which affected the drift performance of the coordinator as an example,CGRO was used to analyze the assembly parameters of the gyro rotor,and the key assembly parameters that affected the selective assembly of the two were obtained.Based on RNGRU,the regression prediction of the mass of the drift-adjusting screw was realized.The experimental results show that the proposed method may achieve the selective assembly for coordinator parts,and the prediction accuracy is superior to the traditional GRU network.

关 键 词：位标器 选配 综合灰关联序 残差门控循环神经网络 

分 类 号：TP205[自动化与计算机技术—检测技术与自动化装置]