文献类型：期刊文献

中文题名：基于高斯混合和概率神经网络的舰船柴油机故障诊断方法

英文题名：Marine Diesel Engine Fault Diagnosis with Unbalance Samples Based on Gaussian Mixture Probabilistic Neural Network

作者：邱其清[1];廖志强[1]

机构：[1]广东海洋大学船舶与海运学院,广东湛江524088

年份：2022

卷号：44

期号：9

起止页码：101

中文期刊名：船舶工程

外文期刊名：Ship Engineering

收录：CSTPCD、、北大核心、CSCD、北大核心2020、CSCD_E2021_2022

基金：国家自然基金项目(52201355);广东海洋大学科研启动经费资助项目。

语种：中文

中文关键词：柴油机;故障诊断;样本不平衡;高斯混合模型;概率神经网络

外文关键词：diesel engine;fault diagnosis;unbalanced samples;gaussian mixture model(GMM);probabilistic neural network(PNN)

中文摘要：针对舰船柴油机智能故障诊断中因故障样本不足而导致的诊断模型准确度不高的问题,提出一种基于高斯混合概率神经网络的舰船柴油机不平衡样本故障诊断方法。首先,采用高斯混合模型(GMM)扩充故障样本,改善样本不平衡问题;其次,基于概率神经网络(PNN)建立柴油机故障分类模型,增强泛化性、提高诊断准确度;最后,工程试验和对比试验证明:研究提出的方法能够准确识别舰船柴油机故障,具有诊断精度高、运行时间短和高泛化性等优点。

外文摘要：Aiming at the problem that the low accuracy of marine diesel engines diagnosis due to insufficient fault samples in the intelligent fault diagnosis model,a fault diagnosis method based on Gaussian mixture model(GMM)and probabilistic neural network(PNN)is proposed.Firstly,Gaussian mixture model(GMM)is used to expand fault samples and improve the sample imbalance problem.Secondly,based on the PNN,the diesel engine fault classification model is established to enhance the generalization and improve diagnosis accuracy.Finally,the engineering experiments and comparative experiments prove that the method can accurately identify marine diesel engine faults,which has the advantages of high diagnostic accuracy,short running time and high generalization.