基于IWOA-BP的火控计算机电源模块故障诊断方法  

Fault diagnosis method of power module of fire control computer based on IWOA-BP

作　　者：邵浩冬 李英顺 王德彪 佟维妍 SHAO Haodong;LI Yingshun;WANG Debiao;TONG Weiyan(School of Chemical Process Automation,Shenyang University of Technology,Shenyang 111003,China;School of Contral Science and Engineering,Dalian University of Technology,Dalian 116200,China;Shenyang Shunyi Technology Co.,Ltd.,Shenyang 110000,China)

机构地区：[1]沈阳工业大学化工过程自动化学院,沈阳111003 [2]大连理工大学控制科学与工程学院,辽宁大连116200 [3]沈阳顺义科技股份有限公司,沈阳110000

出　　处：《兵器装备工程学报》2025年第3期224-231,共8页Journal of Ordnance Equipment Engineering

基　　金：辽宁省科学技术计划项目(2022JH1/10400007)。

摘　　要：火控计算机是火控系统的核心,其对于火控系统的正常运行发挥着重要作用,因此对坦克火控计算机电源模块进行故障诊断是一项很重要的任务。为了提高诊断准确率和效率,引入了Sine-Tent-Cosine混沌映射和自适应惯性权重对原始的鲸鱼算法(WOA)进行改进与优化,利用改进后的算法对BP神经网络的权重、阈值进行参数寻优,构建了IWOA-BP火控计算机电源模块故障诊断模型,与PSO-BP、ANT-BP、WOA-BP几种诊断模型进行实验对比。多次实验结果表明:改进后的IWOA-BP模型在4种模型中效率最高,运行时间仅为8.72 s,在对火控计算机电源模块的5种故障进行诊断时,该模型的平均准确率达到了96.4%,相较于PSO-BP、ANT-BP和WOA-BP几种诊断模型准确率分别提升了3.65%、5.7%和5.93%。The fire control computer is the core of the fire control system,and it plays an important role in the normal operation of the fire control system,so it is a very important task to diagnose the power module of the tank fire control computer.To enhance both accuracy and efficiency of diagnosis,the Sine-Tent-Cosine chaos mapping and adaptive inertia weights were integrated into the original Whale Algorithm(WOA)to improve and optimize it.Subsequently,the optimized algorithm was employed to fine-tune the weights and thresholds of the BP neural network,resulting in the establishment of the IWOA-BP fire control computer power module fault diagnosis model.It was experimentally compared with several diagnostic models of WOA-BP,ANT-BP and PSO-BP.Numerous experimental results have shown that the improved IWOA-BP model is the most efficient among four models,with a running time of only 8.72 seconds.In diagnosing five types of faults in the fire control computer power module,the model achieved an average accuracy of 96.4%,which represents an increase of 3.65%,5.7%,and 5.93%in accuracy compared to the PSO-BP,ANT-BP,and WOA-BP diagnostic models,respectively.

关 键 词：故障诊断 鲸鱼优化算法 Sine-Tent-Cosine混沌映射 自适应惯性权重 BP神经网络 

分 类 号：TJ811.2[兵器科学与技术—武器系统与运用工程]